Datasets:

sugam11

/

covid-qa

like 0

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

How safe is MVA?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is the status of MVA influenza vaccine?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is NYVAC?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

How is NYVAC grown?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

How safe is NYVAC?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What would limit the use of poxvirus vectored vaccines?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

what is the advantage of the NYVAC as an influenza virus?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

Where is poxvirus vaccine being used?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What have the studies on NP shown for the protection against influenza challege?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is the goal of vaccine?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What has enabled the development of one size fits all vaccine?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

Why is a revision of current vaccines is needed?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is an example of an improved vaccine regime?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What can provide an improved vaccine regime?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What features can be created for creating vectored vaccines?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

How can sustained immunity be generated?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is the advantage of vectored vaccines?

Virus-Vectored Influenza Virus Vaccines https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4147686/ SHA: f6d2afb2ec44d8656972ea79f8a833143bbeb42b Authors: Tripp, Ralph A.; Tompkins, S. Mark Date: 2014-08-07 DOI: 10.3390/v6083055 License: cc-by Abstract: Despite the availability of an inactivated vaccine that has been licensed for >50 years, the influenza virus continues to cause morbidity and mortality worldwide. Constant evolution of circulating influenza virus strains and the emergence of new strains diminishes the effectiveness of annual vaccines that rely on a match with circulating influenza strains. Thus, there is a continued need for new, efficacious vaccines conferring cross-clade protection to avoid the need for biannual reformulation of seasonal influenza vaccines. Recombinant virus-vectored vaccines are an appealing alternative to classical inactivated vaccines because virus vectors enable native expression of influenza antigens, even from virulent influenza viruses, while expressed in the context of the vector that can improve immunogenicity. In addition, a vectored vaccine often enables delivery of the vaccine to sites of inductive immunity such as the respiratory tract enabling protection from influenza virus infection. Moreover, the ability to readily manipulate virus vectors to produce novel influenza vaccines may provide the quickest path toward a universal vaccine protecting against all influenza viruses. This review will discuss experimental virus-vectored vaccines for use in humans, comparing them to licensed vaccines and the hurdles faced for licensure of these next-generation influenza virus vaccines. Text: Seasonal influenza is a worldwide health problem causing high mobility and substantial mortality [1] [2] [3] [4] . Moreover, influenza infection often worsens preexisting medical conditions [5] [6] [7] . Vaccines against circulating influenza strains are available and updated annually, but many issues are still present, including low efficacy in the populations at greatest risk of complications from influenza virus infection, i.e., the young and elderly [8, 9] . Despite increasing vaccination rates, influenza-related hospitalizations are increasing [8, 10] , and substantial drug resistance has developed to two of the four currently approved anti-viral drugs [11, 12] . While adjuvants have the potential to improve efficacy and availability of current inactivated vaccines, live-attenuated and virus-vectored vaccines are still considered one of the best options for the induction of broad and efficacious immunity to the influenza virus [13] . The general types of influenza vaccines available in the United States are trivalent inactivated influenza vaccine (TIV), quadrivalent influenza vaccine (QIV), and live attenuated influenza vaccine (LAIV; in trivalent and quadrivalent forms). There are three types of inactivated vaccines that include whole virus inactivated, split virus inactivated, and subunit vaccines. In split virus vaccines, the virus is disrupted by a detergent. In subunit vaccines, HA and NA have been further purified by removal of other viral components. TIV is administered intramuscularly and contains three or four inactivated viruses, i.e., two type A strains (H1 and H3) and one or two type B strains. TIV efficacy is measured by induction of humoral responses to the hemagglutinin (HA) protein, the major surface and attachment glycoprotein on influenza. Serum antibody responses to HA are measured by the hemagglutination-inhibition (HI) assay, and the strain-specific HI titer is considered the gold-standard correlate of immunity to influenza where a four-fold increase in titer post-vaccination, or a HI titer of ≥1:40 is considered protective [4, 14] . Protection against clinical disease is mainly conferred by serum antibodies; however, mucosal IgA antibodies also may contribute to resistance against infection. Split virus inactivated vaccines can induce neuraminidase (NA)-specific antibody responses [15] [16] [17] , and anti-NA antibodies have been associated with protection from infection in humans [18] [19] [20] [21] [22] . Currently, NA-specific antibody responses are not considered a correlate of protection [14] . LAIV is administered as a nasal spray and contains the same three or four influenza virus strains as inactivated vaccines but on an attenuated vaccine backbone [4] . LAIV are temperature-sensitive and cold-adapted so they do not replicate effectively at core body temperature, but replicate in the mucosa of the nasopharynx [23] . LAIV immunization induces serum antibody responses, mucosal antibody responses (IgA), and T cell responses. While robust serum antibody and nasal wash (mucosal) antibody responses are associated with protection from infection, other immune responses, such as CD8 + cytotoxic lymphocyte (CTL) responses may contribute to protection and there is not a clear correlate of immunity for LAIV [4, 14, 24] . Currently licensed influenza virus vaccines suffer from a number of issues. The inactivated vaccines rely on specific antibody responses to the HA, and to a lesser extent NA proteins for protection. The immunodominant portions of the HA and NA molecules undergo a constant process of antigenic drift, a natural accumulation of mutations, enabling virus evasion from immunity [9, 25] . Thus, the circulating influenza A and B strains are reviewed annually for antigenic match with current vaccines, Replacement of vaccine strains may occur regularly, and annual vaccination is recommended to assure protection [4, 26, 27] . For the northern hemisphere, vaccine strain selection occurs in February and then manufacturers begin production, taking at least six months to produce the millions of vaccine doses required for the fall [27] . If the prediction is imperfect, or if manufacturers have issues with vaccine production, vaccine efficacy or availability can be compromised [28] . LAIV is not recommended for all populations; however, it is generally considered to be as effective as inactivated vaccines and may be more efficacious in children [4, 9, 24] . While LAIV relies on antigenic match and the HA and NA antigens are replaced on the same schedule as the TIV [4, 9] , there is some suggestion that LAIV may induce broader protection than TIV due to the diversity of the immune response consistent with inducing virus-neutralizing serum and mucosal antibodies, as well as broadly reactive T cell responses [9, 23, 29] . While overall both TIV and LAIV are considered safe and effective, there is a recognized need for improved seasonal influenza vaccines [26] . Moreover, improved understanding of immunity to conserved influenza virus antigens has raised the possibility of a universal vaccine, and these universal antigens will likely require novel vaccines for effective delivery [30] [31] [32] . Virus-vectored vaccines share many of the advantages of LAIV, as well as those unique to the vectors. Recombinant DNA systems exist that allow ready manipulation and modification of the vector genome. This in turn enables modification of the vectors to attenuate the virus or enhance immunogenicity, in addition to adding and manipulating the influenza virus antigens. Many of these vectors have been extensively studied or used as vaccines against wild type forms of the virus. Finally, each of these vaccine vectors is either replication-defective or causes a self-limiting infection, although like LAIV, safety in immunocompromised individuals still remains a concern [4, 13, [33] [34] [35] . Table 1 summarizes the benefits and concerns of each of the virus-vectored vaccines discussed here. There are 53 serotypes of adenovirus, many of which have been explored as vaccine vectors. A live adenovirus vaccine containing serotypes 4 and 7 has been in use by the military for decades, suggesting adenoviruses may be safe for widespread vaccine use [36] . However, safety concerns have led to the majority of adenovirus-based vaccine development to focus on replication-defective vectors. Adenovirus 5 (Ad5) is the most-studied serotype, having been tested for gene delivery and anti-cancer agents, as well as for infectious disease vaccines. Adenovirus vectors are attractive as vaccine vectors because their genome is very stable and there are a variety of recombinant systems available which can accommodate up to 10 kb of recombinant genetic material [37] . Adenovirus is a non-enveloped virus which is relatively stable and can be formulated for long-term storage at 4 °C, or even storage up to six months at room temperature [33] . Adenovirus vaccines can be grown to high titers, exceeding 10 1° plaque forming units (PFU) per mL when cultured on 293 or PER.C6 cells [38] , and the virus can be purified by simple methods [39] . Adenovirus vaccines can also be delivered via multiple routes, including intramuscular injection, subcutaneous injection, intradermal injection, oral delivery using a protective capsule, and by intranasal delivery. Importantly, the latter two delivery methods induce robust mucosal immune responses and may bypass preexisting vector immunity [33] . Even replication-defective adenovirus vectors are naturally immunostimulatory and effective adjuvants to the recombinant antigen being delivered. Adenovirus has been extensively studied as a vaccine vector for human disease. The first report using adenovirus as a vaccine vector for influenza demonstrated immunogenicity of recombinant adenovirus 5 (rAd5) expressing the HA of a swine influenza virus, A/Swine/Iowa/1999 (H3N2). Intramuscular immunization of mice with this construct induced robust neutralizing antibody responses and protected mice from challenge with a heterologous virus, A/Hong Kong/1/1968 (H3N2) [40] . Replication defective rAd5 vaccines expressing influenza HA have also been tested in humans. A rAd5-HA expressing the HA from A/Puerto Rico/8/1934 (H1N1; PR8) was delivered to humans epicutaneously or intranasally and assayed for safety and immunogenicity. The vaccine was well tolerated and induced seroconversion with the intranasal administration had a higher conversion rate and higher geometric meant HI titers [41] . While clinical trials with rAd vectors have overall been successful, demonstrating safety and some level of efficacy, rAd5 as a vector has been negatively overshadowed by two clinical trial failures. The first trial was a gene therapy examination where high-dose intravenous delivery of an Ad vector resulted in the death of an 18-year-old male [42, 43] . The second clinical failure was using an Ad5-vectored HIV vaccine being tested as a part of a Step Study, a phase 2B clinical trial. In this study, individuals were vaccinated with the Ad5 vaccine vector expressing HIV-1 gag, pol, and nef genes. The vaccine induced HIV-specific T cell responses; however, the study was stopped after interim analysis suggested the vaccine did not achieve efficacy and individuals with high preexisting Ad5 antibody titers might have an increased risk of acquiring HIV-1 [44] [45] [46] . Subsequently, the rAd5 vaccine-associated risk was confirmed [47] . While these two instances do not suggest Ad-vector vaccines are unsafe or inefficacious, the umbra cast by the clinical trials notes has affected interest for all adenovirus vaccines, but interest still remains. Immunization with adenovirus vectors induces potent cellular and humoral immune responses that are initiated through toll-like receptor-dependent and independent pathways which induce robust pro-inflammatory cytokine responses. Recombinant Ad vaccines expressing HA antigens from pandemic H1N1 (pH1N1), H5 and H7 highly pathogenic avian influenza (HPAI) virus (HPAIV), and H9 avian influenza viruses have been tested for efficacy in a number of animal models, including chickens, mice, and ferrets, and been shown to be efficacious and provide protection from challenge [48, 49] . Several rAd5 vectors have been explored for delivery of non-HA antigens, influenza nucleoprotein (NP) and matrix 2 (M2) protein [29, [50] [51] [52] . The efficacy of non-HA antigens has led to their inclusion with HA-based vaccines to improve immunogenicity and broaden breadth of both humoral and cellular immunity [53, 54] . However, as both CD8 + T cell and neutralizing antibody responses are generated by the vector and vaccine antigens, immunological memory to these components can reduce efficacy and limit repeated use [48] . One drawback of an Ad5 vector is the potential for preexisting immunity, so alternative adenovirus serotypes have been explored as vectors, particularly non-human and uncommon human serotypes. Non-human adenovirus vectors include those from non-human primates (NHP), dogs, sheep, pigs, cows, birds and others [48, 55] . These vectors can infect a variety of cell types, but are generally attenuated in humans avoiding concerns of preexisting immunity. Swine, NHP and bovine adenoviruses expressing H5 HA antigens have been shown to induce immunity comparable to human rAd5-H5 vaccines [33, 56] . Recombinant, replication-defective adenoviruses from low-prevalence serotypes have also been shown to be efficacious. Low prevalence serotypes such as adenovirus types 3, 7, 11, and 35 can evade anti-Ad5 immune responses while maintaining effective antigen delivery and immunogenicity [48, 57] . Prime-boost strategies, using DNA or protein immunization in conjunction with an adenovirus vaccine booster immunization have also been explored as a means to avoided preexisting immunity [52] . Adeno-associated viruses (AAV) were first explored as gene therapy vectors. Like rAd vectors, rAAV have broad tropism infecting a variety of hosts, tissues, and proliferating and non-proliferating cell types [58] . AAVs had been generally not considered as vaccine vectors because they were widely considered to be poorly immunogenic. A seminal study using AAV-2 to express a HSV-2 glycoprotein showed this virus vaccine vector effectively induced potent CD8 + T cell and serum antibody responses, thereby opening the door to other rAAV vaccine-associated studies [59, 60] . AAV vector systems have a number of engaging properties. The wild type viruses are non-pathogenic and replication incompetent in humans and the recombinant AAV vector systems are even further attenuated [61] . As members of the parvovirus family, AAVs are small non-enveloped viruses that are stable and amenable to long-term storage without a cold chain. While there is limited preexisting immunity, availability of non-human strains as vaccine candidates eliminates these concerns. Modifications to the vector have increased immunogenicity, as well [60] . There are limited studies using AAVs as vaccine vectors for influenza. An AAV expressing an HA antigen was first shown to induce protective in 2001 [62] . Later, a hybrid AAV derived from two non-human primate isolates (AAVrh32.33) was used to express influenza NP and protect against PR8 challenge in mice [63] . Most recently, following the 2009 H1N1 influenza virus pandemic, rAAV vectors were generated expressing the HA, NP and matrix 1 (M1) proteins of A/Mexico/4603/2009 (pH1N1), and in murine immunization and challenge studies, the rAAV-HA and rAAV-NP were shown to be protective; however, mice vaccinated with rAAV-HA + NP + M1 had the most robust protection. Also, mice vaccinated with rAAV-HA + rAAV-NP + rAAV-M1 were also partially protected against heterologous (PR8, H1N1) challenge [63] . Most recently, an AAV vector was used to deliver passive immunity to influenza [64, 65] . In these studies, AAV (AAV8 and AAV9) was used to deliver an antibody transgene encoding a broadly cross-protective anti-influenza monoclonal antibody for in vivo expression. Both intramuscular and intranasal delivery of the AAVs was shown to protect against a number of influenza virus challenges in mice and ferrets, including H1N1 and H5N1 viruses [64, 65] . These studies suggest that rAAV vectors are promising vaccine and immunoprophylaxis vectors. To this point, while approximately 80 phase I, I/II, II, or III rAAV clinical trials are open, completed, or being reviewed, these have focused upon gene transfer studies and so there is as yet limited safety data for use of rAAV as vaccines [66] . Alphaviruses are positive-sense, single-stranded RNA viruses of the Togaviridae family. A variety of alphaviruses have been developed as vaccine vectors, including Semliki Forest virus (SFV), Sindbis (SIN) virus, Venezuelan equine encephalitis (VEE) virus, as well as chimeric viruses incorporating portions of SIN and VEE viruses. The replication defective vaccines or replicons do not encode viral structural proteins, having these portions of the genome replaces with transgenic material. The structural proteins are provided in cell culture production systems. One important feature of the replicon systems is the self-replicating nature of the RNA. Despite the partial viral genome, the RNAs are self-replicating and can express transgenes at very high levels [67] . SIN, SFV, and VEE have all been tested for efficacy as vaccine vectors for influenza virus [68] [69] [70] [71] . A VEE-based replicon system encoding the HA from PR8 was demonstrated to induce potent HA-specific immune response and protected from challenge in a murine model, despite repeated immunization with the vector expressing a control antigen, suggesting preexisting immunity may not be an issue for the replicon vaccine [68] . A separate study developed a VEE replicon system expressing the HA from A/Hong Kong/156/1997 (H5N1) and demonstrated varying efficacy after in ovo vaccination or vaccination of 1-day-old chicks [70] . A recombinant SIN virus was use as a vaccine vector to deliver a CD8 + T cell epitope only. The well-characterized NP epitope was transgenically expressed in the SIN system and shown to be immunogenic in mice, priming a robust CD8 + T cell response and reducing influenza virus titer after challenge [69] . More recently, a VEE replicon system expressing the HA protein of PR8 was shown to protect young adult (8-week-old) and aged (12-month-old) mice from lethal homologous challenge [72] . The VEE replicon systems are particularly appealing as the VEE targets antigen-presenting cells in the lymphatic tissues, priming rapid and robust immune responses [73] . VEE replicon systems can induce robust mucosal immune responses through intranasal or subcutaneous immunization [72] [73] [74] , and subcutaneous immunization with virus-like replicon particles (VRP) expressing HA-induced antigen-specific systemic IgG and fecal IgA antibodies [74] . VRPs derived from VEE virus have been developed as candidate vaccines for cytomegalovirus (CMV). A phase I clinical trial with the CMV VRP showed the vaccine was immunogenic, inducing CMV-neutralizing antibody responses and potent T cell responses. Moreover, the vaccine was well tolerated and considered safe [75] . A separate clinical trial assessed efficacy of repeated immunization with a VRP expressing a tumor antigen. The vaccine was safe and despite high vector-specific immunity after initial immunization, continued to boost transgene-specific immune responses upon boost [76] . While additional clinical data is needed, these reports suggest alphavirus replicon systems or VRPs may be safe and efficacious, even in the face of preexisting immunity. Baculovirus has been extensively used to produce recombinant proteins. Recently, a baculovirus-derived recombinant HA vaccine was approved for human use and was first available for use in the United States for the 2013-2014 influenza season [4] . Baculoviruses have also been explored as vaccine vectors. Baculoviruses have a number of advantages as vaccine vectors. The viruses have been extensively studied for protein expression and for pesticide use and so are readily manipulated. The vectors can accommodate large gene insertions, show limited cytopathic effect in mammalian cells, and have been shown to infect and express genes of interest in a spectrum of mammalian cells [77] . While the insect promoters are not effective for mammalian gene expression, appropriate promoters can be cloned into the baculovirus vaccine vectors. Baculovirus vectors have been tested as influenza vaccines, with the first reported vaccine using Autographa californica nuclear polyhedrosis virus (AcNPV) expressing the HA of PR8 under control of the CAG promoter (AcCAG-HA) [77] . Intramuscular, intranasal, intradermal, and intraperitoneal immunization or mice with AcCAG-HA elicited HA-specific antibody responses, however only intranasal immunization provided protection from lethal challenge. Interestingly, intranasal immunization with the wild type AcNPV also resulted in protection from PR8 challenge. The robust innate immune response to the baculovirus provided non-specific protection from subsequent influenza virus infection [78] . While these studies did not demonstrate specific protection, there were antigen-specific immune responses and potential adjuvant effects by the innate response. Baculovirus pseudotype viruses have also been explored. The G protein of vesicular stomatitis virus controlled by the insect polyhedron promoter and the HA of A/Chicken/Hubei/327/2004 (H5N1) HPAIV controlled by a CMV promoter were used to generate the BV-G-HA. Intramuscular immunization of mice or chickens with BV-G-HA elicited strong HI and VN serum antibody responses, IFN-γ responses, and protected from H5N1 challenge [79] . A separate study demonstrated efficacy using a bivalent pseudotyped baculovirus vector [80] . Baculovirus has also been used to generate an inactivated particle vaccine. The HA of A/Indonesia/CDC669/2006(H5N1) was incorporated into a commercial baculovirus vector controlled by the e1 promoter from White Spot Syndrome Virus. The resulting recombinant virus was propagated in insect (Sf9) cells and inactivated as a particle vaccine [81, 82] . Intranasal delivery with cholera toxin B as an adjuvant elicited robust HI titers and protected from lethal challenge [81] . Oral delivery of this encapsulated vaccine induced robust serum HI titers and mucosal IgA titers in mice, and protected from H5N1 HPAIV challenge. More recently, co-formulations of inactivated baculovirus vectors have also been shown to be effective in mice [83] . While there is growing data on the potential use of baculovirus or pseudotyped baculovirus as a vaccine vector, efficacy data in mammalian animal models other than mice is lacking. There is also no data on the safety in humans, reducing enthusiasm for baculovirus as a vaccine vector for influenza at this time. Newcastle disease virus (NDV) is a single-stranded, negative-sense RNA virus that causes disease in poultry. NDV has a number of appealing qualities as a vaccine vector. As an avian virus, there is little or no preexisting immunity to NDV in humans and NDV propagates to high titers in both chicken eggs and cell culture. As a paramyxovirus, there is no DNA phase in the virus lifecycle reducing concerns of integration events, and the levels of gene expression are driven by the proximity to the leader sequence at the 3' end of the viral genome. This gradient of gene expression enables attenuation through rearrangement of the genome, or by insertion of transgenes within the genome. Finally, pathogenicity of NDV is largely determined by features of the fusion protein enabling ready attenuation of the vaccine vector [84] . Reverse genetics, a method that allows NDV to be rescued from plasmids expressing the viral RNA polymerase and nucleocapsid proteins, was first reported in 1999 [85, 86] . This process has enabled manipulation of the NDV genome as well as incorporation of transgenes and the development of NDV vectors. Influenza was the first infectious disease targeted with a recombinant NDV (rNDV) vector. The HA protein of A/WSN/1933 (H1N1) was inserted into the Hitchner B1 vaccine strain. The HA protein was expressed on infected cells and was incorporated into infectious virions. While the virus was attenuated compared to the parental vaccine strain, it induced a robust serum antibody response and protected against homologous influenza virus challenge in a murine model of infection [87] . Subsequently, rNDV was tested as a vaccine vector for HPAIV having varying efficacy against H5 and H7 influenza virus infections in poultry [88] [89] [90] [91] [92] [93] [94] . These vaccines have the added benefit of potentially providing protection against both the influenza virus and NDV infection. NDV has also been explored as a vaccine vector for humans. Two NHP studies assessed the immunogenicity and efficacy of an rNDV expressing the HA or NA of A/Vietnam/1203/2004 (H5N1; VN1203) [95, 96] . Intranasal and intratracheal delivery of the rNDV-HA or rNDV-NA vaccines induced both serum and mucosal antibody responses and protected from HPAIV challenge [95, 96] . NDV has limited clinical data; however, phase I and phase I/II clinical trials have shown that the NDV vector is well-tolerated, even at high doses delivered intravenously [44, 97] . While these results are promising, additional studies are needed to advance NDV as a human vaccine vector for influenza. Parainfluenza virus type 5 (PIV5) is a paramyxovirus vaccine vector being explored for delivery of influenza and other infectious disease vaccine antigens. PIV5 has only recently been described as a vaccine vector [98] . Similar to other RNA viruses, PIV5 has a number of features that make it an attractive vaccine vector. For example, PIV5 has a stable RNA genome and no DNA phase in virus replication cycle reducing concerns of host genome integration or modification. PIV5 can be grown to very high titers in mammalian vaccine cell culture substrates and is not cytopathic allowing for extended culture and harvest of vaccine virus [98, 99] . Like NDV, PIV5 has a 3'-to 5' gradient of gene expression and insertion of transgenes at different locations in the genome can variably attenuate the virus and alter transgene expression [100] . PIV5 has broad tropism, infecting many cell types, tissues, and species without causing clinical disease, although PIV5 has been associated with -kennel cough‖ in dogs [99] . A reverse genetics system for PIV5 was first used to insert the HA gene from A/Udorn/307/72 (H3N2) into the PIV5 genome between the hemagglutinin-neuraminidase (HN) gene and the large (L) polymerase gene. Similar to NDV, the HA was expressed at high levels in infected cells and replicated similarly to the wild type virus, and importantly, was not pathogenic in immunodeficient mice [98] . Additionally, a single intranasal immunization in a murine model of influenza infection was shown to induce neutralizing antibody responses and protect against a virus expressing homologous HA protein [98] . PIV5 has also been explored as a vaccine against HPAIV. Recombinant PIV5 vaccines expressing the HA or NP from VN1203 were tested for efficacy in a murine challenge model. Mice intranasally vaccinated with a single dose of PIV5-H5 vaccine had robust serum and mucosal antibody responses, and were protected from lethal challenge. Notably, although cellular immune responses appeared to contribute to protection, serum antibody was sufficient for protection from challenge [100, 101] . Intramuscular immunization with PIV5-H5 was also shown to be effective at inducing neutralizing antibody responses and protecting against lethal influenza virus challenge [101] . PIV5 expressing the NP protein of HPAIV was also efficacious in the murine immunization and challenge model, where a single intranasal immunization induced robust CD8 + T cell responses and protected against homologous (H5N1) and heterosubtypic (H1N1) virus challenge [102] . Currently there is no clinical safety data for use of PIV5 in humans. However, live PIV5 has been a component of veterinary vaccines for -kennel cough‖ for >30 years, and veterinarians and dog owners are exposed to live PIV5 without reported disease [99] . This combined with preclinical data from a variety of animal models suggests that PIV5 as a vector is likely to be safe in humans. As preexisting immunity is a concern for all virus-vectored vaccines, it should be noted that there is no data on the levels of preexisting immunity to PIV5 in humans. However, a study evaluating the efficacy of a PIV5-H3 vaccine in canines previously vaccinated against PIV5 (kennel cough) showed induction of robust anti-H3 serum antibody responses as well as high serum antibody levels to the PIV5 vaccine, suggesting preexisting immunity to the PIV5 vector may not affect immunogenicity of vaccines even with repeated use [99] . Poxvirus vaccines have a long history and the notable hallmark of being responsible for eradication of smallpox. The termination of the smallpox virus vaccination program has resulted in a large population of poxvirus-naï ve individuals that provides the opportunity for the use of poxviruses as vectors without preexisting immunity concerns [103] . Poxvirus-vectored vaccines were first proposed for use in 1982 with two reports of recombinant vaccinia viruses encoding and expressing functional thymidine kinase gene from herpes virus [104, 105] . Within a year, a vaccinia virus encoding the HA of an H2N2 virus was shown to express a functional HA protein (cleaved in the HA1 and HA2 subunits) and be immunogenic in rabbits and hamsters [106] . Subsequently, all ten of the primary influenza proteins have been expressed in vaccine virus [107] . Early work with intact vaccinia virus vectors raised safety concerns, as there was substantial reactogenicity that hindered recombinant vaccine development [108] . Two vaccinia vectors were developed to address these safety concerns. The modified vaccinia virus Ankara (MVA) strain was attenuated by passage 530 times in chick embryo fibroblasts cultures. The second, New York vaccinia virus (NYVAC) was a plaque-purified clone of the Copenhagen vaccine strain rationally attenuated by deletion of 18 open reading frames [109] [110] [111] . Modified vaccinia virus Ankara (MVA) was developed prior to smallpox eradication to reduce or prevent adverse effects of other smallpox vaccines [109] . Serial tissue culture passage of MVA resulted in loss of 15% of the genome, and established a growth restriction for avian cells. The defects affected late stages in virus assembly in non-avian cells, a feature enabling use of the vector as single-round expression vector in non-permissive hosts. Interestingly, over two decades ago, recombinant MVA expressing the HA and NP of influenza virus was shown to be effective against lethal influenza virus challenge in a murine model [112] . Subsequently, MVA expressing various antigens from seasonal, pandemic (A/California/04/2009, pH1N1), equine (A/Equine/Kentucky/1/81 H3N8), and HPAI (VN1203) viruses have been shown to be efficacious in murine, ferret, NHP, and equine challenge models [113] . MVA vaccines are very effective stimulators of both cellular and humoral immunity. For example, abortive infection provides native expression of the influenza antigens enabling robust antibody responses to native surface viral antigens. Concurrently, the intracellular influenza peptides expressed by the pox vector enter the class I MHC antigen processing and presentation pathway enabling induction of CD8 + T cell antiviral responses. MVA also induces CD4 + T cell responses further contributing to the magnitude of the antigen-specific effector functions [107, [112] [113] [114] [115] . MVA is also a potent activator of early innate immune responses further enhancing adaptive immune responses [116] . Between early smallpox vaccine development and more recent vaccine vector development, MVA has undergone extensive safety testing and shown to be attenuated in severely immunocompromised animals and safe for use in children, adults, elderly, and immunocompromised persons. With extensive pre-clinical data, recombinant MVA vaccines expressing influenza antigens have been tested in clinical trials and been shown to be safe and immunogenic in humans [117] [118] [119] . These results combined with data from other (non-influenza) clinical and pre-clinical studies support MVA as a leading viral-vectored candidate vaccine. The NYVAC vector is a highly attenuated vaccinia virus strain. NYVAC is replication-restricted; however, it grows in chick embryo fibroblasts and Vero cells enabling vaccine-scale production. In non-permissive cells, critical late structural proteins are not produced stopping replication at the immature virion stage [120] . NYVAC is very attenuated and considered safe for use in humans of all ages; however, it predominantly induces a CD4 + T cell response which is different compared to MVA [114] . Both MVA and NYVAC provoke robust humoral responses, and can be delivered mucosally to induce mucosal antibody responses [121] . There has been only limited exploration of NYVAC as a vaccine vector for influenza virus; however, a vaccine expressing the HA from A/chicken/Indonesia/7/2003 (H5N1) was shown to induce potent neutralizing antibody responses and protect against challenge in swine [122] . While there is strong safety and efficacy data for use of NYVAC or MVA-vectored influenza vaccines, preexisting immunity remains a concern. Although the smallpox vaccination campaign has resulted in a population of poxvirus-naï ve people, the initiation of an MVA or NYVAC vaccination program for HIV, influenza or other pathogens will rapidly reduce this susceptible population. While there is significant interest in development of pox-vectored influenza virus vaccines, current influenza vaccination strategies rely upon regular immunization with vaccines matched to circulating strains. This would likely limit the use and/or efficacy of poxvirus-vectored influenza virus vaccines for regular and seasonal use [13] . Intriguingly, NYVAC may have an advantage for use as an influenza vaccine vector, because immunization with this vector induces weaker vaccine-specific immune responses compared to other poxvirus vaccines, a feature that may address the concerns surrounding preexisting immunity [123] . While poxvirus-vectored vaccines have not yet been approved for use in humans, there is a growing list of licensed poxvirus for veterinary use that include fowlpox-and canarypox-vectored vaccines for avian and equine influenza viruses, respectively [124, 125] . The fowlpox-vectored vaccine expressing the avian influenza virus HA antigen has the added benefit of providing protection against fowlpox infection. Currently, at least ten poxvirus-vectored vaccines have been licensed for veterinary use [126] . These poxvirus vectors have the potential for use as vaccine vectors in humans, similar to the first use of cowpox for vaccination against smallpox [127] . The availability of these non-human poxvirus vectors with extensive animal safety and efficacy data may address the issues with preexisting immunity to the human vaccine strains, although the cross-reactivity originally described with cowpox could also limit use. Influenza vaccines utilizing vesicular stomatitis virus (VSV), a rhabdovirus, as a vaccine vector have a number of advantages shared with other RNA virus vaccine vectors. Both live and replication-defective VSV vaccine vectors have been shown to be immunogenic [128, 129] , and like Paramyxoviridae, the Rhabdoviridae genome has a 3'-to-5' gradient of gene expression enabling attention by selective vaccine gene insertion or genome rearrangement [130] . VSV has a number of other advantages including broad tissue tropism, and the potential for intramuscular or intranasal immunization. The latter delivery method enables induction of mucosal immunity and elimination of needles required for vaccination. Also, there is little evidence of VSV seropositivity in humans eliminating concerns of preexisting immunity, although repeated use may be a concern. Also, VSV vaccine can be produced using existing mammalian vaccine manufacturing cell lines. Influenza antigens were first expressed in a VSV vector in 1997. Both the HA and NA were shown to be expressed as functional proteins and incorporated into the recombinant VSV particles [131] . Subsequently, VSV-HA, expressing the HA protein from A/WSN/1933 (H1N1) was shown to be immunogenic and protect mice from lethal influenza virus challenge [129] . To reduce safety concerns, attenuated VSV vectors were developed. One candidate vaccine had a truncated VSV G protein, while a second candidate was deficient in G protein expression and relied on G protein expressed by a helper vaccine cell line to the provide the virus receptor. Both vectors were found to be attenuated in mice, but maintained immunogenicity [128] . More recently, single-cycle replicating VSV vaccines have been tested for efficacy against H5N1 HPAIV. VSV vectors expressing the HA from A/Hong Kong/156/97 (H5N1) were shown to be immunogenic and induce cross-reactive antibody responses and protect against challenge with heterologous H5N1 challenge in murine and NHP models [132] [133] [134] . VSV vectors are not without potential concerns. VSV can cause disease in a number of species, including humans [135] . The virus is also potentially neuroinvasive in some species [136] , although NHP studies suggest this is not a concern in humans [137] . Also, while the incorporation of the influenza antigen in to the virion may provide some benefit in immunogenicity, changes in tropism or attenuation could arise from incorporation of different influenza glycoproteins. There is no evidence for this, however [134] . Currently, there is no human safety data for VSV-vectored vaccines. While experimental data is promising, additional work is needed before consideration for human influenza vaccination. Current influenza vaccines rely on matching the HA antigen of the vaccine with circulating strains to provide strain-specific neutralizing antibody responses [4, 14, 24] . There is significant interest in developing universal influenza vaccines that would not require annual reformulation to provide protective robust and durable immunity. These vaccines rely on generating focused immune responses to highly conserved portions of the virus that are refractory to mutation [30] [31] [32] . Traditional vaccines may not be suitable for these vaccination strategies; however, vectored vaccines that have the ability to be readily modified and to express transgenes are compatible for these applications. The NP and M2 proteins have been explored as universal vaccine antigens for decades. Early work with recombinant viral vectors demonstrated that immunization with vaccines expressing influenza antigens induced potent CD8 + T cell responses [107, [138] [139] [140] [141] . These responses, even to the HA antigen, could be cross-protective [138] . A number of studies have shown that immunization with NP expressed by AAV, rAd5, alphavirus vectors, MVA, or other vector systems induces potent CD8 + T cell responses and protects against influenza virus challenge [52, 63, 69, 102, 139, 142] . As the NP protein is highly conserved across influenza A viruses, NP-specific T cells can protect against heterologous and even heterosubtypic virus challenges [30] . The M2 protein is also highly conserved and expressed on the surface of infected cells, although to a lesser extent on the surface of virus particles [30] . Much of the vaccine work in this area has focused on virus-like or subunit particles expressing the M2 ectodomain; however, studies utilizing a DNA-prime, rAd-boost strategies to vaccinate against the entire M2 protein have shown the antigen to be immunogenic and protective [50] . In these studies, antibodies to the M2 protein protected against homologous and heterosubtypic challenge, including a H5N1 HPAIV challenge. More recently, NP and M2 have been combined to induce broadly cross-reactive CD8 + T cell and antibody responses, and rAd5 vaccines expressing these antigens have been shown to protect against pH1N1 and H5N1 challenges [29, 51] . Historically, the HA has not been widely considered as a universal vaccine antigen. However, the recent identification of virus neutralizing monoclonal antibodies that cross-react with many subtypes of influenza virus [143] has presented the opportunity to design vaccine antigens to prime focused antibody responses to the highly conserved regions recognized by these monoclonal antibodies. The majority of these broadly cross-reactive antibodies recognize regions on the stalk of the HA protein [143] . The HA stalk is generally less immunogenic compared to the globular head of the HA protein so most approaches have utilized -headless‖ HA proteins as immunogens. HA stalk vaccines have been designed using DNA and virus-like particles [144] and MVA [142] ; however, these approaches are amenable to expression in any of the viruses vectors described here. The goal of any vaccine is to protect against infection and disease, while inducing population-based immunity to reduce or eliminate virus transmission within the population. It is clear that currently licensed influenza vaccines have not fully met these goals, nor those specific to inducing long-term, robust immunity. There are a number of vaccine-related issues that must be addressed before population-based influenza vaccination strategies are optimized. The concept of a -one size fits all‖ vaccine needs to be updated, given the recent ability to probe the virus-host interface through RNA interference approaches that facilitate the identification of host genes affecting virus replication, immunity, and disease. There is also a need for revision of the current influenza virus vaccine strategies for at-risk populations, particularly those at either end of the age spectrum. An example of an improved vaccine regime might include the use of a vectored influenza virus vaccine that expresses the HA, NA and M and/or NP proteins for the two currently circulating influenza A subtypes and both influenza B strains so that vaccine take and vaccine antigen levels are not an issue in inducing protective immunity. Recombinant live-attenuated or replication-deficient influenza viruses may offer an advantage for this and other approaches. Vectored vaccines can be constructed to express full-length influenza virus proteins, as well as generate conformationally restricted epitopes, features critical in generating appropriate humoral protection. Inclusion of internal influenza antigens in a vectored vaccine can also induce high levels of protective cellular immunity. To generate sustained immunity, it is an advantage to induce immunity at sites of inductive immunity to natural infection, in this case the respiratory tract. Several vectored vaccines target the respiratory tract. Typically, vectored vaccines generate antigen for weeks after immunization, in contrast to subunit vaccination. This increased presence and level of vaccine antigen contributes to and helps sustain a durable memory immune response, even augmenting the selection of higher affinity antibody secreting cells. The enhanced memory response is in part linked to the intrinsic augmentation of immunity induced by the vector. Thus, for weaker antigens typical of HA, vectored vaccines have the capacity to overcome real limitations in achieving robust and durable protection. Meeting the mandates of seasonal influenza vaccine development is difficult, and to respond to a pandemic strain is even more challenging. Issues with influenza vaccine strain selection based on recently circulating viruses often reflect recommendations by the World Health Organization (WHO)-a process that is cumbersome. The strains of influenza A viruses to be used in vaccine manufacture are not wild-type viruses but rather reassortants that are hybrid viruses containing at least the HA and NA gene segments from the target strains and other gene segments from the master strain, PR8, which has properties of high growth in fertilized hen's eggs. This additional process requires more time and quality control, and specifically for HPAI viruses, it is a process that may fail because of the nature of those viruses. In contrast, viral-vectored vaccines are relatively easy to manipulate and produce, and have well-established safety profiles. There are several viral-based vectors currently employed as antigen delivery systems, including poxviruses, adenoviruses baculovirus, paramyxovirus, rhabdovirus, and others; however, the majority of human clinical trials assessing viral-vectored influenza vaccines use poxvirus and adenovirus vectors. While each of these vector approaches has unique features and is in different stages of development, the combined successes of these approaches supports the virus-vectored vaccine approach as a whole. Issues such as preexisting immunity and cold chain requirements, and lingering safety concerns will have to be overcome; however, each approach is making progress in addressing these issues, and all of the approaches are still viable. Virus-vectored vaccines hold particular promise for vaccination with universal or focused antigens where traditional vaccination methods are not suited to efficacious delivery of these antigens. The most promising approaches currently in development are arguably those targeting conserved HA stalk region epitopes. Given the findings to date, virus-vectored vaccines hold great promise and may overcome the current limitations of influenza vaccines.

What is the enhanced memory immune response linked to?

Effectiveness of zinc supplementation on diarrhea and average daily gain in pre-weaned dairy calves: A double-blind, block-randomized, placebo-controlled clinical trial https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6619766/ SHA: ef20a0cd67ce018cf061f154bd8be9d0e58d0f23 Authors: Feldmann, Hillary R.; Williams, Deniece R.; Champagne, John D.; Lehenbauer, Terry W.; Aly, Sharif S. Date: 2019-07-10 DOI: 10.1371/journal.pone.0219321 License: cc-by Abstract: The objective of this clinical trial was to evaluate the effectiveness of zinc supplementation on diarrhea and average daily weight gain (ADG) in pre-weaned dairy calves. A total of 1,482 healthy Holstein heifer and bull calves from a large California dairy were enrolled at 24 to 48 hours of age until hutch exit at approximately 90 days of age. Calves were block-randomized by time to one of three treatments: 1) placebo, 2) zinc methionine (ZM), or 3) zinc sulfate (ZS) administered in milk once daily for 14 days. Serum total protein at enrollment and body weight at birth, treatment end, and hutch exit were measured. Fecal consistency was assessed daily for 28 days post-enrollment. For a random sample of 127 calves, serum zinc concentrations before and after treatment and a fecal antigen ELISA at diarrhea start and resolution for Escherichia coli K99, rotavirus, coronavirus, and Cryptosporidium parvum were performed. Linear regression showed that ZM-treated bull calves had 22 g increased ADG compared to placebo-treated bulls (P = 0.042). ZM-treated heifers had 9 g decreased ADG compared to placebo-treated heifers (P = 0.037), after adjusting for average birth weight. Sex-stratified models showed that high birth weight heifers treated with ZM gained more than placebo-treated heifers of the same birth weight, which suggests a dose-response effect rather than a true sex-specific effect of ZM on ADG. Cox regression showed that ZM and ZS-treated calves had a 14.7% (P = 0.015) and 13.9% (P = 0.022) reduced hazard of diarrhea, respectively, compared to placebo-treated calves. Calves supplemented for at least the first five days of diarrhea with ZM and ZS had a 21.4% (P = 0.027) and 13.0% (P = 0.040) increased hazard of cure from diarrhea, respectively, compared to placebo-treated calves. Logistic regression showed that the odds of microbiological cure at diarrhea resolution for rotavirus, C. parvum, or any single fecal pathogen was not different between treatment groups. Zinc supplementation delayed diarrhea and expedited diarrhea recovery in pre-weaned calves. Additionally, zinc improved weight gain differentially in bulls compared to heifers, indicating a research need for sex-specific dosing. Text: Introduction Diarrhea is the leading cause of morbidity and mortality and the most common reason for antimicrobial drug treatments in pre-weaned dairy heifers [1, 2] . A USDA survey of preweaned dairy heifers reported that 24% experienced diarrhea and 18% received antimicrobial treatment for it [1] . Diarrhea is also a leading cause of morbidity and the second foremost cause of mortality in children with over 1 billion cases and a half a million deaths annually [3, 4] . Zinc supplementation in children decreases the incidence, duration, and severity of diarrhea, increases recovery rates, decreases the use of antibiotics and antidiarrheal medications, and reduces mortality [5] [6] [7] [8] [9] [10] . In a clinical trial that established a non-toxic zinc dose and investigated its therapeutic use for diarrhea in neonatal dairy calves, zinc-treated calves had numerically quicker clinical recovery, increased weight gain, and higher odds of fecal clearance of Cryptosporidium parvum between diarrhea onset and recovery compared with placebotreated calves [11] . As a result, zinc supplementation may be beneficial for prevention of diarrhea in dairy calves and, thus, minimize antimicrobial use. However, studies investigating zinc's potential effectiveness are lacking. In children, both organic (zinc acetate, zinc gluconate, and zinc methionine) and inorganic (zinc sulfate and zinc oxide) zinc formulations are beneficial in the prevention and treatment of childhood diarrhea [12] [13] [14] [15] . However, differing bioavailability was observed in several animal studies [16] [17] [18] [19] . In addition, the underlying mechanism of action of oral zinc is unknown [6] . Hence, contrasting the effect, if any, of organic compared to inorganic zinc formulations in pre-weaned calves may help identify differences in mode of action. The objective of this clinical trial was to compare average daily weight gain (ADG) and the incidence and duration of diarrhea in pre-weaned dairy calves randomly assigned to receive either organic zinc methionine (ZM), inorganic zinc sulfate (ZS), or a placebo in milk once daily for 14 days. By elucidating the potential role of zinc supplementation in prevention of diarrhea in preweaned dairy calves, calf morbidity, mortality, and antimicrobial usage may be mitigated. A double-blind, block randomized, placebo-controlled clinical trial was conducted between December 14, 2015 and June 15, 2016 on a large dairy in California's San Joaquin Valley. The dairy was selected based on the owner and calf manager's willingness to participate in the study. The dairy herd was composed of 5,500 lactating cows, predominantly Holsteins, and housed approximately 1,600 pre-weaned calves. Approximately 75% of the calves were born on the participating dairy and 25% were born on an affiliated dairy located approximately 10 miles away. Calves enrolled in the trial included healthy Holstein heifer or bull calves 24 to 48 hours of age. Calves were determined to be healthy via visual examination by a veterinarian (HF) or a trained researcher. Calves were excluded if they had obvious morbidities or congenital defects, were non-Holstein, born on the affiliated dairy, younger than 24 hours of age or older than 48 hours of age at the time of enrollment. Calves from the affiliated dairy were excluded due to differences in physical location and management practices of pre-partum cows. All procedures were approved by the University of California Davis Institutional Animal Care and Use Committee (protocol number 18067 Approved: March 6, 2014) . 107 g between treatment groups (Stata, College Station, TX). After allowing for 15% attrition and assuming 50% incidence of diarrhea based on study authors expert opinion, and a difference in ADG of 107g [11] , a sample size of 500 calves per treatment group (n = 1500 total) was deemed required. Newborn calves were removed from the dam within an hour of birth and placed in a strawbedded, group calf pen where their navels were dipped in an iodine-based solution. Each calf received 4 liters of colostrum within 1 hour of birth and a second colostrum feeding (2 liters) 6-10 hours after birth. Colostrum was refrigerated for < 48 hours and heated in a hot water bath prior to feeding using an esophageal tube feeder. Within 18 hours of birth, pre-weaned calves were transported to individual metal hutches initially bedded with almond shells. Straw hay was later added to wet and muddy hutches throughout the pre-weaning period. For the first 14 days of life, pre-weaned calves were bottle-fed 1.9 liters of milk twice daily and 1.9 liters of a commercial oral electrolyte solution (Calva Lyte; Calva Products, Inc., Acampo, CA) once daily between milk feedings. Milk consisted of a combination of pasteurized waste milk, rehydrated commercial milk replacer powder (Strauss Feeds LLC, Watertown, WI), tetracycline and neomycin powder, and additional supplements (S1 Table) . The proportion of pasteurized waste milk to milk replacer varied with each feeding, as the volume of waste milk varied with changes in the number and production of cows contributing to the waste milk tank. After 14 days of age, calves were bottle-fed 2.8 liters of milk twice daily. Calves with clinical diarrhea received 1.9 liters of a commercial oral electrolyte solution (NuLife; Genex Cooperative, Inc., Shawano, WI) once daily between milk feedings. A list of ingredients that make up the two oral electrolyte solutions can be found in S2 Table. All pre-weaned calves had free choice access to water and a calf starter grain mix. Calves were gradually weaned over a 10-day period, starting at approximately 60 days of age after which calves received a grower grain mix until 90 days of age when they were moved to group pens. Each calf received 1 mL of a selenium supplement (MU-SE; Merck Animal Health, Boxmeer, Netherlands) intramuscularly within 24 hours of age and an intranasal vaccine (Inforce 3, Zoetis, Inc., Florham Park, NJ) within 48 hours of age and again near the time of weaning. Approximately 8.3% (n = 126) of all enrolled calves were also vaccinated using an autogenous Moraxella bovis/bovoculi bacterin vaccine (Newport Laboratories, Inc., Worthington, MN) for the prevention of pinkeye at 5 and 7 weeks of age. All pre-weaned calves were evaluated daily by dairy personnel for calfhood diseases and treated according to standard on-farm treatment protocols. With regard to diarrhea therapy, calves less than two weeks of age with clinical diarrhea received an oral mixture of 118.5 mL (2.08 g) bismuth subsalicylate (Bismusal Suspension, Durvet, Inc., Blue Springs, MO) and 31.5 mL (1575 mg) spectinomycin (SpectoGard, Bimeda, Inc., Le Sueur, MN) once daily for two days. Calves older than two weeks of age with clinical diarrhea received oral sulfamethoxazole (1600 mg)/trimethoprim (320 mg) (Amneal Pharmaceuticls of NY, Hauppauge, NY) once daily for 2 to 3 days. Repeated treatment was at the discretion of the calf manager. (inductively coupled plasma mass spectrometry). Financial limitations restricted the ability to test more than two samples of each dietary component. Due to variation of zinc content in duplicate samples, the maximum concentration was used to estimate the daily zinc intake during the first 14 days of life (S3 Table) . In blocks of 36, enrolled calves were randomized using a random number generator (Microsoft Corporation, Redmond, WA) to one of three treatment groups: 1) placebo, 2) ZM, or 3) ZS to be administered in the morning milk feeding once daily for 14 days, starting on the day after enrollment. During the 14-day zinc treatment period, study calves that did not drink the entire milk bottle were tube-fed the remaining milk by trained technicians using an esophageal feeder disinfected between uses. The ZM treatment group received 0.45 g zinc methionine complex (equivalent to 80 mg of elemental zinc) as the product Zinpro180 (Zinpro Corporation, Eden Prairie, MN) combined with 0.44 g milk replacer powder. The ZS treatment group received 0.22 g zinc sulfate monohydrate (equivalent to 80 mg of elemental zinc) (Sigma-Aldrich Company, St. Louis, MO) combined with 0.44 g milk replacer powder. The placebo treatment group received approximately 0.44 g fresh milk replacer powder. Zinc supplementation was based on a previously published clinical trial, toxicological studies, and nutritional guidelines [11, [22] [23] [24] [25] . The milk replacer powder used in treatment preparation was the same product used in the pre-weaned calf milk ration. Treatments were weighed (GX-2000 precision scale; A&D Co Ltd., San Jose, CA) at the Dairy Epidemiology Laboratory at the University of California, Davis Veterinary Medicine Teaching and Research Center (VMTRC; Aly Lab) in Tulare, CA and placed in 2.0 mL microcentrifuge tubes with polypropylene snap caps (Fisher Scientific, Pittsburgh, PA). Prior to study commencement, a color was randomly and permanently assigned to each treatment group, after which treatment tubes, calf milk bottles and calf hutches were marked with either pink, orange, or yellow ink. The study investigators and technicians responsible for treatment preparation, allocation, administration and data collection were blinded to the color assignment until completion of the trial. For each calf, the study period started at enrollment (24 to 48 hours of age) and ended when the calf exited the hutch (approximately 90 days of age) and from here onwards will be referred to as the "pre-weaning period." Calf enrollment and study procedures were performed daily at the time of morning feeding. At enrollment, calf characteristics, including sex, birth date, time of first colostrum feeding, and treatment color were recorded. Attitude and feces were assessed daily until 28 days post-enrollment using previously published methods [11] by two study investigators, a veterinarian and a trained researcher. Attitude scoring was based on a threepoint scale. A calf with an attitude score of 1 was bright, alert, and readily stood with stimulation; a calf with a score of 2 was quiet, alert, and stood only with moderate stimulation; a calf with a score of 3 exhibited a dull mentation and remained recumbent in response to stimulation. Fecal scoring was performed only on fresh feces and was also based on a three-point scale, as 1 (solid), 2 (semi-formed/loose), or 3 (watery). If no fresh feces were observed in the hutch, "none seen" (NS) was recorded. Body weight was measured using a digital scale at birth, end of treatment, and hutch exit by farm employees with the exception of end of treatment weights, which were recorded by study investigators. Treatments for farm-diagnosed illnesses were performed and recorded on hutch cards by the calf manager. Study investigators regularly recorded this information from cards in addition to extracting treatment event reports from DairyComp 305 (Valley Agricultural Software, Tulare, CA). Though daily diagnosis and treatment of study calves was performed and recorded by the calf manager, a veterinarian was responsible for examining and determining whether study calves met specific criteria for euthanasia. A calf was euthanized if morbidity was severe enough to significantly depress appetite, hydration status, attitude, mentation, and/or ambulatory capability and the calf showed limited to no immediate response to therapy or supportive care. Study calves were euthanized by the calf manager using an on-farm captive bolt protocol established by the herd veterinarian within 3 hours of the decision to euthanize. Enrolled calves that died prior to hutch exit were necropsied within 24 hours of death by a veterinarian. All calves were monitored throughout the study period for evidence of zinc toxicity. At the end of the study period, calves were cared for at the dairy in accordance with standard commercial operations. Using the same random number generator, a random sample of 127 calves was selected for additional biologic sampling. Approximately 8 to 10% of the study population was selected due to the financial constraints of additional laboratory testing. Serum zinc concentration at baseline and in response to treatment were evaluated for the three treatment groups. Feces collected on the first day of diarrhea and at diarrhea resolution were evaluated for four fecal pathogens (Escherichia. coli K99, bovine rotavirus and coronavirus, and Cryptosporidium parvum oocysts). Serum total protein. At enrollment, blood from each calf was collected from the jugular vein using a 20 gauge 1-inch multi-sample needle (Exelint International Co., Redondo Beach, CA) and placed into a 10 mL red top serum tube (BD Vacutainer, Franklin Lakes, NJ) for determination of total protein. Samples remained at room temperature for up to 12 hours until clotting and were then centrifuged (International Equipment Company, CRU-5000, Needham Heights, MA) for 15 minutes. Total protein (g/dL) was measured by a single investigator (HF) on decanted serum using a handheld refractometer (Sper Scientific, Model 300005, Scottsdale, AZ). Serum zinc. For the 127 randomly-sampled calves, additional blood was collected at enrollment and on the last day of treatment, as described above, and placed into 6.0 mL trace element tubes (BD Vacutainer, Franklin Lakes, NJ). Serum was extracted, as described above, and placed in a 2.0 mL microcentrifuge tube with a polypropylene snap cap (Fisher Scientific, Pittsburgh, PA) and stored at −20˚C until analysis. Using the same random number generator, 36 of the 127 sampled calves were randomly selected for analysis due to limited financial resources. The pre-and post-zinc supplementation serum samples from each calf (n = 72) were analyzed for zinc concentration (ppm) by ICP-OES (inductively coupled plasma-optical emission spectroscopy) at the CAHFS Laboratory. Quality control samples, including method blanks, laboratory control spikes, and reference Sigma serum, were run with each set of study samples. Fecal analysis. For 127 randomly-sampled calves at the first diarrhea episode, fecal samples were collected at two time points, the first day of diarrhea (fecal score > 1) and the day diarrhea resolved (second day of fecal score = 1). Using new gloves and sterile lubricant, fresh feces was collected by digital rectal stimulation into 20 mL polypropylene twist-top jars (The Cary Company, Addison, IL) and stored at -20˚C until analysis. Fecal samples were tested at the Dairy Epidemiology Laboratory, VMTRC by a veterinarian for E. coli K99, bovine rotavirus and coronavirus, and C. parvum oocysts using a commercial kit (Pathasure Enteritis 4; Biovet, Quebec, Canada) that is highly specific (> 90%) and sensitive (E. coli K99, 93%; rotavirus, 100%; coronavirus, 77%) [26, 27] . For calves with a first-day diarrhea sample on or before 7 days of age, both fecal samples were tested for all four pathogens. For calves with a first-day diarrhea sample after 7 days of age, both fecal samples were tested for three pathogens (C. parvum, bovine rotavirus and coronavirus). Samples from calves older than 7 days of age were not tested for E. coli K99 based on calves' susceptibility [28] . Testing was performed according to kit manufacturer guidelines, and a low-temperature incubator (Fisher Scientific, Model 146, Pittsburgh, PA) was used during incubation periods. Test results were recorded as positive or negative using control wells for color comparison. If the color change was darker than the negative control, the sample was considered positive. Milk zinc. For each of the 107 study days (December 15, 2015 to March 31, 2016) of zinc supplementation, approximately 1.5 mL of treated milk from two bottles of each treatment group were randomly collected into 2.0 mL microcentrifuge tubes with polypropylene snap caps (Fisher Scientific, Pittsburgh, PA), and stored at −20˚C until analysis. At the time of analysis, milk samples were thawed at 4˚C, vortexed, pooled by week and treatment group, and analyzed for zinc concentration (ppm) by ICP-MS (inductively coupled plasma mass spectrometry) at the CAHFS Laboratory. Quality control samples, including method blanks, laboratory control spikes, National Institute of Standards and Technology (NIST) reference materials (NIST 1640), and a spiked milk sample, were run with each set of study samples. Data analysis was performed using R Statistic Software version 3.3.1 and Stata IC 14.2 (College Station, TX). Statistical differences were determined at the 5% level of significance using per protocol analysis. An ANOVA was used to compare calves in each treatment group at enrollment with respect to birth weight (kg), serum total protein (g/dL), attitude score, and fecal score. Oral zinc dose at the start and end of treatment was calculated as the zinc supplementation dose (80 mg) divided by calf body weight (kg) at birth and on the last day of treatment, respectively. An ANOVA was used to compare oral zinc dose (mg/kg) at treatment start and end as well as mean body weight (kg) at birth, end of treatment, and hutch exit between treatment groups and between bulls and heifers. A Chi-Square test of Independence was used to compare the proportions of calves by sex as well as mortality between treatment groups. For all analyses, Tukey's Honest Significant Difference method was used to generate pairwise comparisons to further characterize significant differences identified by ANOVA. Residual diagnostics, including Residuals vs. Fitted, Scale-Location, Normal Q-Q, and Cook's distances plots, were used to validate all ANOVA model assumptions. The non-parametric Kruskal-Wallis Rank Sum test was used when assumptions were violated. For the randomly-sampled calves, Fisher exact tests were used to compare fecal pathogen prevalence on the first day of diarrhea and at diarrhea resolution between treatment groups. Pairwise comparisons with Bonferroni adjustment were used to identify specific differences in pathogen prevalence. An ANOVA was used to compare serum zinc concentration before and after treatment between treatment groups and between bulls and heifers. A Kruskal-Wallis Rank Sum test was used to compare rank sums of zinc concentrations in pooled milk samples from different treatment groups. Post-hoc Nemenyi-tests for pairwise multiple comparisons of ranked data were used to identify specific differences in zinc concentrations between groups. For all regression models in this study, univariate regression was first used to evaluate associations between individual predictor variables and outcomes. All variables with statistical and/or biological significance were initially included in multivariate regression models. The final models were built using a manual backwards elimination procedure, with a significance level of P > 0.05 as the removal criterion. Confounding was assessed using the method of change of estimates, where a 10% or greater change in the estimate of the treatment group regression coefficient between the models with and without the confounder variable was used as evidence of confounding [29] [30] [31] [32] . Variables identified as confounders were included in the final model. All possible interactions between treatment group and predictor variables were explored and retained in the final model if statistically significant. Microbiological cure. For the randomly-sampled calves, logistic regression was used to evaluate associations between microbiological cure and treatment group. Other predictor variables of interest included sex, serum total protein, and age on the first day of diarrhea. Microbiological cure was defined as a negative fecal ELISA test at resolution of clinical diarrhea for calves with a positive ELISA test on the first day of diarrhea for at least one of the four fecal pathogens (E. coli K99, bovine rotavirus and coronavirus, and C. parvum). Models were generated for each fecal pathogen individually and an overall model, which evaluated microbiological cure at clinical diarrhea resolution for calves that tested positive for any single pathogen on the first day of diarrhea. Serum total protein and calf age at first diarrhea were included in all final models to control for potential confounding by passive transfer status and age. Mean daily weight change. Linear regression was used to evaluate associations between ADG (kg) and treatment group during the treatment and pre-weaning periods separately. Other predictor variables of interest included sex, birth weight (kg), serum total protein, number of days having diarrhea, age, and volume (L) of milk, Calva, and NuLife electrolytes fed at either end of treatment or hutch exit. For each calf, ADG during the treatment and pre-weaning period was calculated as the difference between birth and end treatment or hutch exit weight, respectively, divided by the number of days between these time points. To explore the possibility of an interaction between treatment group, sex, and birth weight, the final linear regression model for ADG during the pre-weaning period was stratified by sex. Age at the end of treatment or hutch exit and number of days with diarrhea were dropped from all final models in favor of improved Akaike information criterion (AIC). Onset of diarrhea and clinical cure. For all survival analyses, diarrhea was defined as a fecal score greater than 1 while diarrhea cure was defined as the second consecutive day of normal feces (fecal score of 1) following the first diarrhea episode. Subsequent episodes of diarrhea were not included in the analysis. Calves that died or did not experience diarrhea or cure from diarrhea were censored. If fresh feces were not observed on daily calf hutch assessment, a fecal score was not recorded for that day and not included in the analyses. Kaplan-Meier analysis was used to determine median days to first diarrhea event and, for those calves that developed diarrhea during the assessment period, median days to clinical diarrhea cure. A Log Rank test of equality was used to compare survivor functions between treatments. Cox Proportional Hazards regression analysis was used to estimate and compare the hazard of diarrhea and diarrhea cure between treatment groups. Sex, age, serum total protein at enrollment, birth weight (kg), antimicrobial therapy, and application of fresh straw to the hutches were evaluated as predictor variables and potential confounders. When modeling the hazard of diarrhea cure, a binary variable termed therapeutic supplementation indicating whether calves were treated with either ZM, ZS or placebo for all or at least the first 5 days of diarrhea was evaluated as an additional covariate. A five-day period was selected by the authors based on clinical experience, as five days represents a reasonable duration over which most therapeutic treatments for calf diarrhea should be applied and be expected to alleviate disease. The proportional hazards assumption that the hazard of diarrhea is independent of time was assessed using analysis of Schoenfeld residuals and testing whether the log hazard-ratio function is constant over time. Any variable found to violate the proportional hazards assumption was included in the final regression model as a time varying covariate. A total of 1,513 calves were enrolled in the trial. However, due to failure to immediately recognize exclusion criteria, 23 calves were excluded shortly after enrollment. In addition, 8 calves were excluded due to treatment errors. Therefore, a total of 1,482 calves (placebo = 500, ZM = 491, ZS = 491) were included in the final analyses. A total of 242 calves (16.3%) had minimal fecal output at the time of enrollment while 125 calves (8.4%) had abnormal fecal scores of 2 or 3 that were described as meconium. All enrolled calves appeared healthy on visual assessment and hence were assumed to have a normal fecal score at the time of enrollment. The three treatment groups at enrollment did not differ significantly in mean birth weight (kg) (P = 0.244), mean serum total protein (g/dL) (P = 0.541), mean attitude score (P = 0.845), mean fecal score (P = 0.522), as shown in Table 1 , or distribution of calf sex (P = 0.472). Of the 1,482 study calves, 21 (1.4%) died during the trial: 5 calves in the placebo group (1.0%), 11 calves in the ZM group (2.2%), and 5 calves in the ZS group (1.0%). Of these 21 calves that died, 14 (66.7%) were bulls and 7 (33.3%) were heifers. Eighteen of the 21 calves (85.7%) were found dead, rather than euthanized, due to acute and spontaneous death without previous obvious clinical signs of disease. The remaining 3 calves were euthanized prior to death due to severe and/or prolonged morbidity. Characteristics and causes of death based on field necropsy of these calves can be found in S4 Table. There was no significant difference in the proportion of calves that died between treatment groups (P = 0.168), though mortality was significantly higher in bulls compared to heifer calves (P = 0.049). Birth weight data were available for all 1,482 calves. Due to calf mortality between enrollment and completion of treatment (n = 4), end treatment weight data were available for 1,478 calves. Similarly, due to calf mortality (n = 21) or missing data for body weight at hutch exit from the dairy's records (n = 40), hutch exit weight data were available for 1,421 calves. A summary of body weight data stratified by treatment group and sex is presented in Table 2 . Within each treatment group, bull calves showed consistently higher birth weight (P < 0.001), end treatment weight (P < 0.001), and exit hutch weight (P < 0.001) compared to heifer calves. However, at birth, end of treatment, or hutch exit there were no differences in body weight between treatment groups for bulls (P > 0.1) or heifers (P > 0.1). The mean attitude scores during the study period were 1.2 across the three treatment groups (P = 0.208). Of 1,482 calves included in the final analysis, A total of 629 treated milk samples were obtained throughout the 107-day study period and pooled by treatment group and week, yielding a total of 16 pooled samples per treatment group. Zinc concentrations (ppm) were significantly higher in pooled milk samples treated with ZM (P < 0.001) and ZS (P < 0.001) compared to placebo-treated samples, and there were no significant differences between ZM and ZS-treated samples (P = 1.000), as shown in S5 Table. Within zinc treatment groups, oral zinc dose at the start and end of treatment is summarized by sex in S6 Table. For both ZM-and ZS-treated calves, oral zinc dose at the start (P < 0.001) and end (P < 0.001) of treatment was significantly higher in heifer versus bull calves. Serum zinc concentrations before and after treatment were obtained from 36 calves (n = 12 for each treatment group) and are summarized in S7 Table. Overall, there were no significant differences in mean pre-treatment serum zinc concentrations between treatment groups (P = 0.233). Mean post-treatment serum zinc concentrations were significantly higher in calves treated with ZM (P < 0.001) and ZS (P = 0.002) compared to placebo-treated calves, and there were no significant differences among calves treated with ZM and ZS (P = 0.406). Stratification of serum zinc data by treatment group and sex demonstrated that for ZM-treated calves, heifers had a numerically higher post-treatment serum zinc concentration compared to bulls, though the difference was not statistically significant (P = 0.199). In contrast, in ZStreated calves, heifers had a numerically lower post-treatment serum zinc concentration compared to bulls, though the difference was also not statistically significant (P = 0.538). Fecal analysis data were analyzed for 92 of the 127 randomly-selected calves. The remaining 35 calves were not included in the analysis due to not acquiring diarrhea during the assessment period (n = 10), death prior to final sampling (n = 1), exclusion due to improper treatment regimen (n = 1), or incorrect sampling day (n = 23).The 92 calves had a mean age at onset of diarrhea of 13.3, 11.0 and 11.3 days for ZM, ZS and placebo-treated groups, respectively; and a mean age at resolution of diarrhea of 18.8, 16.1 and 15.7 days for ZM, ZS and placebo-treated groups, respectively. There were no significant differences in the prevalence of E. coli K99 (P = 0.694), rotavirus (P = 0.331), coronavirus (P = 0.819), or C. parvum (P = 0.719) fecal shedding on the first day of diarrhea between treatment groups (S8 Table) . There were no significant differences in the prevalence of E. coli K99 (P = 0.256), rotavirus (P = 0.344), or coronavirus (P = 1.000) fecal shedding at resolution of diarrhea between treatment groups, though there was a difference in C. parvum (P = 0.006) fecal shedding between treatment groups (S9 Table) . The prevalence of C. parvum fecal shedding at resolution of diarrhea was significantly higher in calves treated with ZM (P = 0.009) and ZS (P = 0.023) compared to placebo-treated calves, and there were no significant differences among calves treated with ZM and ZS (P = 1.000). Results of logistic regression models for overall and pathogen-specific microbiological cure are presented in Tables 3-5 . For pathogen-specific cure, all calves that tested positive on the first day of diarrhea for coronavirus (n = 4) tested negative at clinical diarrhea resolution. For calves that tested positive on the first day of diarrhea for E. coli K99 (n = 9), all placebo-treated calves (n = 2) tested negative at clinical diarrhea resolution while half (n = 2) of all ZS-treated calves tested either positive or negative at clinical diarrhea resolution, resulting in omission of ZS treatment variable from the model due to collinearity. Hence, logistic regression analyses for microbiological cure in calves that tested positive for coronavirus and E. coli K99 were not possible. For 59 calves that tested positive for rotavirus on the first day of diarrhea (Table 3) , calves treated with ZM had a 50% increased odds of testing negative at diarrhea resolution compared to placebo-treated calves, though this difference was not significant (P = 0.549). Likewise, calves treated with ZS had 100% increased odds (2 times the odds) of testing negative for rotavirus at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.314). However, this model demonstrated a significant main effect of serum total protein, such that for every 1 unit (g/dL) increase in serum total protein at enrollment, the odds of microbiological cure of rotavirus decreased by 79% (P = 0.026). For 40 calves that tested positive for Cryptosporidium parvum on the first day of diarrhea (Table 4) , calves treated with ZM had an 87% reduced odds of testing negative at diarrhea resolution Effect of zinc on diarrhea in calves compared to placebo-treated calves, though this difference was not significant (P = 0.119). Likewise, calves treated with ZS had a 74% reduced odds of testing negative for Cryptosporidium parvum at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.183). For 55 calves that tested positive for any one of the four fecal pathogens (E. coli K99, rotavirus, coronavirus, Cryptosporidium parvum) on the first day of diarrhea (Table 5) , calves treated with ZM had a 25% increased odds of testing negative at diarrhea resolution compared to placebo-treated calves, though this difference was not significant (P = 0.769). Likewise, calves treated with ZS had a 52% increased odds of testing negative for Cryptosporidium parvum at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.633). However, this model demonstrated that heifer calves had 71% lower odds of microbiological cure of any single fecal pathogen compared to bull calves (P = 0.076). A total of 1,482 calves were included in the linear regression model results for ADG during the treatment period, which are presented in Table 6 . There was no significant difference in ADG for ZM-or ZS-treated calves compared to placebo-treated calves, though there were significant main effects of sex, birth weight, and milk volume. Specifically, heifer calves gained 70 g bodyweight per day less compared to bull calves (P < 0.001). For every 1 kg increase in birth weight, calves gained 16 g per day less than their herd mates (P < 0.001). For every 1 L increase in milk volume per day during the treatment period, calves gained an additional 13 g per day (P < 0.001). Table 7 summarizes the linear regression analysis of ADG during the pre-weaning period for 1,421 calves which showed a significant difference in ADG for ZM-treated calves compared to placebo-treated calves and in bull versus heifer calves. Milk volume had a significant effect on ADG, such that for every 1 L increase in milk volume per day during the treatment period, calves gained an additional 2 g bodyweight per day (P = 0.001). Results of the final model showed a significant main effect for ZM-treated bull calves and a significant interaction term for ZM treatment by sex. After controlling for milk volume received during the treatment Table 6 calves (n = 1,482) Effect of zinc on diarrhea in calves period, ZM-treated bulls gained 22 g body weight per day on average more than placebotreated bull calves (P = 0.042) and ZM-treated heifers gained 12 g less body weight per day on average compared to placebo-treated heifers (P = 0.019). When considering the model coefficients for treatment group, sex, and their interaction, bull calves treated with ZM gained 454 g per day (0.432 + 0.022) while female calves treated with ZM gained 0.404 g per day (0.432 + 0.022-0.016-0.034), hence 50 g per day more gain in male calves compared to heifers treated with ZM (P = 0.019). For ZS-treated calves, there was a numerical decrease in weight gain of 5 g per day in bulls and 11 g per day in heifers compared to placebo-treated calves, though the differences were not significant (P = 0.673 bulls; P = 0.681 heifers). Linear regression models of ADG during the pre-weaning period were stratified by sex in order to avoid interpreting a three-way interaction between treatment group, sex, and birth weight. In the heifer model (S10 Table) , the interaction between ZM treatment and birth weight was significant which implied that birth weight modified the effect of ZM treatment on ADG. At a 29 kg birth weight (two standard deviations below the mean), ZM-treated heifers gained 49 g body weight per day on average less than placebo-treated heifers (P = 0.037). However, at a 49 kg birth weight (two standard deviations above the mean), ZM-treated heifers gained 30 g body weight per day on average more than placebo-treated heifers (P = 0.037). In the bull calves model (S11 Table) there was no significant interaction between treatment group and birth weight. A total of 1,482 calves were included in the Kaplan-Meier survival analysis of time to first diarrhea event (Fig 1) . There were no significant differences in median age at onset of diarrhea, specifically, 8, 8 and 7 days for the ZM, ZS and placebo-treated calves, respectively (P = 0.402). Cox proportional hazard regression model for diarrhea hazard are presented in Table 8 . After controlling for age, calves treated with ZM had a 14.7% reduced hazard of diarrhea compared to placebo-treated calves (P = 0.015). Calves treated with ZS had 13.9% reduced hazard of diarrhea compared to placebo-treated calves (P = 0.022). calves (n = 1,421) A total of 1,394 calves were included in the Kaplan-Meier survival analysis of time to clinical diarrhea cure (Fig 2) , as 88 calves failed to acquire diarrhea during the assessment period. There were no significant differences in the median days to diarrhea cure which was 7 days across all 3 treatment groups (P = 0.264). Cox proportional hazard regression model for diarrhea cure hazard are presented in Table 9 . Results of the final model showed a significant interaction term between treatment and therapeutic supplementation as well as the need for age as a time varying covariate. When considering calves that did not receive supplementation, respective to each of the 3 groups, for at least the first five days of diarrhea there was no significant difference between either ZM-and ZS-treated calves compared to placebo-treated calves (P = 0.223 ZM, P = 0.134 ZS). However, when considering calves that were supplemented for at least the first five days of diarrhea, ZM-treated calves experienced a 21.4% higher hazard of cure from diarrhea compared to placebo-treated calves (P = 0.027). Likewise, ZS-treated calves experienced a 13.0% higher hazard of cure from diarrhea compared to placebo-treated calves (P = 0.040). The current trial demonstrated evidence for the beneficial effect of ZM on ADG and neonatal diarrhea as well as an effect of ZS on diarrhea in dairy calves during the pre-weaning period. It is important to consider these results in the context of the entire pre-weaning and hutch period. On average, after 90 days from birth to hutch exit, placebo-treated bull calves gained 38.88 kg body weight while ZM-treated bull calves gained an additional 1.98 kg (40.86 kg). In contrast, the effect of zinc on weight gain in treated heifers depended on birth weight. Low birth weight heifers treated with ZM gained on average less than a placebo-treated heifer of the same birth weight. In contrast, high birth weight heifers treated with ZM gained more than placebo-treated heifers of the same birth weight. The switch in direction of the association between ZM treatment and ADG in heifer calves depending on birth weight suggests a doseresponse effect rather than a true sex-specific effect of ZM on ADG. Hence, low birth weight calves (including heifers) may require a lower dose of ZM to mitigate any negative effect of what is otherwise a suitable dose for higher birth weight calves. These findings are in agreement with a previous randomized clinical trial testing the effect of daily oral zinc in diarrheic neonatal Holstein calves which, showed that ZM-treated calves had a numerically, though not significantly increased ADG compared to calves treated with zinc oxide or placebo due to small sample size [11] . In general, our trial findings are in agreement with the large body of human literature supporting the use of oral zinc for the prevention and treatment of diarrhea and impaired growth in children [5, 10, 33] . Zinc supplementation is widely accepted by global health organizations as a vital component of therapy for childhood diarrhea [3, 4] , however, recent reviews of the literature demonstrated heterogeneity in study results on the basis of age, baseline zinc status, geographic location, and supplementation regimen [10, 34] . Similar to our findings, a sex-specific response to zinc supplementation has been demonstrated in several human studies. Zinc gluconate administered for diarrhea prevention reduced the incidence of dysentery in treated boys but not girls [35] ; when given therapeutically, it reduced diarrhea duration and frequency more dramatically in boys compared to girls [36] . Similarly, zinc sulfate was shown to improve Effect of zinc on diarrhea in calves diarrhea outcomes in boys but improved growth rates in girls [13] . Broadly, these differences between male and female responses to zinc supplementation are not understood, though theories regarding differences in immune function and response [13, 35] , diarrhea etiology [13] , and nutrient requirements [35] have been proposed. In the current study, ZM-treated bulls demonstrated increased ADG compared to placebo-treated bulls while ZM-treated heifers demonstrated decreased ADG compared to placebo-treated heifers. However, due to a significant interaction between ZM treatment and birth weight, this reduction in ADG in ZMtreated heifers was overcome with increasing birth weight, such that ZM-treated heifers with birth weights above 42 kg experienced increased ADG during the pre-weaning period, compared to placebo-treated heifers with birth weights above 42 kg. Differences in the growth response to ZM supplementation between bull and heifer calves may have been related to its effect on feed intake. Previous research on the effects of feeding various doses of oral zinc oxide to pre-ruminant dairy calves demonstrated that high levels of oral zinc supplementation resulted in reduced feed intake [23] . In the current trial, oral ZM dose was estimated to be significantly higher in heifers compared to bulls due to the significantly lower birth weight of heifers. Additionally, serum zinc concentrations in ZM-treated heifers were numerically higher than that of bulls, though this difference was not significant, likely due to the small sample size. Perhaps the higher zinc dose in heifers was associated with reduced feed intake, leading to reduced growth, and that this effect was more pronounced for ZM compared to ZS. The fact that ZM-treated heifers with birth weights approaching those of average bull calves (and, therefore, a similar zinc dose to that in bulls) experienced an increase in ADG over placebo-treated heifers similar to that of bull calves partially supports this theory. Although management practices on the study dairy were designed to be identical for both bulls and heifers, it is possible that subtle, unrecognized differences in nutritional and health management may also have contributed to sex-specific differences in weight gain. Nevertheless, future trials are warranted to investigate the potential differences in the dose-response to zinc supplementation between bulls and heifers. We hypothesized that ADG would be increased in zinc-supplemented calves compared to placebo-supplemented calves due to the potential preventive and therapeutic effects of zinc supplementation on neonatal diarrhea. In other words, calf diarrhea is mitigated by zinc supplementation and, therefore, on the causal pathway between zinc and ADG. However, considering the similarly-reduced hazard of diarrhea and increased hazard of cure from diarrhea in both ZM and ZS treatment groups but a lack of effect of ZS on ADG, it is likely that the effect of ZM on ADG is not solely mediated through its effects on diarrhea. Differences in effectiveness between organic and inorganic formulations also may exist. In fact, the underlying mechanism of action of oral zinc remains unknown [6] . Several theories of the mechanisms of action of zinc in the prevention and treatment of childhood diarrhea exist, including a mucosal-protective role, a diarrhea-induced zinc deficiency, an essential element in cell-mediated immunity, and a modifier of intra-luminal electrolyte secretion and absorption [6, [37] [38] [39] . The clinical and practical implications of effects of ZM supplementation on ADG and diarrhea must be considered. Pre-weaned calf diarrhea remains an ongoing issue for the dairy industry. The deleterious effects on calf health and performance and the resulting economic burden create a strong incentive to treat and prevent diarrhea in pre-weaned calves. On large dairy operations like those in California's Central Valley, small changes in disease incidence and duration as well as animal growth and performance can have profound economic consequences. As a non-antimicrobial product, zinc may become increasingly attractive as antimicrobials in livestock feed are under increased scrutiny and regulation due to concerns about antimicrobial resistance [2, 40] . Prevalence of C. parvum fecal shedding in a random sample of 92 study calves at onset and resolution of diarrhea was significantly higher in calves treated with zinc compared to Placebo-treated calves. In contrast, a previous study where calves that tested positive for C. parvum at the start of diarrhea and were treated with ZM had 16 times higher odds of being fecal ELISA negative at exit compared to the Placebo group (P = 0.08; power = 72.3%) [11] . The difference in findings may be due to the differences in the timing of diarrhea across treatment groups. For the current study's random sample of calves that acquired, survived, and were sampled on the correct days, the mean age of calves on both onset and resolution of diarrhea was higher for ZM and ZS calves compared to placebo-treated calves. Although C. parvum oocyst shedding in infected calves can occur as early as 3 days of age, peak shedding occurs at about 14 days of age [41] . It is possible that the increase in prevalence of C. parvum shedding in ZM and ZS treated groups was due to the increased age of zinc-treated calves compared to placebo-treated calves at resolution of diarrhea. The latter explanation is also supported by our findings that the odds of microbiological cure from C. parvum significantly decreased in older calves, with no significant differences in the odds of cure between treatment groups. In addition, the current testing did not estimate the concentration of C. parvum shedding which may still differ between treatment groups. Despite the large sample size, the current trial was limited to a single California dairy, which may represent other large dairies but does not reflect all the dairy management systems in California or elsewhere. Additionally, our results show that calves respond to zinc supplementation for diarrhea prevention differently depending on chemical formulation and calf sex. The latter could be due to differences in body weight between bulls and heifers and may point towards the need for sex-specific dosing. Furthermore, the current research did not evaluate the potential economic utility of zinc supplementation. Future studies on more accurate dosing of zinc by calf sex, the practical feasibility of weight-based dosing, and the expected cost-effectiveness of zinc administration as part of the management of pre-weaned dairy calves are warranted. Finally, our clinical trial was performed on a single, large, predominately Holstein, California dairy over a six-month period, which precluded our ability to evaluate differences due to season or breed. Hence, future studies to assess any modifying effect of breed and seasonal differences on the effect of zinc on calf health and weight gain are also needed. The current double blind, block-randomized placebo controlled clinical trial tested the effect of a prophylactic daily oral zinc supplementation in neonatal Holstein calves. Bull calves treated with ZM had a significantly increased ADG (22 g per day) during the pre-weaning period compared to placebo-treated bulls. In comparison, ZM-treated heifers had significantly lower average daily gain (9 g per day) compared to placebo-treated heifers, although higher ZM doses in low birthweight heifers may explain the lower ADG. Calves treated with either ZM or ZS had significantly lower risks of diarrhea and significantly higher risk of cure from diarrhea over the first 30 days of life compared to placebo-treated calves and hence the current trial demonstrated that zinc supplementation delayed diarrhea and expedited diarrhea recovery in pre-weaned calves. Additionally, zinc improved weight gain differentially in bulls compared to heifers, indicating the need for further research to investigate zinc dosing in calves. Supporting information S1 (DOCX) S1 Dataset. Raw data collected from trial, organized as separate excel sheets for enrollment, daily assessment, serum total protein, birth weight, exit treatment weight, exit trial weight, serum zinc testing, fecal samples, fecal testing, milk testing, and dead calves. (XLSX)

What was the purpose of the research?

Effectiveness of zinc supplementation on diarrhea and average daily gain in pre-weaned dairy calves: A double-blind, block-randomized, placebo-controlled clinical trial https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6619766/ SHA: ef20a0cd67ce018cf061f154bd8be9d0e58d0f23 Authors: Feldmann, Hillary R.; Williams, Deniece R.; Champagne, John D.; Lehenbauer, Terry W.; Aly, Sharif S. Date: 2019-07-10 DOI: 10.1371/journal.pone.0219321 License: cc-by Abstract: The objective of this clinical trial was to evaluate the effectiveness of zinc supplementation on diarrhea and average daily weight gain (ADG) in pre-weaned dairy calves. A total of 1,482 healthy Holstein heifer and bull calves from a large California dairy were enrolled at 24 to 48 hours of age until hutch exit at approximately 90 days of age. Calves were block-randomized by time to one of three treatments: 1) placebo, 2) zinc methionine (ZM), or 3) zinc sulfate (ZS) administered in milk once daily for 14 days. Serum total protein at enrollment and body weight at birth, treatment end, and hutch exit were measured. Fecal consistency was assessed daily for 28 days post-enrollment. For a random sample of 127 calves, serum zinc concentrations before and after treatment and a fecal antigen ELISA at diarrhea start and resolution for Escherichia coli K99, rotavirus, coronavirus, and Cryptosporidium parvum were performed. Linear regression showed that ZM-treated bull calves had 22 g increased ADG compared to placebo-treated bulls (P = 0.042). ZM-treated heifers had 9 g decreased ADG compared to placebo-treated heifers (P = 0.037), after adjusting for average birth weight. Sex-stratified models showed that high birth weight heifers treated with ZM gained more than placebo-treated heifers of the same birth weight, which suggests a dose-response effect rather than a true sex-specific effect of ZM on ADG. Cox regression showed that ZM and ZS-treated calves had a 14.7% (P = 0.015) and 13.9% (P = 0.022) reduced hazard of diarrhea, respectively, compared to placebo-treated calves. Calves supplemented for at least the first five days of diarrhea with ZM and ZS had a 21.4% (P = 0.027) and 13.0% (P = 0.040) increased hazard of cure from diarrhea, respectively, compared to placebo-treated calves. Logistic regression showed that the odds of microbiological cure at diarrhea resolution for rotavirus, C. parvum, or any single fecal pathogen was not different between treatment groups. Zinc supplementation delayed diarrhea and expedited diarrhea recovery in pre-weaned calves. Additionally, zinc improved weight gain differentially in bulls compared to heifers, indicating a research need for sex-specific dosing. Text: Introduction Diarrhea is the leading cause of morbidity and mortality and the most common reason for antimicrobial drug treatments in pre-weaned dairy heifers [1, 2] . A USDA survey of preweaned dairy heifers reported that 24% experienced diarrhea and 18% received antimicrobial treatment for it [1] . Diarrhea is also a leading cause of morbidity and the second foremost cause of mortality in children with over 1 billion cases and a half a million deaths annually [3, 4] . Zinc supplementation in children decreases the incidence, duration, and severity of diarrhea, increases recovery rates, decreases the use of antibiotics and antidiarrheal medications, and reduces mortality [5] [6] [7] [8] [9] [10] . In a clinical trial that established a non-toxic zinc dose and investigated its therapeutic use for diarrhea in neonatal dairy calves, zinc-treated calves had numerically quicker clinical recovery, increased weight gain, and higher odds of fecal clearance of Cryptosporidium parvum between diarrhea onset and recovery compared with placebotreated calves [11] . As a result, zinc supplementation may be beneficial for prevention of diarrhea in dairy calves and, thus, minimize antimicrobial use. However, studies investigating zinc's potential effectiveness are lacking. In children, both organic (zinc acetate, zinc gluconate, and zinc methionine) and inorganic (zinc sulfate and zinc oxide) zinc formulations are beneficial in the prevention and treatment of childhood diarrhea [12] [13] [14] [15] . However, differing bioavailability was observed in several animal studies [16] [17] [18] [19] . In addition, the underlying mechanism of action of oral zinc is unknown [6] . Hence, contrasting the effect, if any, of organic compared to inorganic zinc formulations in pre-weaned calves may help identify differences in mode of action. The objective of this clinical trial was to compare average daily weight gain (ADG) and the incidence and duration of diarrhea in pre-weaned dairy calves randomly assigned to receive either organic zinc methionine (ZM), inorganic zinc sulfate (ZS), or a placebo in milk once daily for 14 days. By elucidating the potential role of zinc supplementation in prevention of diarrhea in preweaned dairy calves, calf morbidity, mortality, and antimicrobial usage may be mitigated. A double-blind, block randomized, placebo-controlled clinical trial was conducted between December 14, 2015 and June 15, 2016 on a large dairy in California's San Joaquin Valley. The dairy was selected based on the owner and calf manager's willingness to participate in the study. The dairy herd was composed of 5,500 lactating cows, predominantly Holsteins, and housed approximately 1,600 pre-weaned calves. Approximately 75% of the calves were born on the participating dairy and 25% were born on an affiliated dairy located approximately 10 miles away. Calves enrolled in the trial included healthy Holstein heifer or bull calves 24 to 48 hours of age. Calves were determined to be healthy via visual examination by a veterinarian (HF) or a trained researcher. Calves were excluded if they had obvious morbidities or congenital defects, were non-Holstein, born on the affiliated dairy, younger than 24 hours of age or older than 48 hours of age at the time of enrollment. Calves from the affiliated dairy were excluded due to differences in physical location and management practices of pre-partum cows. All procedures were approved by the University of California Davis Institutional Animal Care and Use Committee (protocol number 18067 Approved: March 6, 2014) . 107 g between treatment groups (Stata, College Station, TX). After allowing for 15% attrition and assuming 50% incidence of diarrhea based on study authors expert opinion, and a difference in ADG of 107g [11] , a sample size of 500 calves per treatment group (n = 1500 total) was deemed required. Newborn calves were removed from the dam within an hour of birth and placed in a strawbedded, group calf pen where their navels were dipped in an iodine-based solution. Each calf received 4 liters of colostrum within 1 hour of birth and a second colostrum feeding (2 liters) 6-10 hours after birth. Colostrum was refrigerated for < 48 hours and heated in a hot water bath prior to feeding using an esophageal tube feeder. Within 18 hours of birth, pre-weaned calves were transported to individual metal hutches initially bedded with almond shells. Straw hay was later added to wet and muddy hutches throughout the pre-weaning period. For the first 14 days of life, pre-weaned calves were bottle-fed 1.9 liters of milk twice daily and 1.9 liters of a commercial oral electrolyte solution (Calva Lyte; Calva Products, Inc., Acampo, CA) once daily between milk feedings. Milk consisted of a combination of pasteurized waste milk, rehydrated commercial milk replacer powder (Strauss Feeds LLC, Watertown, WI), tetracycline and neomycin powder, and additional supplements (S1 Table) . The proportion of pasteurized waste milk to milk replacer varied with each feeding, as the volume of waste milk varied with changes in the number and production of cows contributing to the waste milk tank. After 14 days of age, calves were bottle-fed 2.8 liters of milk twice daily. Calves with clinical diarrhea received 1.9 liters of a commercial oral electrolyte solution (NuLife; Genex Cooperative, Inc., Shawano, WI) once daily between milk feedings. A list of ingredients that make up the two oral electrolyte solutions can be found in S2 Table. All pre-weaned calves had free choice access to water and a calf starter grain mix. Calves were gradually weaned over a 10-day period, starting at approximately 60 days of age after which calves received a grower grain mix until 90 days of age when they were moved to group pens. Each calf received 1 mL of a selenium supplement (MU-SE; Merck Animal Health, Boxmeer, Netherlands) intramuscularly within 24 hours of age and an intranasal vaccine (Inforce 3, Zoetis, Inc., Florham Park, NJ) within 48 hours of age and again near the time of weaning. Approximately 8.3% (n = 126) of all enrolled calves were also vaccinated using an autogenous Moraxella bovis/bovoculi bacterin vaccine (Newport Laboratories, Inc., Worthington, MN) for the prevention of pinkeye at 5 and 7 weeks of age. All pre-weaned calves were evaluated daily by dairy personnel for calfhood diseases and treated according to standard on-farm treatment protocols. With regard to diarrhea therapy, calves less than two weeks of age with clinical diarrhea received an oral mixture of 118.5 mL (2.08 g) bismuth subsalicylate (Bismusal Suspension, Durvet, Inc., Blue Springs, MO) and 31.5 mL (1575 mg) spectinomycin (SpectoGard, Bimeda, Inc., Le Sueur, MN) once daily for two days. Calves older than two weeks of age with clinical diarrhea received oral sulfamethoxazole (1600 mg)/trimethoprim (320 mg) (Amneal Pharmaceuticls of NY, Hauppauge, NY) once daily for 2 to 3 days. Repeated treatment was at the discretion of the calf manager. (inductively coupled plasma mass spectrometry). Financial limitations restricted the ability to test more than two samples of each dietary component. Due to variation of zinc content in duplicate samples, the maximum concentration was used to estimate the daily zinc intake during the first 14 days of life (S3 Table) . In blocks of 36, enrolled calves were randomized using a random number generator (Microsoft Corporation, Redmond, WA) to one of three treatment groups: 1) placebo, 2) ZM, or 3) ZS to be administered in the morning milk feeding once daily for 14 days, starting on the day after enrollment. During the 14-day zinc treatment period, study calves that did not drink the entire milk bottle were tube-fed the remaining milk by trained technicians using an esophageal feeder disinfected between uses. The ZM treatment group received 0.45 g zinc methionine complex (equivalent to 80 mg of elemental zinc) as the product Zinpro180 (Zinpro Corporation, Eden Prairie, MN) combined with 0.44 g milk replacer powder. The ZS treatment group received 0.22 g zinc sulfate monohydrate (equivalent to 80 mg of elemental zinc) (Sigma-Aldrich Company, St. Louis, MO) combined with 0.44 g milk replacer powder. The placebo treatment group received approximately 0.44 g fresh milk replacer powder. Zinc supplementation was based on a previously published clinical trial, toxicological studies, and nutritional guidelines [11, [22] [23] [24] [25] . The milk replacer powder used in treatment preparation was the same product used in the pre-weaned calf milk ration. Treatments were weighed (GX-2000 precision scale; A&D Co Ltd., San Jose, CA) at the Dairy Epidemiology Laboratory at the University of California, Davis Veterinary Medicine Teaching and Research Center (VMTRC; Aly Lab) in Tulare, CA and placed in 2.0 mL microcentrifuge tubes with polypropylene snap caps (Fisher Scientific, Pittsburgh, PA). Prior to study commencement, a color was randomly and permanently assigned to each treatment group, after which treatment tubes, calf milk bottles and calf hutches were marked with either pink, orange, or yellow ink. The study investigators and technicians responsible for treatment preparation, allocation, administration and data collection were blinded to the color assignment until completion of the trial. For each calf, the study period started at enrollment (24 to 48 hours of age) and ended when the calf exited the hutch (approximately 90 days of age) and from here onwards will be referred to as the "pre-weaning period." Calf enrollment and study procedures were performed daily at the time of morning feeding. At enrollment, calf characteristics, including sex, birth date, time of first colostrum feeding, and treatment color were recorded. Attitude and feces were assessed daily until 28 days post-enrollment using previously published methods [11] by two study investigators, a veterinarian and a trained researcher. Attitude scoring was based on a threepoint scale. A calf with an attitude score of 1 was bright, alert, and readily stood with stimulation; a calf with a score of 2 was quiet, alert, and stood only with moderate stimulation; a calf with a score of 3 exhibited a dull mentation and remained recumbent in response to stimulation. Fecal scoring was performed only on fresh feces and was also based on a three-point scale, as 1 (solid), 2 (semi-formed/loose), or 3 (watery). If no fresh feces were observed in the hutch, "none seen" (NS) was recorded. Body weight was measured using a digital scale at birth, end of treatment, and hutch exit by farm employees with the exception of end of treatment weights, which were recorded by study investigators. Treatments for farm-diagnosed illnesses were performed and recorded on hutch cards by the calf manager. Study investigators regularly recorded this information from cards in addition to extracting treatment event reports from DairyComp 305 (Valley Agricultural Software, Tulare, CA). Though daily diagnosis and treatment of study calves was performed and recorded by the calf manager, a veterinarian was responsible for examining and determining whether study calves met specific criteria for euthanasia. A calf was euthanized if morbidity was severe enough to significantly depress appetite, hydration status, attitude, mentation, and/or ambulatory capability and the calf showed limited to no immediate response to therapy or supportive care. Study calves were euthanized by the calf manager using an on-farm captive bolt protocol established by the herd veterinarian within 3 hours of the decision to euthanize. Enrolled calves that died prior to hutch exit were necropsied within 24 hours of death by a veterinarian. All calves were monitored throughout the study period for evidence of zinc toxicity. At the end of the study period, calves were cared for at the dairy in accordance with standard commercial operations. Using the same random number generator, a random sample of 127 calves was selected for additional biologic sampling. Approximately 8 to 10% of the study population was selected due to the financial constraints of additional laboratory testing. Serum zinc concentration at baseline and in response to treatment were evaluated for the three treatment groups. Feces collected on the first day of diarrhea and at diarrhea resolution were evaluated for four fecal pathogens (Escherichia. coli K99, bovine rotavirus and coronavirus, and Cryptosporidium parvum oocysts). Serum total protein. At enrollment, blood from each calf was collected from the jugular vein using a 20 gauge 1-inch multi-sample needle (Exelint International Co., Redondo Beach, CA) and placed into a 10 mL red top serum tube (BD Vacutainer, Franklin Lakes, NJ) for determination of total protein. Samples remained at room temperature for up to 12 hours until clotting and were then centrifuged (International Equipment Company, CRU-5000, Needham Heights, MA) for 15 minutes. Total protein (g/dL) was measured by a single investigator (HF) on decanted serum using a handheld refractometer (Sper Scientific, Model 300005, Scottsdale, AZ). Serum zinc. For the 127 randomly-sampled calves, additional blood was collected at enrollment and on the last day of treatment, as described above, and placed into 6.0 mL trace element tubes (BD Vacutainer, Franklin Lakes, NJ). Serum was extracted, as described above, and placed in a 2.0 mL microcentrifuge tube with a polypropylene snap cap (Fisher Scientific, Pittsburgh, PA) and stored at −20˚C until analysis. Using the same random number generator, 36 of the 127 sampled calves were randomly selected for analysis due to limited financial resources. The pre-and post-zinc supplementation serum samples from each calf (n = 72) were analyzed for zinc concentration (ppm) by ICP-OES (inductively coupled plasma-optical emission spectroscopy) at the CAHFS Laboratory. Quality control samples, including method blanks, laboratory control spikes, and reference Sigma serum, were run with each set of study samples. Fecal analysis. For 127 randomly-sampled calves at the first diarrhea episode, fecal samples were collected at two time points, the first day of diarrhea (fecal score > 1) and the day diarrhea resolved (second day of fecal score = 1). Using new gloves and sterile lubricant, fresh feces was collected by digital rectal stimulation into 20 mL polypropylene twist-top jars (The Cary Company, Addison, IL) and stored at -20˚C until analysis. Fecal samples were tested at the Dairy Epidemiology Laboratory, VMTRC by a veterinarian for E. coli K99, bovine rotavirus and coronavirus, and C. parvum oocysts using a commercial kit (Pathasure Enteritis 4; Biovet, Quebec, Canada) that is highly specific (> 90%) and sensitive (E. coli K99, 93%; rotavirus, 100%; coronavirus, 77%) [26, 27] . For calves with a first-day diarrhea sample on or before 7 days of age, both fecal samples were tested for all four pathogens. For calves with a first-day diarrhea sample after 7 days of age, both fecal samples were tested for three pathogens (C. parvum, bovine rotavirus and coronavirus). Samples from calves older than 7 days of age were not tested for E. coli K99 based on calves' susceptibility [28] . Testing was performed according to kit manufacturer guidelines, and a low-temperature incubator (Fisher Scientific, Model 146, Pittsburgh, PA) was used during incubation periods. Test results were recorded as positive or negative using control wells for color comparison. If the color change was darker than the negative control, the sample was considered positive. Milk zinc. For each of the 107 study days (December 15, 2015 to March 31, 2016) of zinc supplementation, approximately 1.5 mL of treated milk from two bottles of each treatment group were randomly collected into 2.0 mL microcentrifuge tubes with polypropylene snap caps (Fisher Scientific, Pittsburgh, PA), and stored at −20˚C until analysis. At the time of analysis, milk samples were thawed at 4˚C, vortexed, pooled by week and treatment group, and analyzed for zinc concentration (ppm) by ICP-MS (inductively coupled plasma mass spectrometry) at the CAHFS Laboratory. Quality control samples, including method blanks, laboratory control spikes, National Institute of Standards and Technology (NIST) reference materials (NIST 1640), and a spiked milk sample, were run with each set of study samples. Data analysis was performed using R Statistic Software version 3.3.1 and Stata IC 14.2 (College Station, TX). Statistical differences were determined at the 5% level of significance using per protocol analysis. An ANOVA was used to compare calves in each treatment group at enrollment with respect to birth weight (kg), serum total protein (g/dL), attitude score, and fecal score. Oral zinc dose at the start and end of treatment was calculated as the zinc supplementation dose (80 mg) divided by calf body weight (kg) at birth and on the last day of treatment, respectively. An ANOVA was used to compare oral zinc dose (mg/kg) at treatment start and end as well as mean body weight (kg) at birth, end of treatment, and hutch exit between treatment groups and between bulls and heifers. A Chi-Square test of Independence was used to compare the proportions of calves by sex as well as mortality between treatment groups. For all analyses, Tukey's Honest Significant Difference method was used to generate pairwise comparisons to further characterize significant differences identified by ANOVA. Residual diagnostics, including Residuals vs. Fitted, Scale-Location, Normal Q-Q, and Cook's distances plots, were used to validate all ANOVA model assumptions. The non-parametric Kruskal-Wallis Rank Sum test was used when assumptions were violated. For the randomly-sampled calves, Fisher exact tests were used to compare fecal pathogen prevalence on the first day of diarrhea and at diarrhea resolution between treatment groups. Pairwise comparisons with Bonferroni adjustment were used to identify specific differences in pathogen prevalence. An ANOVA was used to compare serum zinc concentration before and after treatment between treatment groups and between bulls and heifers. A Kruskal-Wallis Rank Sum test was used to compare rank sums of zinc concentrations in pooled milk samples from different treatment groups. Post-hoc Nemenyi-tests for pairwise multiple comparisons of ranked data were used to identify specific differences in zinc concentrations between groups. For all regression models in this study, univariate regression was first used to evaluate associations between individual predictor variables and outcomes. All variables with statistical and/or biological significance were initially included in multivariate regression models. The final models were built using a manual backwards elimination procedure, with a significance level of P > 0.05 as the removal criterion. Confounding was assessed using the method of change of estimates, where a 10% or greater change in the estimate of the treatment group regression coefficient between the models with and without the confounder variable was used as evidence of confounding [29] [30] [31] [32] . Variables identified as confounders were included in the final model. All possible interactions between treatment group and predictor variables were explored and retained in the final model if statistically significant. Microbiological cure. For the randomly-sampled calves, logistic regression was used to evaluate associations between microbiological cure and treatment group. Other predictor variables of interest included sex, serum total protein, and age on the first day of diarrhea. Microbiological cure was defined as a negative fecal ELISA test at resolution of clinical diarrhea for calves with a positive ELISA test on the first day of diarrhea for at least one of the four fecal pathogens (E. coli K99, bovine rotavirus and coronavirus, and C. parvum). Models were generated for each fecal pathogen individually and an overall model, which evaluated microbiological cure at clinical diarrhea resolution for calves that tested positive for any single pathogen on the first day of diarrhea. Serum total protein and calf age at first diarrhea were included in all final models to control for potential confounding by passive transfer status and age. Mean daily weight change. Linear regression was used to evaluate associations between ADG (kg) and treatment group during the treatment and pre-weaning periods separately. Other predictor variables of interest included sex, birth weight (kg), serum total protein, number of days having diarrhea, age, and volume (L) of milk, Calva, and NuLife electrolytes fed at either end of treatment or hutch exit. For each calf, ADG during the treatment and pre-weaning period was calculated as the difference between birth and end treatment or hutch exit weight, respectively, divided by the number of days between these time points. To explore the possibility of an interaction between treatment group, sex, and birth weight, the final linear regression model for ADG during the pre-weaning period was stratified by sex. Age at the end of treatment or hutch exit and number of days with diarrhea were dropped from all final models in favor of improved Akaike information criterion (AIC). Onset of diarrhea and clinical cure. For all survival analyses, diarrhea was defined as a fecal score greater than 1 while diarrhea cure was defined as the second consecutive day of normal feces (fecal score of 1) following the first diarrhea episode. Subsequent episodes of diarrhea were not included in the analysis. Calves that died or did not experience diarrhea or cure from diarrhea were censored. If fresh feces were not observed on daily calf hutch assessment, a fecal score was not recorded for that day and not included in the analyses. Kaplan-Meier analysis was used to determine median days to first diarrhea event and, for those calves that developed diarrhea during the assessment period, median days to clinical diarrhea cure. A Log Rank test of equality was used to compare survivor functions between treatments. Cox Proportional Hazards regression analysis was used to estimate and compare the hazard of diarrhea and diarrhea cure between treatment groups. Sex, age, serum total protein at enrollment, birth weight (kg), antimicrobial therapy, and application of fresh straw to the hutches were evaluated as predictor variables and potential confounders. When modeling the hazard of diarrhea cure, a binary variable termed therapeutic supplementation indicating whether calves were treated with either ZM, ZS or placebo for all or at least the first 5 days of diarrhea was evaluated as an additional covariate. A five-day period was selected by the authors based on clinical experience, as five days represents a reasonable duration over which most therapeutic treatments for calf diarrhea should be applied and be expected to alleviate disease. The proportional hazards assumption that the hazard of diarrhea is independent of time was assessed using analysis of Schoenfeld residuals and testing whether the log hazard-ratio function is constant over time. Any variable found to violate the proportional hazards assumption was included in the final regression model as a time varying covariate. A total of 1,513 calves were enrolled in the trial. However, due to failure to immediately recognize exclusion criteria, 23 calves were excluded shortly after enrollment. In addition, 8 calves were excluded due to treatment errors. Therefore, a total of 1,482 calves (placebo = 500, ZM = 491, ZS = 491) were included in the final analyses. A total of 242 calves (16.3%) had minimal fecal output at the time of enrollment while 125 calves (8.4%) had abnormal fecal scores of 2 or 3 that were described as meconium. All enrolled calves appeared healthy on visual assessment and hence were assumed to have a normal fecal score at the time of enrollment. The three treatment groups at enrollment did not differ significantly in mean birth weight (kg) (P = 0.244), mean serum total protein (g/dL) (P = 0.541), mean attitude score (P = 0.845), mean fecal score (P = 0.522), as shown in Table 1 , or distribution of calf sex (P = 0.472). Of the 1,482 study calves, 21 (1.4%) died during the trial: 5 calves in the placebo group (1.0%), 11 calves in the ZM group (2.2%), and 5 calves in the ZS group (1.0%). Of these 21 calves that died, 14 (66.7%) were bulls and 7 (33.3%) were heifers. Eighteen of the 21 calves (85.7%) were found dead, rather than euthanized, due to acute and spontaneous death without previous obvious clinical signs of disease. The remaining 3 calves were euthanized prior to death due to severe and/or prolonged morbidity. Characteristics and causes of death based on field necropsy of these calves can be found in S4 Table. There was no significant difference in the proportion of calves that died between treatment groups (P = 0.168), though mortality was significantly higher in bulls compared to heifer calves (P = 0.049). Birth weight data were available for all 1,482 calves. Due to calf mortality between enrollment and completion of treatment (n = 4), end treatment weight data were available for 1,478 calves. Similarly, due to calf mortality (n = 21) or missing data for body weight at hutch exit from the dairy's records (n = 40), hutch exit weight data were available for 1,421 calves. A summary of body weight data stratified by treatment group and sex is presented in Table 2 . Within each treatment group, bull calves showed consistently higher birth weight (P < 0.001), end treatment weight (P < 0.001), and exit hutch weight (P < 0.001) compared to heifer calves. However, at birth, end of treatment, or hutch exit there were no differences in body weight between treatment groups for bulls (P > 0.1) or heifers (P > 0.1). The mean attitude scores during the study period were 1.2 across the three treatment groups (P = 0.208). Of 1,482 calves included in the final analysis, A total of 629 treated milk samples were obtained throughout the 107-day study period and pooled by treatment group and week, yielding a total of 16 pooled samples per treatment group. Zinc concentrations (ppm) were significantly higher in pooled milk samples treated with ZM (P < 0.001) and ZS (P < 0.001) compared to placebo-treated samples, and there were no significant differences between ZM and ZS-treated samples (P = 1.000), as shown in S5 Table. Within zinc treatment groups, oral zinc dose at the start and end of treatment is summarized by sex in S6 Table. For both ZM-and ZS-treated calves, oral zinc dose at the start (P < 0.001) and end (P < 0.001) of treatment was significantly higher in heifer versus bull calves. Serum zinc concentrations before and after treatment were obtained from 36 calves (n = 12 for each treatment group) and are summarized in S7 Table. Overall, there were no significant differences in mean pre-treatment serum zinc concentrations between treatment groups (P = 0.233). Mean post-treatment serum zinc concentrations were significantly higher in calves treated with ZM (P < 0.001) and ZS (P = 0.002) compared to placebo-treated calves, and there were no significant differences among calves treated with ZM and ZS (P = 0.406). Stratification of serum zinc data by treatment group and sex demonstrated that for ZM-treated calves, heifers had a numerically higher post-treatment serum zinc concentration compared to bulls, though the difference was not statistically significant (P = 0.199). In contrast, in ZStreated calves, heifers had a numerically lower post-treatment serum zinc concentration compared to bulls, though the difference was also not statistically significant (P = 0.538). Fecal analysis data were analyzed for 92 of the 127 randomly-selected calves. The remaining 35 calves were not included in the analysis due to not acquiring diarrhea during the assessment period (n = 10), death prior to final sampling (n = 1), exclusion due to improper treatment regimen (n = 1), or incorrect sampling day (n = 23).The 92 calves had a mean age at onset of diarrhea of 13.3, 11.0 and 11.3 days for ZM, ZS and placebo-treated groups, respectively; and a mean age at resolution of diarrhea of 18.8, 16.1 and 15.7 days for ZM, ZS and placebo-treated groups, respectively. There were no significant differences in the prevalence of E. coli K99 (P = 0.694), rotavirus (P = 0.331), coronavirus (P = 0.819), or C. parvum (P = 0.719) fecal shedding on the first day of diarrhea between treatment groups (S8 Table) . There were no significant differences in the prevalence of E. coli K99 (P = 0.256), rotavirus (P = 0.344), or coronavirus (P = 1.000) fecal shedding at resolution of diarrhea between treatment groups, though there was a difference in C. parvum (P = 0.006) fecal shedding between treatment groups (S9 Table) . The prevalence of C. parvum fecal shedding at resolution of diarrhea was significantly higher in calves treated with ZM (P = 0.009) and ZS (P = 0.023) compared to placebo-treated calves, and there were no significant differences among calves treated with ZM and ZS (P = 1.000). Results of logistic regression models for overall and pathogen-specific microbiological cure are presented in Tables 3-5 . For pathogen-specific cure, all calves that tested positive on the first day of diarrhea for coronavirus (n = 4) tested negative at clinical diarrhea resolution. For calves that tested positive on the first day of diarrhea for E. coli K99 (n = 9), all placebo-treated calves (n = 2) tested negative at clinical diarrhea resolution while half (n = 2) of all ZS-treated calves tested either positive or negative at clinical diarrhea resolution, resulting in omission of ZS treatment variable from the model due to collinearity. Hence, logistic regression analyses for microbiological cure in calves that tested positive for coronavirus and E. coli K99 were not possible. For 59 calves that tested positive for rotavirus on the first day of diarrhea (Table 3) , calves treated with ZM had a 50% increased odds of testing negative at diarrhea resolution compared to placebo-treated calves, though this difference was not significant (P = 0.549). Likewise, calves treated with ZS had 100% increased odds (2 times the odds) of testing negative for rotavirus at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.314). However, this model demonstrated a significant main effect of serum total protein, such that for every 1 unit (g/dL) increase in serum total protein at enrollment, the odds of microbiological cure of rotavirus decreased by 79% (P = 0.026). For 40 calves that tested positive for Cryptosporidium parvum on the first day of diarrhea (Table 4) , calves treated with ZM had an 87% reduced odds of testing negative at diarrhea resolution Effect of zinc on diarrhea in calves compared to placebo-treated calves, though this difference was not significant (P = 0.119). Likewise, calves treated with ZS had a 74% reduced odds of testing negative for Cryptosporidium parvum at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.183). For 55 calves that tested positive for any one of the four fecal pathogens (E. coli K99, rotavirus, coronavirus, Cryptosporidium parvum) on the first day of diarrhea (Table 5) , calves treated with ZM had a 25% increased odds of testing negative at diarrhea resolution compared to placebo-treated calves, though this difference was not significant (P = 0.769). Likewise, calves treated with ZS had a 52% increased odds of testing negative for Cryptosporidium parvum at diarrhea resolution compared to placebo-treated calves, though this difference was also not significant (P = 0.633). However, this model demonstrated that heifer calves had 71% lower odds of microbiological cure of any single fecal pathogen compared to bull calves (P = 0.076). A total of 1,482 calves were included in the linear regression model results for ADG during the treatment period, which are presented in Table 6 . There was no significant difference in ADG for ZM-or ZS-treated calves compared to placebo-treated calves, though there were significant main effects of sex, birth weight, and milk volume. Specifically, heifer calves gained 70 g bodyweight per day less compared to bull calves (P < 0.001). For every 1 kg increase in birth weight, calves gained 16 g per day less than their herd mates (P < 0.001). For every 1 L increase in milk volume per day during the treatment period, calves gained an additional 13 g per day (P < 0.001). Table 7 summarizes the linear regression analysis of ADG during the pre-weaning period for 1,421 calves which showed a significant difference in ADG for ZM-treated calves compared to placebo-treated calves and in bull versus heifer calves. Milk volume had a significant effect on ADG, such that for every 1 L increase in milk volume per day during the treatment period, calves gained an additional 2 g bodyweight per day (P = 0.001). Results of the final model showed a significant main effect for ZM-treated bull calves and a significant interaction term for ZM treatment by sex. After controlling for milk volume received during the treatment Table 6 calves (n = 1,482) Effect of zinc on diarrhea in calves period, ZM-treated bulls gained 22 g body weight per day on average more than placebotreated bull calves (P = 0.042) and ZM-treated heifers gained 12 g less body weight per day on average compared to placebo-treated heifers (P = 0.019). When considering the model coefficients for treatment group, sex, and their interaction, bull calves treated with ZM gained 454 g per day (0.432 + 0.022) while female calves treated with ZM gained 0.404 g per day (0.432 + 0.022-0.016-0.034), hence 50 g per day more gain in male calves compared to heifers treated with ZM (P = 0.019). For ZS-treated calves, there was a numerical decrease in weight gain of 5 g per day in bulls and 11 g per day in heifers compared to placebo-treated calves, though the differences were not significant (P = 0.673 bulls; P = 0.681 heifers). Linear regression models of ADG during the pre-weaning period were stratified by sex in order to avoid interpreting a three-way interaction between treatment group, sex, and birth weight. In the heifer model (S10 Table) , the interaction between ZM treatment and birth weight was significant which implied that birth weight modified the effect of ZM treatment on ADG. At a 29 kg birth weight (two standard deviations below the mean), ZM-treated heifers gained 49 g body weight per day on average less than placebo-treated heifers (P = 0.037). However, at a 49 kg birth weight (two standard deviations above the mean), ZM-treated heifers gained 30 g body weight per day on average more than placebo-treated heifers (P = 0.037). In the bull calves model (S11 Table) there was no significant interaction between treatment group and birth weight. A total of 1,482 calves were included in the Kaplan-Meier survival analysis of time to first diarrhea event (Fig 1) . There were no significant differences in median age at onset of diarrhea, specifically, 8, 8 and 7 days for the ZM, ZS and placebo-treated calves, respectively (P = 0.402). Cox proportional hazard regression model for diarrhea hazard are presented in Table 8 . After controlling for age, calves treated with ZM had a 14.7% reduced hazard of diarrhea compared to placebo-treated calves (P = 0.015). Calves treated with ZS had 13.9% reduced hazard of diarrhea compared to placebo-treated calves (P = 0.022). calves (n = 1,421) A total of 1,394 calves were included in the Kaplan-Meier survival analysis of time to clinical diarrhea cure (Fig 2) , as 88 calves failed to acquire diarrhea during the assessment period. There were no significant differences in the median days to diarrhea cure which was 7 days across all 3 treatment groups (P = 0.264). Cox proportional hazard regression model for diarrhea cure hazard are presented in Table 9 . Results of the final model showed a significant interaction term between treatment and therapeutic supplementation as well as the need for age as a time varying covariate. When considering calves that did not receive supplementation, respective to each of the 3 groups, for at least the first five days of diarrhea there was no significant difference between either ZM-and ZS-treated calves compared to placebo-treated calves (P = 0.223 ZM, P = 0.134 ZS). However, when considering calves that were supplemented for at least the first five days of diarrhea, ZM-treated calves experienced a 21.4% higher hazard of cure from diarrhea compared to placebo-treated calves (P = 0.027). Likewise, ZS-treated calves experienced a 13.0% higher hazard of cure from diarrhea compared to placebo-treated calves (P = 0.040). The current trial demonstrated evidence for the beneficial effect of ZM on ADG and neonatal diarrhea as well as an effect of ZS on diarrhea in dairy calves during the pre-weaning period. It is important to consider these results in the context of the entire pre-weaning and hutch period. On average, after 90 days from birth to hutch exit, placebo-treated bull calves gained 38.88 kg body weight while ZM-treated bull calves gained an additional 1.98 kg (40.86 kg). In contrast, the effect of zinc on weight gain in treated heifers depended on birth weight. Low birth weight heifers treated with ZM gained on average less than a placebo-treated heifer of the same birth weight. In contrast, high birth weight heifers treated with ZM gained more than placebo-treated heifers of the same birth weight. The switch in direction of the association between ZM treatment and ADG in heifer calves depending on birth weight suggests a doseresponse effect rather than a true sex-specific effect of ZM on ADG. Hence, low birth weight calves (including heifers) may require a lower dose of ZM to mitigate any negative effect of what is otherwise a suitable dose for higher birth weight calves. These findings are in agreement with a previous randomized clinical trial testing the effect of daily oral zinc in diarrheic neonatal Holstein calves which, showed that ZM-treated calves had a numerically, though not significantly increased ADG compared to calves treated with zinc oxide or placebo due to small sample size [11] . In general, our trial findings are in agreement with the large body of human literature supporting the use of oral zinc for the prevention and treatment of diarrhea and impaired growth in children [5, 10, 33] . Zinc supplementation is widely accepted by global health organizations as a vital component of therapy for childhood diarrhea [3, 4] , however, recent reviews of the literature demonstrated heterogeneity in study results on the basis of age, baseline zinc status, geographic location, and supplementation regimen [10, 34] . Similar to our findings, a sex-specific response to zinc supplementation has been demonstrated in several human studies. Zinc gluconate administered for diarrhea prevention reduced the incidence of dysentery in treated boys but not girls [35] ; when given therapeutically, it reduced diarrhea duration and frequency more dramatically in boys compared to girls [36] . Similarly, zinc sulfate was shown to improve Effect of zinc on diarrhea in calves diarrhea outcomes in boys but improved growth rates in girls [13] . Broadly, these differences between male and female responses to zinc supplementation are not understood, though theories regarding differences in immune function and response [13, 35] , diarrhea etiology [13] , and nutrient requirements [35] have been proposed. In the current study, ZM-treated bulls demonstrated increased ADG compared to placebo-treated bulls while ZM-treated heifers demonstrated decreased ADG compared to placebo-treated heifers. However, due to a significant interaction between ZM treatment and birth weight, this reduction in ADG in ZMtreated heifers was overcome with increasing birth weight, such that ZM-treated heifers with birth weights above 42 kg experienced increased ADG during the pre-weaning period, compared to placebo-treated heifers with birth weights above 42 kg. Differences in the growth response to ZM supplementation between bull and heifer calves may have been related to its effect on feed intake. Previous research on the effects of feeding various doses of oral zinc oxide to pre-ruminant dairy calves demonstrated that high levels of oral zinc supplementation resulted in reduced feed intake [23] . In the current trial, oral ZM dose was estimated to be significantly higher in heifers compared to bulls due to the significantly lower birth weight of heifers. Additionally, serum zinc concentrations in ZM-treated heifers were numerically higher than that of bulls, though this difference was not significant, likely due to the small sample size. Perhaps the higher zinc dose in heifers was associated with reduced feed intake, leading to reduced growth, and that this effect was more pronounced for ZM compared to ZS. The fact that ZM-treated heifers with birth weights approaching those of average bull calves (and, therefore, a similar zinc dose to that in bulls) experienced an increase in ADG over placebo-treated heifers similar to that of bull calves partially supports this theory. Although management practices on the study dairy were designed to be identical for both bulls and heifers, it is possible that subtle, unrecognized differences in nutritional and health management may also have contributed to sex-specific differences in weight gain. Nevertheless, future trials are warranted to investigate the potential differences in the dose-response to zinc supplementation between bulls and heifers. We hypothesized that ADG would be increased in zinc-supplemented calves compared to placebo-supplemented calves due to the potential preventive and therapeutic effects of zinc supplementation on neonatal diarrhea. In other words, calf diarrhea is mitigated by zinc supplementation and, therefore, on the causal pathway between zinc and ADG. However, considering the similarly-reduced hazard of diarrhea and increased hazard of cure from diarrhea in both ZM and ZS treatment groups but a lack of effect of ZS on ADG, it is likely that the effect of ZM on ADG is not solely mediated through its effects on diarrhea. Differences in effectiveness between organic and inorganic formulations also may exist. In fact, the underlying mechanism of action of oral zinc remains unknown [6] . Several theories of the mechanisms of action of zinc in the prevention and treatment of childhood diarrhea exist, including a mucosal-protective role, a diarrhea-induced zinc deficiency, an essential element in cell-mediated immunity, and a modifier of intra-luminal electrolyte secretion and absorption [6, [37] [38] [39] . The clinical and practical implications of effects of ZM supplementation on ADG and diarrhea must be considered. Pre-weaned calf diarrhea remains an ongoing issue for the dairy industry. The deleterious effects on calf health and performance and the resulting economic burden create a strong incentive to treat and prevent diarrhea in pre-weaned calves. On large dairy operations like those in California's Central Valley, small changes in disease incidence and duration as well as animal growth and performance can have profound economic consequences. As a non-antimicrobial product, zinc may become increasingly attractive as antimicrobials in livestock feed are under increased scrutiny and regulation due to concerns about antimicrobial resistance [2, 40] . Prevalence of C. parvum fecal shedding in a random sample of 92 study calves at onset and resolution of diarrhea was significantly higher in calves treated with zinc compared to Placebo-treated calves. In contrast, a previous study where calves that tested positive for C. parvum at the start of diarrhea and were treated with ZM had 16 times higher odds of being fecal ELISA negative at exit compared to the Placebo group (P = 0.08; power = 72.3%) [11] . The difference in findings may be due to the differences in the timing of diarrhea across treatment groups. For the current study's random sample of calves that acquired, survived, and were sampled on the correct days, the mean age of calves on both onset and resolution of diarrhea was higher for ZM and ZS calves compared to placebo-treated calves. Although C. parvum oocyst shedding in infected calves can occur as early as 3 days of age, peak shedding occurs at about 14 days of age [41] . It is possible that the increase in prevalence of C. parvum shedding in ZM and ZS treated groups was due to the increased age of zinc-treated calves compared to placebo-treated calves at resolution of diarrhea. The latter explanation is also supported by our findings that the odds of microbiological cure from C. parvum significantly decreased in older calves, with no significant differences in the odds of cure between treatment groups. In addition, the current testing did not estimate the concentration of C. parvum shedding which may still differ between treatment groups. Despite the large sample size, the current trial was limited to a single California dairy, which may represent other large dairies but does not reflect all the dairy management systems in California or elsewhere. Additionally, our results show that calves respond to zinc supplementation for diarrhea prevention differently depending on chemical formulation and calf sex. The latter could be due to differences in body weight between bulls and heifers and may point towards the need for sex-specific dosing. Furthermore, the current research did not evaluate the potential economic utility of zinc supplementation. Future studies on more accurate dosing of zinc by calf sex, the practical feasibility of weight-based dosing, and the expected cost-effectiveness of zinc administration as part of the management of pre-weaned dairy calves are warranted. Finally, our clinical trial was performed on a single, large, predominately Holstein, California dairy over a six-month period, which precluded our ability to evaluate differences due to season or breed. Hence, future studies to assess any modifying effect of breed and seasonal differences on the effect of zinc on calf health and weight gain are also needed. The current double blind, block-randomized placebo controlled clinical trial tested the effect of a prophylactic daily oral zinc supplementation in neonatal Holstein calves. Bull calves treated with ZM had a significantly increased ADG (22 g per day) during the pre-weaning period compared to placebo-treated bulls. In comparison, ZM-treated heifers had significantly lower average daily gain (9 g per day) compared to placebo-treated heifers, although higher ZM doses in low birthweight heifers may explain the lower ADG. Calves treated with either ZM or ZS had significantly lower risks of diarrhea and significantly higher risk of cure from diarrhea over the first 30 days of life compared to placebo-treated calves and hence the current trial demonstrated that zinc supplementation delayed diarrhea and expedited diarrhea recovery in pre-weaned calves. Additionally, zinc improved weight gain differentially in bulls compared to heifers, indicating the need for further research to investigate zinc dosing in calves. Supporting information S1 (DOCX) S1 Dataset. Raw data collected from trial, organized as separate excel sheets for enrollment, daily assessment, serum total protein, birth weight, exit treatment weight, exit trial weight, serum zinc testing, fecal samples, fecal testing, milk testing, and dead calves. (XLSX)

What preventative measure has been taken to decrease the incidence of diarrhea in children?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the death toll in the 1918-1919 Spanish Influenza epidemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

How many people were infected during the 1918 Spanish Influenza epidemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the case fatality rate in the 1918 Spanish Influenza epidemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the death toll in the 1918-1919 Spanish Influenza epidemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Are the modern day Influenza viruses related to the 1918 Spanish Influenza virus?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Why is the Spanish Influenza virus the Mother of the modern influenza viruses?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

When was it determined that the 1918 pandemic was caused by the H1N1 Influenza virus?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Did the Spanish Influenza or Swine flu or the H1N1 virus disappear in humans for some time?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

When did the Swine Flu (Spanish Influenza) virus reappear in humans?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What descendant lineages of the swine flu (Spanish Influenza) virus were identified in 2006?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Are the modern descendant influenza viruses as dangerous as the 1918 parent swine flu (Spanish Influenza) H1N1 virus?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

How dangerous are the modern H1N1 (swine flu) and the H3N2 (Influenza A) viruses compared to the 1918 H1N1 (swine flu Spanish Influenza) viruses?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Are the descendant H1N1 strains of the 1918 H1N1 swine flu (Spanish Influenza) virus, still prevalent?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Is the origin and epidemiology of the 1918 swine flu (Spanish Influenza) known?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What is the geographical origin of the H1N1 swine flu ?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Is the geographical origin of the 1918 H1N1 swine flu known?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What is an unique feature of the 1918 swine flu?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What season or time of the year do the new strains of influenza emerge?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Once appeared, when do the influenza like diseases occur in subsequent years?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

When did the first wave of the H1N1 swine flu (Spanish Influenza) occur?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the death rate in the first wave of the 1918 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

When were the second and the third wave of the 1918-1919 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the primary difference between the first wave and the 2nd and 3rd wave of the 1918-1919 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Why the human influenza viruses do not disappear after herd immunity is developed?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What are the circumstances that promote the spread of influenza virus?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Do seasonal temperatures and humidity explain the appearance of the three waves of the 1918 swine flu?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Which virus samples from the 1918 swine flu pandemic have been identified?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Are viruses in the first and third waves of the 1918 swine flu pandemic same or derived from the virus from the second wave of the swine flu?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Was the 1918 swine flu virus novel to humans are was it derived from older viruses?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Do avian flu viruses change over long periods?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What is the typical age profile of mortality in Influenza diseases?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the age profile of mortality in the 1918 swine flu?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Which age group was most susceptible to die during the 1918 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What was the death rate among children during the 1918 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

What theory provides partial explanation for the age-specific profile of the death rate in the 1918 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Is there a difference in the pathologic feature and course of disease between modern influenza pandemics and the 1918 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Could the 1918 swine flu virus been controlled by modern day drugs or vaccines?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Why was there such a high death rate in the 19118 swine flu pandemic?

1918 Influenza: the Mother of All Pandemics Jeffery K. Taubenberger" and David M. Morens1- The “Spanish" influenza pandemic of 1918—1919, which caused :50 million deaths worldwide, remains an ominous warning to public health. Many questions about its origins, its unusual epidemiologic features, and the basis of its pathogenicity remain unanswered. The public health implications of the pandemic therefore remain in doubt even as we now grapple with the feared emergence of a pandemic caused by H5N1 or other virus. However, new information about the 1918 virus is emerging, for example, sequencing of the entire genome from archival autopsy tis- sues. But, the viral genome alone is unlikely to provide answers to some critical questions. Understanding the 1918 pandemic and its implications for future pandemics requires careful experimentation and in-depth historical analysis. ”Curiouser and curiouser/ ” criedAlice Lewis Carroll, Alice’s Adventures in Wonderland, 1865 An estimated one third of the world’s population (or z500 million persons) were infected and had clinical- ly apparent illnesses (1,2) during the 191871919 influenza pandemic. The disease was exceptionally severe. Case- fatality rates were >2.5%, compared to <0.1% in other influenza pandemics (3,4). Total deaths were estimated at z50 million (577) and were arguably as high as 100 mil- lion (7). The impact of this pandemic was not limited to 191871919. All influenza A pandemics since that time, and indeed almost all cases of influenza A worldwide (except- ing human infections from avian Viruses such as H5N1 and H7N7), have been caused by descendants of the 1918 Virus, including “drifted” H1N1 Viruses and reassorted H2N2 and H3N2 Viruses. The latter are composed of key genes from the 1918 Virus, updated by subsequently-incor— porated avian influenza genes that code for novel surface *Armed Forces Institute of Pathology, Rockville, Maryland, USA; and TNational Institutes of Health, Bethesda, Maryland, USA proteins, making the 1918 Virus indeed the “mother” of all pandemics. In 1918, the cause of human influenza and its links to avian and swine influenza were unknown. Despite clinical and epidemiologic similarities to influenza pandemics of 1889, 1847, and even earlier, many questioned whether such an explosively fatal disease could be influenza at all. That question did not begin to be resolved until the 1930s, when closely related influenza Viruses (now known to be H1N1 Viruses) were isolated, first from pigs and shortly thereafter from humans. Seroepidemiologic studies soon linked both of these viruses to the 1918 pandemic (8). Subsequent research indicates that descendants of the 1918 Virus still persists enzootically in pigs. They probably also circulated continuously in humans, undergoing gradual antigenic drift and causing annual epidemics, until the 1950s. With the appearance of a new H2N2 pandemic strain in 1957 (“Asian flu”), the direct H1N1 Viral descen- dants 0f the 1918 pandemic strain disappeared from human circulation entirely, although the related lineage persisted enzootically in pigs. But in 1977, human H1N1 Viruses suddenly “reemerged” from a laboratory freezer (9). They continue to circulate endemically and epidemically. Thus in 2006, 2 major descendant lineages of the 1918 H1N1 Virus, as well as 2 additional reassortant lineages, persist naturally: a human epidemic/endemic H1N1 line- age, a porcine enzootic H1N1 lineage (so-called classic swine flu), and the reassorted human H3N2 Virus lineage, which like the human H1N1 Virus, has led to a porcine H3N2 lineage. None of these Viral descendants, however, approaches the pathogenicity of the 1918 parent Virus. Apparently, the porcine H1N1 and H3N2 lineages uncom- monly infect humans, and the human H1N1 and H3N2 lin- eages have both been associated with substantially lower rates ofillness and death than the virus of 1918. In fact, cur- rent H1N1 death rates are even lower than those for H3N2 lineage strains (prevalent from 1968 until the present). H1N1 Viruses descended from the 1918 strain, as well as H3N2 Viruses, have now been cocirculating worldwide for 29 years and show little evidence of imminent extinction. Trying To Understand What Happened By the early 1990s, 75 years of research had failed to answer a most basic question about the 1918 pandemic: why was it so fatal? No Virus from 1918 had been isolated, but all of its apparent descendants caused substantially milder human disease. Moreover, examination of mortality data from the 1920s suggests that within a few years after 1918, influenza epidemics had settled into a pattern of annual epidemicity associated with strain drifting and sub- stantially lowered death rates. Did some critical Viral genet- ic event produce a 1918 Virus of remarkable pathogenicity and then another critical genetic event occur soon after the 1918 pandemic to produce an attenuated H1N1 Virus? In 1995, a scientific team identified archival influenza autopsy materials collected in the autumn of 1918 and began the slow process of sequencing small Viral RNA fragments to determine the genomic structure of the causative influenza Virus (10). These efforts have now determined the complete genomic sequence of 1 Virus and partial sequences from 4 others. The primary data from the above studies (11717) and a number of reviews covering different aspects of the 1918 pandemic have recently been published ([8720) and confirm that the 1918 Virus is the likely ancestor of all 4 of the human and swine H1N1 and H3N2 lineages, as well as the “extinct” H2N2 lineage. No known mutations correlated with high pathogenicity in other human or animal influenza Viruses have been found in the 1918 genome, but ongoing studies to map Virulence factors are yielding interesting results. The 1918 sequence data, however, leave unanswered questions about the ori- gin of the Virus (19) and about the epidemiology of the pandemic. When and Where Did the 1918 Influenza Pandemic Arise? Before and after 1918, most influenza pandemics developed in Asia and spread from there to the rest of the world. Confounding definite assignment of a geographic point of origin, the 1918 pandemic spread more or less simultaneously in 3 distinct waves during an z12-month period in 191871919, in Europe, Asia, and North America (the first wave was best described in the United States in March 1918). Historical and epidemiologic data are inade- quate to identify the geographic origin of the Virus (21), and recent phylogenetic analysis of the 1918 Viral genome does not place the Virus in any geographic context ([9). Although in 1918 influenza was not a nationally reportable disease and diagnostic criteria for influenza and pneumonia were vague, death rates from influenza and pneumonia in the United States had risen sharply in 1915 and 1916 because of a major respiratory disease epidemic beginning in December 1915 (22). Death rates then dipped slightly in 1917. The first pandemic influenza wave appeared in the spring of 1918, followed in rapid succes- sion by much more fatal second and third waves in the fall and winter of 191871919, respectively (Figure 1). Is it pos- sible that a poorly-adapted H1N1 Virus was already begin- ning to spread in 1915, causing some serious illnesses but not yet sufficiently fit to initiate a pandemic? Data consis- tent with this possibility were reported at the time from European military camps (23), but a counter argument is that if a strain with a new hemagglutinin (HA) was caus- ing enough illness to affect the US national death rates from pneumonia and influenza, it should have caused a pandemic sooner, and when it eventually did, in 1918, many people should have been immune or at least partial- ly immunoprotected. “Herald” events in 1915, 1916, and possibly even in early 1918, if they occurred, would be dif- ficult to identify. The 1918 influenza pandemic had another unique fea- ture, the simultaneous (or nearly simultaneous) infection of humans and swine. The Virus of the 1918 pandemic like- ly expressed an antigenically novel subtype to which most humans and swine were immunologically naive in 1918 (12,20). Recently published sequence and phylogenetic analyses suggest that the genes encoding the HA and neu- raminidase (NA) surface proteins of the 1918 Virus were derived from an avianlike influenza Virus shortly before the start of the pandemic and that the precursor Virus had not circulated widely in humans or swine in the few decades before (12,15, 24). More recent analyses of the other gene segments of the Virus also support this conclu- sion. Regression analyses of human and swine influenza sequences obtained from 1930 to the present place the ini- tial circulation of the 1918 precursor Virus in humans at approximately 191571918 (20). Thus, the precursor was probably not circulating widely in humans until shortly before 1918, nor did it appear to have jumped directly from any species of bird studied to date (19). In summary, its origin remains puzzling. Were the 3 Waves in 1918—1 919 Caused by the Same Virus? If So, How and Why? Historical records since the 16th century suggest that new influenza pandemics may appear at any time of year, not necessarily in the familiar annual winter patterns of interpandemic years, presumably because newly shifted influenza Viruses behave differently when they find a uni- versal or highly susceptible human population. Thereafter, confronted by the selection pressures of population immu- nity, these pandemic Viruses begin to drift genetically and eventually settle into a pattern of annual epidemic recur- rences caused by the drifted Virus variants. Figure 1. Three pandemic waves: weekly combined influenza and pneumonia mortality, United Kingdom, 1918—1919 (21). In the 1918-1919 pandemic, a first or spring wave began in March 1918 and spread unevenly through the United States, Europe, and possibly Asia over the next 6 months (Figure 1). Illness rates were high, but death rates in most locales were not appreciably above normal. A sec- ond or fall wave spread globally from September to November 1918 and was highly fatal. In many nations, a third wave occurred in early 1919 (21). Clinical similari- ties led contemporary observers to conclude initially that they were observing the same disease in the successive waves. The milder forms of illness in all 3 waves were identical and typical of influenza seen in the 1889 pandem- ic and in prior interpandemic years. In retrospect, even the rapid progressions from uncomplicated influenza infec- tions to fatal pneumonia, a hallmark of the 191871919 fall and winter waves, had been noted in the relatively few severe spring wave cases. The differences between the waves thus seemed to be primarily in the much higher fre- quency of complicated, severe, and fatal cases in the last 2 waves. But 3 extensive pandemic waves of influenza within 1 year, occurring in rapid succession, with only the briefest of quiescent intervals between them, was unprecedented. The occurrence, and to some extent the severity, of recur- rent annual outbreaks, are driven by Viral antigenic drift, with an antigenic variant Virus emerging to become domi- nant approximately every 2 to 3 years. Without such drift, circulating human influenza Viruses would presumably disappear once herd immunity had reached a critical threshold at which further Virus spread was sufficiently limited. The timing and spacing of influenza epidemics in interpandemic years have been subjects of speculation for decades. Factors believed to be responsible include partial herd immunity limiting Virus spread in all but the most favorable circumstances, which include lower environ- mental temperatures and human nasal temperatures (bene- ficial to thermolabile Viruses such as influenza), optimal humidity, increased crowding indoors, and imperfect ven- tilation due to closed windows and suboptimal airflow. However, such factors cannot explain the 3 pandemic waves of 1918-1919, which occurred in the spring-sum- mer, summer—fall, and winter (of the Northern Hemisphere), respectively. The first 2 waves occurred at a time of year normally unfavorable to influenza Virus spread. The second wave caused simultaneous outbreaks in the Northern and Southern Hemispheres from September to November. Furthermore, the interwave peri- ods were so brief as to be almost undetectable in some locales. Reconciling epidemiologically the steep drop in cases in the first and second waves with the sharp rises in cases of the second and third waves is difficult. Assuming even transient postinfection immunity, how could suscep- tible persons be too few to sustain transmission at 1 point, and yet enough to start a new explosive pandemic wave a few weeks later? Could the Virus have mutated profoundly and almost simultaneously around the world, in the short periods between the successive waves? Acquiring Viral drift sufficient to produce new influenza strains capable of escaping population immunity is believed to take years of global circulation, not weeks of local circulation. And hav- ing occurred, such mutated Viruses normally take months to spread around the world. At the beginning of other “off season” influenza pan- demics, successive distinct waves within a year have not been reported. The 1889 pandemic, for example, began in the late spring of 1889 and took several months to spread throughout the world, peaking in northern Europe and the United States late in 1889 or early in 1890. The second recurrence peaked in late spring 1891 (more than a year after the first pandemic appearance) and the third in early 1892 (21 ). As was true for the 1918 pandemic, the second 1891 recurrence produced of the most deaths. The 3 recur- rences in 1889-1892, however, were spread over >3 years, in contrast to 191871919, when the sequential waves seen in individual countries were typically compressed into z879 months. What gave the 1918 Virus the unprecedented ability to generate rapidly successive pandemic waves is unclear. Because the only 1918 pandemic Virus samples we have yet identified are from second-wave patients ([6), nothing can yet be said about whether the first (spring) wave, or for that matter, the third wave, represented circulation of the same Virus or variants of it. Data from 1918 suggest that persons infected in the second wave may have been pro- tected from influenza in the third wave. But the few data bearing on protection during the second and third waves after infection in the first wave are inconclusive and do lit- tle to resolve the question of whether the first wave was caused by the same Virus or whether major genetic evolu- tionary events were occurring even as the pandemic exploded and progressed. Only influenza RNAipositive human samples from before 1918, and from all 3 waves, can answer this question. What Was the Animal Host Origin of the Pandemic Virus? Viral sequence data now suggest that the entire 1918 Virus was novel to humans in, or shortly before, 1918, and that it thus was not a reassortant Virus produced from old existing strains that acquired 1 or more new genes, such as those causing the 1957 and 1968 pandemics. On the con- trary, the 1918 Virus appears to be an avianlike influenza Virus derived in toto from an unknown source (17,19), as its 8 genome segments are substantially different from contemporary avian influenza genes. Influenza Virus gene sequences from a number offixed specimens ofwild birds collected circa 1918 show little difference from avian Viruses isolated today, indicating that avian Viruses likely undergo little antigenic change in their natural hosts even over long periods (24,25). For example, the 1918 nucleoprotein (NP) gene sequence is similar to that ofviruses found in wild birds at the amino acid level but very divergent at the nucleotide level, which suggests considerable evolutionary distance between the sources of the 1918 NP and of currently sequenced NP genes in wild bird strains (13,19). One way of looking at the evolutionary distance of genes is to com- pare ratios of synonymous to nonsynonymous nucleotide substitutions. A synonymous substitution represents a silent change, a nucleotide change in a codon that does not result in an amino acid replacement. A nonsynonymous substitution is a nucleotide change in a codon that results in an amino acid replacement. Generally, a Viral gene sub- jected to immunologic drift pressure or adapting to a new host exhibits a greater percentage of nonsynonymous mutations, while a Virus under little selective pressure accumulates mainly synonymous changes. Since little or no selection pressure is exerted on synonymous changes, they are thought to reflect evolutionary distance. Because the 1918 gene segments have more synony- mous changes from known sequences of wild bird strains than expected, they are unlikely to have emerged directly from an avian influenza Virus similar to those that have been sequenced so far. This is especially apparent when one examines the differences at 4-fold degenerate codons, the subset of synonymous changes in which, at the third codon position, any of the 4 possible nucleotides can be substituted without changing the resulting amino acid. At the same time, the 1918 sequences have too few amino acid difierences from those of wild-bird strains to have spent many years adapting only in a human or swine intermedi- ate host. One possible explanation is that these unusual gene segments were acquired from a reservoir of influenza Virus that has not yet been identified or sampled. All of these findings beg the question: where did the 1918 Virus come from? In contrast to the genetic makeup of the 1918 pandem- ic Virus, the novel gene segments of the reassorted 1957 and 1968 pandemic Viruses all originated in Eurasian avian Viruses (26); both human Viruses arose by the same mech- anismireassortment of a Eurasian wild waterfowl strain with the previously circulating human H1N1 strain. Proving the hypothesis that the Virus responsible for the 1918 pandemic had a markedly different origin requires samples of human influenza strains circulating before 1918 and samples of influenza strains in the wild that more closely resemble the 1918 sequences. What Was the Biological Basis for 1918 Pandemic Virus Pathogenicity? Sequence analysis alone does not ofier clues to the pathogenicity of the 1918 Virus. A series of experiments are under way to model Virulence in Vitro and in animal models by using Viral constructs containing 1918 genes produced by reverse genetics. Influenza Virus infection requires binding of the HA protein to sialic acid receptors on host cell surface. The HA receptor-binding site configuration is different for those influenza Viruses adapted to infect birds and those adapted to infect humans. Influenza Virus strains adapted to birds preferentially bind sialic acid receptors with 01 (273) linked sugars (27729). Human-adapted influenza Viruses are thought to preferentially bind receptors with 01 (2%) link- ages. The switch from this avian receptor configuration requires of the Virus only 1 amino acid change (30), and the HAs of all 5 sequenced 1918 Viruses have this change, which suggests that it could be a critical step in human host adaptation. A second change that greatly augments Virus binding to the human receptor may also occur, but only 3 of5 1918 HA sequences have it (16). This means that at least 2 H1N1 receptor-binding vari- ants cocirculated in 1918: 1 with high—affinity binding to the human receptor and 1 with mixed-affinity binding to both avian and human receptors. No geographic or chrono- logic indication eXists to suggest that one of these variants was the precursor of the other, nor are there consistent dif- ferences between the case histories or histopathologic fea- tures of the 5 patients infected with them. Whether the Viruses were equally transmissible in 1918, whether they had identical patterns of replication in the respiratory tree, and whether one or both also circulated in the first and third pandemic waves, are unknown. In a series of in Vivo experiments, recombinant influen- za Viruses containing between 1 and 5 gene segments of the 1918 Virus have been produced. Those constructs bearing the 1918 HA and NA are all highly pathogenic in mice (31). Furthermore, expression microarray analysis performed on whole lung tissue of mice infected with the 1918 HA/NA recombinant showed increased upregulation of genes involved in apoptosis, tissue injury, and oxidative damage (32). These findings are unexpected because the Viruses with the 1918 genes had not been adapted to mice; control experiments in which mice were infected with modern human Viruses showed little disease and limited Viral replication. The lungs of animals infected with the 1918 HA/NA construct showed bronchial and alveolar epithelial necrosis and a marked inflammatory infiltrate, which suggests that the 1918 HA (and possibly the NA) contain Virulence factors for mice. The Viral genotypic basis of this pathogenicity is not yet mapped. Whether pathogenicity in mice effectively models pathogenicity in humans is unclear. The potential role of the other 1918 pro- teins, singularly and in combination, is also unknown. Experiments to map further the genetic basis of Virulence of the 1918 Virus in various animal models are planned. These experiments may help define the Viral component to the unusual pathogenicity of the 1918 Virus but cannot address whether specific host factors in 1918 accounted for unique influenza mortality patterns. Why Did the 1918 Virus Kill So Many Healthy Young Ad ults? The curve of influenza deaths by age at death has histor- ically, for at least 150 years, been U-shaped (Figure 2), exhibiting mortality peaks in the very young and the very old, with a comparatively low frequency of deaths at all ages in between. In contrast, age-specific death rates in the 1918 pandemic exhibited a distinct pattern that has not been documented before or since: a “W—shaped” curve, similar to the familiar U-shaped curve but with the addition of a third (middle) distinct peak of deaths in young adults z20410 years of age. Influenza and pneumonia death rates for those 1534 years of age in 191871919, for example, were 20 times higher than in previous years (35). Overall, near- ly half of the influenza—related deaths in the 1918 pandem- ic were in young adults 20410 years of age, a phenomenon unique to that pandemic year. The 1918 pandemic is also unique among influenza pandemics in that absolute risk of influenza death was higher in those <65 years of age than in those >65; persons <65 years of age accounted for >99% of all excess influenza—related deaths in 191871919. In com- parison, the <65-year age group accounted for 36% of all excess influenza—related deaths in the 1957 H2N2 pandem- ic and 48% in the 1968 H3N2 pandemic (33). A sharper perspective emerges when 1918 age-specific influenza morbidity rates (21) are used to adj ust the W- shaped mortality curve (Figure 3, panels, A, B, and C [35,37]). Persons 65 years of age in 1918 had a dispro- portionately high influenza incidence (Figure 3, panel A). But even after adjusting age-specific deaths by age-specif— ic clinical attack rates (Figure 3, panel B), a W—shaped curve with a case-fatality peak in young adults remains and is significantly different from U-shaped age-specific case- fatality curves typically seen in other influenza years, e.g., 192871929 (Figure 3, panel C). Also, in 1918 those 5 to 14 years of age accounted for a disproportionate number of influenza cases, but had a much lower death rate from influenza and pneumonia than other age groups. To explain this pattern, we must look beyond properties of the Virus to host and environmental factors, possibly including immunopathology (e.g., antibody-dependent infection enhancement associated with prior Virus exposures [38]) and exposure to risk cofactors such as coinfecting agents, medications, and environmental agents. One theory that may partially explain these findings is that the 1918 Virus had an intrinsically high Virulence, tem- pered only in those patients who had been born before 1889, e.g., because of exposure to a then-circulating Virus capable of providing partial immunoprotection against the 1918 Virus strain only in persons old enough (>35 years) to have been infected during that prior era (35). But this the- ory would present an additional paradox: an obscure pre- cursor Virus that left no detectable trace today would have had to have appeared and disappeared before 1889 and then reappeared more than 3 decades later. Epidemiologic data on rates of clinical influenza by age, collected between 1900 and 1918, provide good evi- dence for the emergence of an antigenically novel influen- za Virus in 1918 (21). Jordan showed that from 1900 to 1917, the 5- to 15-year age group accounted for 11% of total influenza cases, while the >65-year age group accounted for 6 % of influenza cases. But in 1918, cases in Figure 2. “U-” and “W—” shaped combined influenza and pneumo- nia mortality, by age at death, per 100,000 persons in each age group, United States, 1911—1918. Influenza- and pneumonia- specific death rates are plotted for the interpandemic years 1911—1917 (dashed line) and for the pandemic year 1918 (solid line) (33,34). Incidence male per 1 .nao persunslage group Mortality per 1.000 persunslige group + Case—fataiity rale 1918—1919 Case fatalily par 100 persons ill wilh P&I pel age group Figure 3. Influenza plus pneumonia (P&l) (combined) age-specific incidence rates per 1,000 persons per age group (panel A), death rates per 1,000 persons, ill and well combined (panel B), and case-fatality rates (panel C, solid line), US Public Health Service house-to-house surveys, 8 states, 1918 (36). A more typical curve of age-specific influenza case-fatality (panel C, dotted line) is taken from US Public Health Service surveys during 1928—1929 (37). the 5 to 15-year-old group jumped to 25% of influenza cases (compatible with exposure to an antigenically novel Virus strain), while the >65-year age group only accounted for 0.6% of the influenza cases, findings consistent with previously acquired protective immunity caused by an identical or closely related Viral protein to which older per- sons had once been exposed. Mortality data are in accord. In 1918, persons >75 years had lower influenza and pneumonia case-fatality rates than they had during the prepandemic period of 191171917. At the other end of the age spectrum (Figure 2), a high proportion of deaths in infancy and early childhood in 1918 mimics the age pat- tern, if not the mortality rate, of other influenza pandemics. Could a 1918-like Pandemic Appear Again? If So, What Could We Do About It? In its disease course and pathologic features, the 1918 pandemic was different in degree, but not in kind, from previous and subsequent pandemics. Despite the extraordi- nary number of global deaths, most influenza cases in 1918 (>95% in most locales in industrialized nations) were mild and essentially indistinguishable from influenza cases today. Furthermore, laboratory experiments with recombi- nant influenza Viruses containing genes from the 1918 Virus suggest that the 1918 and 1918-like Viruses would be as sensitive as other typical Virus strains to the Food and Drug Administrationiapproved antiinfluenza drugs riman- tadine and oseltamivir. However, some characteristics of the 1918 pandemic appear unique: most notably, death rates were 5 7 20 times higher than expected. Clinically and pathologically, these high death rates appear to be the result of several factors, including a higher proportion of severe and complicated infections of the respiratory tract, rather than involvement of organ systems outside the normal range of the influenza Virus. Also, the deaths were concentrated in an unusually young age group. Finally, in 1918, 3 separate recurrences of influenza followed each other with unusual rapidity, resulting in 3 explosive pandemic waves within a year’s time (Figure 1). Each of these unique characteristics may reflect genetic features of the 1918 Virus, but understand- ing them will also require examination of host and envi- ronmental factors. Until we can ascertain which of these factors gave rise to the mortality patterns observed and learn more about the formation of the pandemic, predictions are only educated guesses. We can only conclude that since it happened once, analogous conditions could lead to an equally devastating pandemic. Like the 1918 Virus, H5N1 is an avian Virus (39), though a distantly related one. The evolutionary path that led to pandemic emergence in 1918 is entirely unknown, but it appears to be different in many respects from the cur- rent situation with H5N1. There are no historical data, either in 1918 or in any other pandemic, for establishing that a pandemic “precursor” Virus caused a highly patho- genic outbreak in domestic poultry, and no highly patho- genic avian influenza (HPAI) Virus, including H5N1 and a number of others, has ever been known to cause a major human epidemic, let alone a pandemic. While data bearing on influenza Virus human cell adaptation (e.g., receptor binding) are beginning to be understood at the molecular level, the basis for Viral adaptation to efficient human-to- human spread, the chief prerequisite for pandemic emer- gence, is unknown for any influenza Virus. The 1918 Virus acquired this trait, but we do not know how, and we cur- rently have no way of knowing whether H5N1 Viruses are now in a parallel process of acquiring human-to-human transmissibility. Despite an explosion of data on the 1918 Virus during the past decade, we are not much closer to understanding pandemic emergence in 2006 than we were in understanding the risk of H1N1 “swine flu” emergence in 1976. Even with modern antiviral and antibacterial drugs, vaccines, and prevention knowledge, the return of a pan- demic Virus equivalent in pathogenicity to the Virus of 1918 would likely kill >100 million people worldwide. A pandemic Virus with the (alleged) pathogenic potential of some recent H5N1 outbreaks could cause substantially more deaths. Whether because of Viral, host or environmental fac- tors, the 1918 Virus causing the first or ‘spring’ wave was not associated with the exceptional pathogenicity of the second (fall) and third (winter) waves. Identification of an influenza RNA-positive case from the first wave could point to a genetic basis for Virulence by allowing differ- ences in Viral sequences to be highlighted. Identification of pre-1918 human influenza RNA samples would help us understand the timing of emergence of the 1918 Virus. Surveillance and genomic sequencing of large numbers of animal influenza Viruses will help us understand the genet- ic basis of host adaptation and the extent of the natural reservoir of influenza Viruses. Understanding influenza pandemics in general requires understanding the 1918 pan- demic in all its historical, epidemiologic, and biologic aspects. Dr Taubenberger is chair of the Department of Molecular Pathology at the Armed Forces Institute of Pathology, Rockville, Maryland. His research interests include the molecular patho- physiology and evolution of influenza Viruses. Dr Morens is an epidemiologist with a long-standing inter- est in emerging infectious diseases, Virology, tropical medicine, and medical history. Since 1999, he has worked at the National Institute of Allergy and Infectious Diseases. References 1. Frost WH. Statistics of influenza morbidity. Public Health Rep. 19203558497. 2. Bumet F, Clark E. Influenza: a survey ofthe last 50 years in the light of modern work on the Virus of epidemic influenza. Melbourne: MacMillan; 1942. 3. Marks G, Beatty WK. Epidemics. New York: Scribners, 1976. 4. Rosenau MJ, Last JM. Maxcy-Rosenau preventative medicine and public health. New York: Appleton-Century-Crofts; 1980. 5. Crosby A. America’s forgotten pandemic. Cambridge (UK): Cambridge University Press;1989. 6. Patterson KD, Pyle GF. The geography and mortality of the 1918 influenza pandemic. Bull Hist Med. 1991;65:4–21. 7. Johnson NPAS, Mueller J. Updating the accounts: global mortality of the 1918–1920 “Spanish” influenza pandemic. Bull Hist Med 2002;76:105–15. 8. Shope RE. The incidence of neutralizing antibodies for swine influenza virus in the sera of human beings of different ages. J Exp Med. 1936;63:669–84. 9. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to “Scandinavian” strains isolated in epidemics of 1950–1951. Virology. 1978;89:632–6. 10. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 “Spanish” influenza virus. Science. 1997;275:1793–6. 11. Basler CF, Reid AH, Dybing JK, Janczewski TA, Fanning TG, Zheng H, et al. Sequence of the 1918 pandemic influenza virus nonstructural gene (NS) segment and characterization of recombinant viruses bearing the 1918 NS genes. Proc Natl Acad Sci U S A 2001;98:2746–51. 12. Reid AH, Fanning TG, Hultin JV, Taubenberger JK. Origin and evolution of the 1918 “Spanish” influenza virus hemagglutinin gene. Proc Natl Acad Sci U S A 1999;96:1651–6. 13. Reid AH, Fanning TG, Janczewski TA, Lourens RM, and Taubenberger JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein gene segment. J Virol. 2004;78:12462–70. 14. Reid AH, Fanning TG, Janczewski TA, McCall S, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus matrix gene segment. J Virol. 2002;76:10717–23. 15. Reid AH, Fanning TG, Janczewski TA, Taubenberger JK. Characterization of the 1918 “Spanish” influenza virus neuraminidase gene. Proc Natl Acad Sci U S A 2000;97:6785–90. 16. Reid AH, Janczewski TA, Lourens RM, Elliot AJ, Daniels RS, Berry CL, et al. 1918 influenza pandemic caused by highly conserved viruses with two receptor-binding variants. Emerg Infect Dis. 2003;9:1249–53. 17. Taubenberger JK, Reid AH, Lourens RM, Wang R, Jin G, Fanning TG. Characterization of the 1918 influenza virus polymerase genes. Nature. 2005;437:889–93. 18. Reid AH, Taubenberger JK. The 1918 flu and other influenza pandemics: “over there” and back again. Lab Invest. 1999;79:95–101. 19. Reid AH, Taubenberger JK, Fanning TG. Evidence of an absence: the genetic origins of the 1918 pandemic influenza virus. Nat Rev Microbiol. 2004;2:909–14. 20. Taubenberger JK, Reid AH, Fanning TG. The 1918 influenza virus: a killer comes into view. Virology. 2000;274:241–5. 21. Jordan E. Epidemic influenza: a survey. Chicago: American Medical Association, 1927. 22. Capps J, Moody A. The recent epidemic of grip. JAMA. 1916;67:1349–50. 33. Oxford JS, Sefton A, Jackson R, Innes W, Daniels RS, Johnson NP. World War I may have allowed the emergence of “Spanish” influenza. Lancet Infect Dis. 2002;2:111–4. 24. Fanning TG, Slemons RD, Reid AH, Janczewski TA, Dean J, Taubenberger JK. 1917 avian influenza virus sequences suggest that the 1918 pandemic virus did not acquire its hemagglutinin directly from birds. J Virol. 2002;76:7860–2. 25. Reid AH, Fanning TG, Slemons RD, Janczewski TA, Dean J, Taubenberger JK. Relationship of pre-1918 avian influenza HA and NP sequences to subsequent avian influenza strains. Avian Dis. 2003;47:921–5. 26. Bean W, Schell M, Katz J, Kawaoka Y, Naeve C, Gorman O, et al. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J Virol. 1992;66:1129–38. 27. Weis W, Brown JH, Cusack S, Paulson JC, Skehel JJ, Wiley DC. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature. 1988;333:426–31. 28. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6′-sialyl(N-acetyllactosamine). Virology. 1997;232: 345–50. 29. Matrosovich M, Gambaryan A, Teneberg S, Piskarev VE, Yamnikova SS, Lvov DK, et al. Avian influenza A viruses differ from human viruses by recognition of sialyloigosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology. 1997;233:224–34. 30. Glaser L, Stevens J, Zamarin D, Wilson IA, Garcia-Sastre A, Tumpey TM, et al. A single amino acid substitution in the 1918 influenza virus hemagglutinin changes the receptor binding specificity. J Virol. 2005;79:11533–6. 31. Kobasa D, Takada A, Shinya K, Hatta M, Halfmann P, Theriault S, et al. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature. 2004;431:703–7. 32. Kash JC, Basler CF, Garcia-Sastre A, Carter V, Billharz R, Swayne DE, et al. Global host immune response: pathogenesis and transcriptional profiling of type A influenza viruses expressing the hemagglutinin and neuraminidase genes from the 1918 pandemic virus. J Virol. 2004;78:9499–511. 33. Grove RD, Hetzel AM. Vital statistics rates in the United States: 1940–1960. Washington: US Government Printing Office, 1968. 34. Linder FE, Grove RD. Vital statistics rates in the United States: 1900–1940. Washington: US Government Printing Office, 1943. 35. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53–60. 36. Frost WH. The epidemiology of influenza. Public Health Rep. 1919;34:1823–61. 37. Collins SD. Age and sex incidence of influenza and pneumonia morbidity and mortality in the epidemic of 1928-1929 with comparative data for the epidemic of 1918–1919. Public Health Rep. 1931;46:1909–37. 38. Majde JA. Influenza: Learn from the past. ASM News. 1996;62:514. 39. Peiris JS, Yu WC, Leung CW, Cheung CY, Ng WF, Nicholls JM, et al. Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet. 2004;363:617–9. Address for correspondence: Jeffery K. Taubenberger, Department of Molecular Pathology, Armed Forces Institute of Pathology, 1413 Research Blvd, Bldg 101, Rm 1057, Rockville, MD 20850-3125, USA; fax. 301-295-9507; email: taubenberger@afip.osd.mil The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Is the molecular basis of human adaptation of a virus understood?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How many severe cases of influenza-related illnesses are reported per year?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How many influenza-related deaths are reported each year?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What is the mortality rate of the H5N1 strain of influenza?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What cells are the main target of the influenza A virus in the lungs?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How many extracellular domains are in the CEAMCAM1 protein?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

Where is CEACAM1 expressed in the body?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What motifs are absent in the short form of CEACAM1 protein?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What are the most common isoforms of CEACAM1?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How do CEACAM1 and CEACAM5 interact?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What are the SRC-family of kinases?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What triggers the release of pro-inflammatory cytokines/chemokines to assist in viral clearance?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

What mediates the anti-apoptosis of neutrophils?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How do natural killer cells fight influenza viruses?

Deep sequencing of primary human lung epithelial cells challenged with H5N1 influenza virus reveals a proviral role for CEACAM1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6195505/ SHA: ef58c6e981a08c85d2c0efb80e5b32b075f660b4 Authors: Ye, Siying; Cowled, Christopher J.; Yap, Cheng-Hon; Stambas, John Date: 2018-10-19 DOI: 10.1038/s41598-018-33605-6 License: cc-by Abstract: Current prophylactic and therapeutic strategies targeting human influenza viruses include vaccines and antivirals. Given variable rates of vaccine efficacy and antiviral resistance, alternative strategies are urgently required to improve disease outcomes. Here we describe the use of HiSeq deep sequencing to analyze host gene expression in primary human alveolar epithelial type II cells infected with highly pathogenic avian influenza H5N1 virus. At 24 hours post-infection, 623 host genes were significantly upregulated, including the cell adhesion molecule CEACAM1. H5N1 virus infection stimulated significantly higher CEACAM1 protein expression when compared to influenza A PR8 (H1N1) virus, suggesting a key role for CEACAM1 in influenza virus pathogenicity. Furthermore, silencing of endogenous CEACAM1 resulted in reduced levels of proinflammatory cytokine/chemokine production, as well as reduced levels of virus replication following H5N1 infection. Our study provides evidence for the involvement of CEACAM1 in a clinically relevant model of H5N1 infection and may assist in the development of host-oriented antiviral strategies. Text: Influenza viruses cause acute and highly contagious seasonal respiratory disease in all age groups. Between 3-5 million cases of severe influenza-related illness and over 250 000 deaths are reported every year. In addition to constant seasonal outbreaks, highly pathogenic avian influenza (HPAI) strains, such as H5N1, remain an ongoing pandemic threat with recent WHO figures showing 454 confirmed laboratory infections and a mortality rate of 53%. It is important to note that humans have very little pre-existing immunity towards avian influenza virus strains. Moreover, there is no commercially available human H5N1 vaccine. Given the potential for H5N1 viruses to trigger a pandemic 1,2 , there is an urgent need to develop novel therapeutic interventions to combat known deficiencies in our ability to control outbreaks. Current seasonal influenza virus prophylactic and therapeutic strategies involve the use of vaccination and antivirals. Vaccine efficacy is highly variable as evidenced by a particularly severe 2017/18 epidemic, and frequent re-formulation of the vaccine is required to combat ongoing mutations in the influenza virus genome. In addition, antiviral resistance has been reported for many circulating strains, including the avian influenza H7N9 virus that emerged in 2013 3, 4 . Influenza A viruses have also been shown to target and hijack multiple host cellular pathways to promote survival and replication 5, 6 . As such, there is increasing evidence to suggest that targeting host pathways will influence virus replication, inflammation, immunity and pathology 5, 7 . Alternative intervention strategies based on modulation of the host response could be used to supplement the current prophylactic and therapeutic protocols. While the impact of influenza virus infection has been relatively well studied in animal models 8, 9 , human cellular responses are poorly defined due to the lack of available human autopsy material, especially from HPAI virus-infected patients. In the present study, we characterized influenza virus infection of primary human alveolar epithelial type II (ATII) cells isolated from normal human lung tissue donated by patients undergoing lung resection. ATII cells are a physiologically relevant infection model as they are a main target for influenza A viruses when entering the respiratory tract 10 . Human host gene expression following HPAI H5N1 virus (A/Chicken/ Vietnam/0008/04) infection of primary ATII cells was analyzed using Illumina HiSeq deep sequencing. In order to gain a better understanding of the mechanisms underlying modulation of host immunity in an anti-inflammatory environment, we also analyzed changes in gene expression following HPAI H5N1 infection in the presence of the reactive oxygen species (ROS) inhibitor, apocynin, a compound known to interfere with NADPH oxidase subunit assembly 5, 6 . The HiSeq analysis described herein has focused on differentially regulated genes following H5N1 infection. Several criteria were considered when choosing a "hit" for further study. These included: (1) Novelty; has this gene been studied before in the context of influenza virus infection/pathogenesis? (2) Immunoregulation; does this gene have a regulatory role in host immune responses so that it has the potential to be manipulated to improve immunity? (3) Therapeutic reagents; are there any existing commercially available therapeutic reagents, such as specific inhibitors or inhibitory antibodies that can be utilized for in vitro and in vivo study in order to optimize therapeutic strategies? (4) Animal models; is there a knock-out mouse model available for in vivo influenza infection studies? Based on these criteria, carcinoembryonic-antigen (CEA)-related cell adhesion molecule 1 (CEACAM1) was chosen as a key gene of interest. CEACAM1 (also known as BGP or CD66) is expressed on epithelial and endothelial cells 11 , as well as B cells, T cells, neutrophils, NK cells, macrophages and dendritic cells (DCs) [12] [13] [14] . Human CEACAM1 has been shown to act as a receptor for several human bacterial and fungal pathogens, including Haemophilus influenza, Escherichia coli, Salmonella typhi and Candida albicans, but has not as yet been implicated in virus entry [15] [16] [17] . There is however emerging evidence to suggest that CEACAM1 is involved in host immunity as enhanced expression in lymphocytes was detected in pregnant women infected with cytomegalovirus 18 and in cervical tissue isolated from patients with papillomavirus infection 19 . Eleven CEACAM1 splice variants have been reported in humans 20 . CEACAM1 isoforms (Uniprot P13688-1 to -11) can differ in the number of immunoglobulin-like domains present, in the presence or absence of a transmembrane domain and/or the length of their cytoplasmic tail (i.e. L, long or S, short). The full-length human CEACAM1 protein (CEACAM1-4L) consists of four extracellular domains (one extracellular immunoglobulin variable-region-like (IgV-like) domain and three immunoglobulin constant region 2-like (IgC2-like) domains), a transmembrane domain, and a long (L) cytoplasmic tail. The long cytoplasmic tail contains two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are absent in the short form 20 . The most common isoforms expressed by human immune cells are CEACAM1-4L and CEACAM1-3L 21 . CEACAM1 interacts homophilically with itself 22 or heterophilically with CEACAM5 (a related CEACAM family member) 23 . The dimeric state allows recruitment of signaling molecules such as SRC-family kinases, including the tyrosine phosphatase SRC homology 2 (SH2)-domain containing protein tyrosine phosphatase 1 (SHP1) and SHP2 members to phosphorylate ITIMs 24 . As such, the presence or absence of ITIMs in CEACAM1 isoforms influences signaling properties and downstream cellular function. CEACAM1 homophilic or heterophilic interactions and ITIM phosphorylation are critical for many biological processes, including regulation of lymphocyte function, immunosurveillance, cell growth and differentiation 25, 26 and neutrophil activation and adhesion to target cells during inflammatory responses 27 . It should be noted that CEACAM1 expression has been modulated in vivo using an anti-CEACAM1 antibody (MRG1) to inhibit CEACAM1-positive melanoma xenograft growth in SCID/NOD mice 28 . MRG1 blocked CEACAM1 homophilic interactions that inhibit T cell effector function, enhancing the killing of CEACAM1+ melanoma cells by T cells 28 . This highlights a potential intervention pathway that can be exploited in other disease processes, including virus infection. In addition, Ceacam1-knockout mice are available for further in vivo infection studies. Our results show that CEACAM1 mRNA and protein expression levels were highly elevated following HPAI H5N1 infection. Furthermore, small interfering RNA (siRNA)-mediated inhibition of CEACAM1 reduced inflammatory cytokine and chemokine production, and more importantly, inhibited H5N1 virus replication in primary human ATII cells and in the continuous human type II respiratory epithelial A549 cell line. Taken together, these observations suggest that CEACAM1 is an attractive candidate for modulating influenza-specific immunity. In summary, our study has identified a novel target that may influence HPAI H5N1 immunity and serves to highlight the importance of manipulating host responses as a way of improving disease outcomes in the context of virus infection. Three experimental groups were included in the HiSeq analysis of H5N1 infection in the presence or absence of the ROS inhibitor, apocynin: (i) uninfected cells treated with 1% DMSO (vehicle control) (ND), (ii) H5N1-infected cells treated with 1% DMSO (HD) and (iii) H5N1-infected cells treated with 1 mM apocynin dissolved in DMSO (HA). These three groups were assessed using pairwise comparisons: ND vs. HD, ND vs. HA, and HD vs. HA. H5N1 infection and apocynin treatment induce differential expression of host genes. ATII cells isolated from human patients 29, 30 were infected with H5N1 on the apical side at a multiplicity of infection (MOI) of 2 for 24 hours and RNA extracted. HiSeq was performed on samples and reads mapped to the human genome where they were then assembled into transcriptomes for differential expression analysis. A total of 13,649 genes were identified with FPKM (fragments per kilobase of exon per million fragments mapped) > 1 in at least one of the three experimental groups. A total of 623 genes were significantly upregulated and 239 genes were significantly downregulated (q value < 0.05, ≥2-fold change) following H5N1 infection (ND vs. HD) ( Fig. 1A ; Table S1 ). HPAI H5N1 infection of ATII cells activated an antiviral state as evidenced by the upregulation of numerous interferon-induced genes, genes associated with pathogen defense, cell proliferation, apoptosis, and metabolism (Table 1; Table S2 ). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping showed that many of the upregulated genes in the HD group were mapped to TNF signaling (hsa04668), Toll-like receptor signaling (hsa04620), cytokine-cytokine receptor interaction (hsa04060) and RIG-I-like receptor signaling (hsa04622) ( In the H5N1-infected and apocynin-treated (HA) group, a large number of genes were also significantly upregulated (509 genes) or downregulated (782 genes) ( Fig. 1B ; Table S1 ) relative to the ND control group. Whilst a subset of genes was differentially expressed in both the HD and HA groups, either being upregulated (247 genes, Fig. 1D ) or downregulated (146 genes, Fig. 1E ), a majority of genes did not in fact overlap between the HD and HA groups (Fig. 1D , E). This suggests that apocynin treatment can affect gene expression independent of H5N1 infection. Gene Ontology (GO) enrichment analysis of genes upregulated by apocynin showed the involvement of the type I interferon signaling pathway (GO:0060337), the defense response to virus (GO:0009615), negative regulation of viral processes (GO:48525) and the response to stress (GO:0006950) ( Table S2 , "ND vs. HA Up"). Genes downregulated by apocynin include those that are involved in cell adhesion (GO:0007155), regulation of cell migration (GO:0030334), regulation of cell proliferation (GO:0042127), signal transduction (GO:0007165) and oxidation-reduction processes (GO:0055114) ( Table S2 , "ND vs. HA Down"). A total of 623 genes were upregulated following H5N1 infection ("ND vs. HD Up", Fig. 1F ). By overlapping the two lists of genes from "ND vs. HD Up" and "HD vs. HA Down", 245 genes were shown to be downregulated in the presence of apocynin (Fig. 1F ). By overlapping three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Up", 55 genes out of the 245 genes (190 plus 55 genes) were present in all three lists (Fig. 1G) , indicating that these 55 genes were significantly inhibited by apocynin but to a level that was still significantly higher than that in uninfected cells. The 55 genes include those involved in influenza A immunity (hsa05164; DDX58, IFIH1, IFNB1, MYD88, PML, STAT2), Jak-STAT signaling (hsa04630; IFNB1, IL15RA, IL22RA1, STAT2), RIG-I-like receptor signaling (hsa04622; DDX58, IFIH1, IFNB1) and Antigen processing and presentation (hsa04612; TAP2, TAP1, HLA-DOB) (Tables S3 and S4) . Therefore, critical immune responses induced following H5N1 infection were not dampened following apocynin treatment. The remaining 190 of 245 genes were not present in the "ND vs. HA Up" list, suggesting that those genes were significantly inhibited by apocynin to a level that was similar to uninfected control cells (Fig. 1G ). The 190 genes include those involved in TNF signaling (hsa04668; CASP10, CCL2, CCL5, CFLAR, CXCL5, END1, IL6, TRAF1, VEGFC), cytokine-cytokine receptor interaction (hsa04060; VEGFC, IL6, CCL2, CXCL5, CXCL16, IL2RG, CD40, CCL5, CCL7, IL1A), NF-kappa B signaling pathway (hsa04064: TRAF1, CFLAR, CARD11, TNFSF13B, TICAM1, CD40) and PI3K-Akt signaling (hsa04151; CCND1, GNB4, IL2RG, IL6, ITGA2, JAK2, LAMA1, MYC, IPK3AP1, TLR2, VEGFC) (Tables S3 and S4 ). This is consistent with the role of apocynin in reducing inflammation 31 . By overlapping the three lists of genes from "ND vs. HD Up", "HD vs. HA Down" and "ND vs. HA Down", 11 genes were found in all three comparisons (Fig. 1H ). This suggests that these 11 genes are upregulated following H5N1 infection and are significantly reduced by apocynin treatment to a level lower than that observed in uninfected control cells (Fig. 1H ). Among these were inflammatory cytokines/chemokines genes, including CXCL5, IL1A, AXL (a member of the TAM receptor family of receptor tyrosine kinases) and TMEM173/STING (Stimulator of IFN Genes) (Table S4) . Our previous study demonstrated that H5N1 infection of A549 cells in the presence of apocynin enhanced expression of negative regulators of cytokine signaling (SOCS), SOCS1 and SOCS3 6 . This, in turn, resulted in a reduction of H5N1-stimulated cytokine and chemokine production (IL6, IFNB1, CXCL10 and CCL5 in A549 cells), which was not attributed to lower virus replication as virus titers were not affected by apocynin treatment 6 . We performed a qRT-PCR analysis on the same RNA samples submitted for HiSeq analysis to validate HiSeq results. IL6 ( Fig. 2A) , IFNB1 (Fig. 2B) , CXCL10 (Fig. 2C ), and CCL5 ( Fig. 2D ) gene expression was significantly elevated in ATII cells following infection and was reduced by the addition of apocynin (except for IFNB1). Consistent with previous findings in A549 cells 6 , H5N1 infection alone induced the expression of SOCS1 as shown by HiSeq and qRT-PCR analysis (Fig. 2E ). Apocynin treatment further increased SOCS1 mRNA expression (Fig. 2E ). Although HiSeq analysis did not detect a statistically significant increase of SOCS1 following apocynin treatment, the Log2 fold-changes in SOCS1 gene expression were similar between the HD and HA groups (4.8-fold vs 4.0-fold) (Fig. 2E ). HiSeq analysis of SOCS3 transcription showed significant increase following H5N1 infection and apocynin treatment (Fig. 2F ). qRT-PCR analysis showed that although SOCS3 mRNA was only slightly increased following H5N1 infection, it was further significantly upregulated in the presence Table 2 . Representatives of over-represented KEGG pathways with a maximum P-value of 0.05 and the number of genes contributing to each pathway that is significantly upregulated following H5N1 infection ("ND vs. HD Up"). The full list of KEGG pathways is presented in Table S3 . of apocynin (Fig. 2F) . Therefore, apocynin also contributes to the reduction of H5N1-stimulated cytokine and chemokine production in ATII cells. Apocynin, a compound that inhibits production of ROS, has been shown to influence influenza-specific responses in vitro 6 and in vivo 5 . Although virus titers are not affected by apocynin treatment in vitro 6 , some anti-viral activity is observed in vivo when mice have been infected with a low pathogenic A/HongKong/X31 H3N2 virus 6 . HiSeq analysis of HPAI H5N1 virus gene transcription showed that although there was a trend for increased influenza virus gene expression following apocynin treatment, only influenza non-structural (NS) gene expression was significantly increased (Fig. 2G) . The reduced cytokine and chemokine production in H5N1-infected ATII cells ( Fig. 2A-F) is unlikely to be associated with lower virus replication. GO enrichment analysis was performed on genes that were significantly upregulated following HPAI H5N1 infection in ATII cells in the presence or absence of apocynin to identify over-presented GO terms. Many of the H5N1-upregulated genes were broadly involved in defense response (GO:0006952), response to external biotic stimulus (GO:0043207), immune system processes (GO:0002376), cytokine-mediated signaling pathway (GO:0019221) and type I interferon signaling pathway (GO:0060337) ( Table 1; Table S2 ). In addition, many of the H5N1-upregulated genes mapped to metabolic pathways (hsa01100), cytokine-cytokine receptor interaction (hsa04060), Influenza A (hsa05164), TNF signaling (hsa04668) or Jak-STAT signaling (hsa04630) (Table S3) . However, not all the H5N1-upregulated genes in these pathways were inhibited by apocynin treatment as mentioned above ( Fig. 1F ; Table S3 ). . Fold-changes following qRT-PCR analysis were calculated using 2 −ΔΔCt method (right Y axis) normalized to β-actin and compared with the ND group. Data from HiSeq was calculated as Log2 fold-change (left Y axis) compared with the ND group. IFNB1 transcription was not detected in ND, therefore HiSeq IFNB1 data from HD and HA groups was expressed as FPKM. *p < 0.05 and **p < 0.01, ***p < 0.001 compared with ND; # p < 0.05, ## p < 0.01, compared with HD. (G) Hiseq analysis of H5N1 influenza virus gene expression profiles with or without apocynin treatment in primary human ATII cells. # p < 0.05, compared with HD. Upregulation of the cell adhesion molecule CEACAM1 in H5N1-infected ATII cells. The cell adhesion molecule CEACAM1 has been shown to be critical for the regulation of immune responses during infection, inflammation and cancer 20 . The CEACAM1 transcript was significantly upregulated following H5N1 infection (Fig. 3A) . In contrast, a related member of the CEACAM family, CEACAM5, was not affected by H5N1 infection (Fig. 3B) . It is also worth noting that more reads were obtained for CEACAM5 (>1000 FPKM) (Fig. 3B ) than CEACAM1 (~7 FPKM) (Fig. 3A) in uninfected ATII cells, which is consistent with their normal expression patterns in human lung tissue 32 . Therefore, although CEACAM1 forms heterodimers with CEACAM5 23 , the higher basal expression of CEACAM5 in ATII cells may explain why its expression was not enhanced by H5N1 infection. Endogenous CEACAM1 protein expression was also analyzed in uninfected or influenza virus-infected A549 (Fig. 3C ) and ATII cells (Fig. 3D ). CEACAM1 protein expression was slightly, but not significantly, increased in A549 cells infected with A/Puerto Rico/8/1934 H1N1 (PR8) virus for 24 or 48 hours when compared to uninfected cells (Fig. 3C ). No significant difference in CEACAM1 protein levels were observed at various MOIs (2, 5 or 10) or between the 24 and 48 hpi timepoints (Fig. 3C) . After examing CEACAM1 protein expression following infection with PR8 virus in A549 cells, CEACAM1 protein expression was then examined in primary human ATII cells infected with HPAI H5N1 and compared to PR8 virus infection (Fig. 3D) . ATII cells were infected with PR8 virus at a MOI of 2, a dose that induced upregulation of cytokines and influenza Matrix (M) gene analyzed by qRT-PCR (data not shown). Lower MOIs of 0.5, 1 and 2 of HPAI H5N1 were tested due to the strong cytopathogenic effect H5N1 causes at higher MOIs. Endogenous CEACAM1 protein levels were significantly and similarly elevated in H5N1-infected ATII cells at the three MOIs tested. CEACAM1 protein expression in ATII cells infected with H5N1 at MOIs of 0.5 were higher at 48 hpi than those observed at 24 hpi (Fig. 3D ). HPAI H5N1 virus infection at MOIs of 0.5, 1 and 2 stimulated higher endogenous levels of CEACAM1 protein expression when compared to PR8 virus infection at a MOI of 2 at the corresponding time point (a maximum ~9-fold increase induced by H5N1 at MOIs of 0.5 and 1 at 48 hpi when compared to PR8 at MOI of 2), suggesting a possible role for CEACAM1 in influenza virus pathogenicity (Fig. 3D ). In order to understand the role of CEACAM1 in influenza pathogenesis, A549 and ATII cells were transfected with siCEACAM1 to knockdown endogenous CEACAM1 protein expression. ATII and A549 cells were transfected with siCEACAM1 or siNeg negative control. The expression of four main CEACAM1 variants, CEACAM1-4L, -4S, -3L and -3S, and CEACAM1 protein were analyzed using SYBR Green qRT-PCR and Western blotting, respectively. SYBR Green qRT-PCR analysis showed that ATII cells transfected with 15 pmol of siCEACAM1 significantly reduced the expression of CEACAM1-4L and -4S when compared to siNeg control, while the expression of CEACAM1-3L and -3S was not altered (Fig. 4A ). CEACAM1 protein expression was reduced by approximately 50% in both ATII and A549 cells following siCEACAM1 transfection when compared with siNeg-transfected cells (Fig. 4B) . Increasing doses of siCEACAM1 (10, 15 and 20 pmol) did not further downregulate CEACAM1 protein expression in A549 cells (Fig. 4B ). As such, 15 pmol of siCEACAM1 was chosen for subsequent knockdown studies in both ATII and A549 cells. It is important to note that the anti-CEACAM1 antibody only detects L isoforms based on epitope information provided by Abcam. Therefore, observed reductions in CEACAM1 protein expression can be attributed mainly to the abolishment of CEACAM1-4L. The functional consequences of CEACAM1 knockdown were then examined in ATII and A549 cells following H5N1 infection. IL6, IFNB1, CXCL10, CCL5 and TNF production was analyzed in H5N1-infected ATII and A549 cells using qRT-PCR. ATII (Fig. 5A ) and A549 cells (Fig. 5B) transfected with siCEACAM1 showed significantly lower expression of IL6, CXCL10 and CCL5 when compared with siNeg-transfected cells. However, the expression of the anti-viral cytokine, IFNB1, was not affected in both cells types. In addition, TNF expression, which can be induced by type I IFNs 33 , was significantly lower in siCEACAM1-transfected A549 cells (Fig. 5B) , but was not affected in siCEACAM1-transfected ATII cells (Fig. 5A) . Hypercytokinemia or "cytokine storm" in H5N1 and H7N9 virus-infected patients is thought to contribute to inflammatory tissue damage 34, 35 . Downregulation of CEACAM1 in the context of severe viral infection may reduce inflammation caused by H5N1 infection without dampening the antiviral response. Furthermore, virus replication was significantly reduced by 5.2-fold in ATII (Figs. 5C) and 4.8-fold in A549 cells (Fig. 5D ) transfected with siCEACAM1 when compared with siNeg-transfected cells. Virus titers in siNeg-transfected control cells were not significantly different from those observed in mock-transfected control cells (Fig. 5C,D) . Influenza viruses utilize host cellular machinery to manipulate normal cell processes in order to promote replication and evade host immune responses. Studies in the field are increasingly focused on understanding and modifying key host factors in order to ameliorate disease. Examples include modulation of ROS to reduce inflammation 5 and inhibition of NFκB and mitogenic Raf/MEK/ERK kinase cascade activation to suppress viral replication 36, 37 . These host targeting strategies will offer an alternative to current interventions that are focused on targeting the virus. In the present study, we analyzed human host gene expression profiles following HPAI H5N1 infection and treatment with the antioxidant, apocynin. As expected, genes that were significantly upregulated following H5N1 infection were involved in biological processes, including cytokine signaling, immunity and apoptosis. In addition, H5N1-upregulated genes were also involved in regulation of protein phosphorylation, cellular metabolism and cell proliferation, which are thought to be exploited by viruses for replication 38 . Apocynin treatment had both anti-viral (Tables S2-S4) 5 and pro-viral impact (Fig. 2G) , which is not surprising as ROS are potent microbicidal agents, as well as important immune signaling molecules at different concentrations 39 . In our hands, apocynin treatment reduced H5N1-induced inflammation, but also impacted the cellular defense response, cytokine production and cytokine-mediated signaling. Importantly, critical antiviral responses were not compromised, i.e. expression of pattern recognition receptors (e.g. DDX58 (RIG-I), TLRs, IFIH1 (MDA5)) was not downregulated (Table S1 ). Given the significant interference of influenza viruses on host immunity, we focused our attention on key regulators of the immune response. Through HiSeq analysis, we identified the cell adhesion molecule CEACAM1 as a critical regulator of immunity. Knockdown of endogenous CEACAM1 inhibited H5N1 virus replication and reduced H5N1-stimulated inflammatory cytokine/chemokine production. H5N1 infection resulted in significant upregulation of a number of inflammatory cytokines/chemokines genes, including AXL and STING, which were significantly reduced by apocynin treatment to a level lower than that observed in uninfected cells (Table S4) . It has been previously demonstrated that anti-AXL antibody treatment of PR8-infected mice significantly reduced lung inflammation and virus titers 40 . STING has been shown to be important for promoting anti-viral responses, as STING-knockout THP-1 cells produce less type I IFN following influenza A virus infection 41 . Reduction of STING gene expression or other anti-viral factors (e.g. IFNB1, MX1, ISG15; Table S1 ) by apocynin, may in part, explain the slight increase in influenza gene transcription following apocynin treatment (Fig. 2G) . These results also suggest that apocynin treatment may reduce H5N1-induced inflammation and apoptosis. Indeed, the anti-inflammatory and anti-apoptotic effects of apocynin have been shown previously in a number of disease models, including diabetes mellitus 42 , myocardial infarction 43 , neuroinflammation 44 and influenza virus infection 6 . Recognition of intracellular viral RNA by pattern recognition receptors (PRRs) triggers the release of pro-inflammatory cytokines/chemokines that recruit innate immune cells, such as neutrophils and NK cells, to the site of infection to assist in viral clearance 45 . Neutrophils exert their cytotoxic function by first attaching to influenza-infected epithelial cells via adhesion molecules, such as CEACAM1 46 . Moreover, studies have indicated that influenza virus infection promotes neutrophil apoptosis 47 , delaying virus elimination 48 . Phosphorylation of CEACAM1 ITIM motifs and activation of caspase-3 is critical for mediating anti-apoptotic events and for promoting survival of neutrophils 27 . This suggests that CEACAM1-mediated anti-apoptotic events may be important for the resolution of influenza virus infection in vivo, which can be further investigated through infection studies with Ceacam1-knockout mice. NK cells play a critical role in innate defense against influenza viruses by recognizing and killing infected cells. Influenza viruses, however, employ several strategies to escape NK effector functions, including modification of influenza hemagglutinin (HA) glycosylation to avoid NK activating receptor binding 49 . Homo-or heterophilic CEACAM1 interactions have been shown to inhibit NK-killing 25, 26 , and are thought to contribute to tumor cell immune evasion 50 . Given these findings, one could suggest the possibility that upregulation of CEACAM1 (to inhibit NK activity) may be a novel and uncharacterized immune evasion strategy employed by influenza viruses. Our laboratory is now investigating the role of CEACAM1 in NK cell function. Small-molecule inhibitors of protein kinases or protein phosphatases (e.g. inhibitors for Src, JAK, SHP2) have been developed as therapies for cancer, inflammation, immune and metabolic diseases 51 . Modulation of CEACAM1 phosphorylation, dimerization and the downstream function with small-molecule inhibitors may assist in dissecting the contribution of CEACAM1 to NK cell activity. The molecular mechanism of CEACAM1 action following infection has also been explored in A549 cells using PR8 virus 52 . Vitenshtein et al. demonstrated that CEACAM1 was upregulated following recognition of viral RNA by RIG-I, and that this upregulation was interferon regulatory factor 3 (IRF3)-dependent. In addition, phosphorylation of CEACAM1 by SHP2 inhibited viral replication by reducing phosphorylation of mammalian target of rapamycin (mTOR) to suppress global cellular protein production. In the present study, we used a more physiologically relevant infection model, primary human ATII cells, to study the role of Further studies will be required to investigate/confirm the molecular mechanisms of CEACAM1 upregulation following influenza virus infection, especially in vivo. As upregulation of CEACAM1 has been observed in other virus infections, such as cytomegalovirus 18 and papillomavirus 19 , it will be important to determine whether a common mechanism of action can be attributed to CEACAM1 in order to determine its functional significance. If this can be established, CEACAM1 could be used as a target for the development of a pan-antiviral agent. In summary, molecules on the cell surface such as CEACAM1 are particularly attractive candidates for therapeutic development, as drugs do not need to cross the cell membrane in order to be effective. Targeting of host-encoded genes in combination with current antivirals and vaccines may be a way of reducing morbidity and mortality associated with influenza virus infection. Our study clearly demonstrates that increased CEACAM1 expression is observed in primary human ATII cells infected with HPAI H5N1 influenza virus. Importantly, knockdown of CEACAM1 expression resulted in a reduction in influenza virus replication and suggests targeting of this molecule may assist in improving disease outcomes. Isolation and culture of primary human ATII cells. Human non-tumor lung tissue samples were donated by anonymous patients undergoing lung resection at University Hospital, Geelong, Australia. The research protocols and human ethics were approved by the Human Ethics Committees of Deakin University, Barwon Health and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Informed consent was obtained from all tissue donors. All research was performed in accordance with the guidelines stated in the National Statement on Ethical Conduct in Human Research (2007) . The sampling of normal lung tissue was confirmed by the Victorian Cancer Biobank, Australia. Lung specimens were preserved in Hartmann's solution (Baxter) for 4-8 hours or O/N at 4 °C to maintain cellular integrity and viability before cells are isolated. Human alveolar epithelial type II (ATII) cells were isolated and cultured using a previously described method 30, 53 with minor modifications. Briefly, lung tissue with visible bronchi was removed and perfused with abundant PBS and submerged in 0.5% Trypsin-EDTA (Gibco) twice for 15 min at 37 °C. The partially digested tissue was sliced into sections and further digested in Hank's Balanced Salt Solution (HBSS) containing elastase (12.9 units/mL; Roche Diagnostics) and DNase I (0.5 mg/mL; Roche Diagnostics) for 60 min at 37 °C. Single cell suspensions were obtained by filtration through a 40 μm cell strainer and cells (including macrophages and fibroblasts) were allowed to attach to tissue-culture treated Petri dishes in a 1:1 mixture of DMEM/F12 medium (Gibco) and small airway growth medium (SAGM) medium (Lonza) containing 5% fetal calf serum (FCS) and 0.5 mg/mL DNase I for 2 hours at 37 °C. Non-adherent cells, including ATII cells, were collected and subjected to centrifugation at 300 g for 20 min on a discontinuous Percoll density gradient (1.089 and 1.040 g/mL). Purified ATII cells from the interface of two density gradients was collected, washed in HBSS, and re-suspended in SAGM medium supplemented with 1% charcoal-filtered FCS (Gibco) and 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). ATII cells were plated on polyester Transwell inserts (0.4 μm pore; Corning) coated with type IV human placenta collagen (0.05 mg/mL; Sigma) at 300,000 cells/cm 2 and cultured under liquid-covered conditions in a humidified incubator (5% CO 2 , 37 °C). Growth medium was changed every 48 hours. These culture conditions suppressed fibroblasts expansion within the freshly isolated ATII cells and encouraged ATII cells to form confluent monolayers with a typical large and somewhat square morphology 54 Cell culture and media. A549 carcinomic human alveolar basal epithelial type II-like cells and Madin-Darby canine kidney (MDCK) cells were provided by the tissue culture facility of Australian Animal Health Laboratory (AAHL), CSIRO. A549 and MDCK cells were maintained in Ham's F12K medium (GIBCO) and RPMI-1640 medium (Invitrogen), respectively, supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) and maintained at 37 °C, 5% CO 2 . Virus and viral infection. HPAI A/chicken/Vietnam/0008/2004 H5N1 (H5N1) was obtained from AAHL, CSIRO. Viral stocks of A/Puerto Rico/8/1934 H1N1 (PR8) were obtained from the University of Melbourne. Virus stocks were prepared using standard inoculation of 10-day-old embryonated eggs. A single stock of virus was prepared for use in all assays. All H5N1 experiments were performed within biosafety level 3 laboratories (BSL3) at AAHL, CSIRO. Cells were infected with influenza A viruses as previously described 6, 29 . Briefly, culture media was removed and cells were washed with warm PBS three times followed by inoculation with virus for 1 hour. Virus was then removed and cells were washed with warm PBS three times, and incubated in the appropriate fresh serum-free culture media containing 0.3% BSA at 37 °C. Uninfected and infected cells were processed identically. For HiSeq analysis, ATII cells from three donors were infected on the apical side with H5N1 at a MOI of 2 for 24 hours in serum-free SAGM medium supplemented with 0.3% bovine serum albumin (BSA) containing 1 mM apocynin dissolved in DMSO or 1% DMSO vehicle control. Uninfected ATII cells incubated in media containing 1% DMSO were used as a negative control. For other subsequent virus infection studies, ATII cells from a different set of three donors (different from those used in HiSeq analysis) or A549 cells from at least three different passages were infected with influenza A viruses at various MOIs as indicated in the text. For H5N1 studies following transfection with siRNA, the infectious dose was optimized to a MOI of 0.01, a dose at which significantly higher CEACAM1 protein expression was induced with minimal cell death at 24 hpi. For PR8 infection studies, a final concentration of 0.5 µg/mL L-1-Tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Worthington) was included in media post-inoculation to assist replication. Virus titers were determined using standard plaque assays in MDCK cells as previously described 55 . RNA extraction, quality control (QC) and HiSeq analysis. ATII cells from three donors were used for HiSeq analysis. Total RNA was extracted from cells using a RNeasy Mini kit (Qiagen). Influenza-infected cells were washed with PBS three times and cells lysed with RLT buffer supplemented with β-mercaptoethanol (10 μL/mL; Gibco). Cell lysates were homogenized with QIAshredder columns followed by on-column DNA digestion with the RNase-Free DNase Set (Qiagen), and RNA extracted according to manufacturer's instructions. Initial QC was conducted to ensure that the quantity and quality of RNA samples for HiSeq analysis met the following criteria; 1) RNA samples had OD260/280 ratios between 1.8 and 2.0 as measured with NanoDrop TM Spectrophotometer (Thermo Scientific); 2) Sample concentrations were at a minimum of 100 ng/μl; 3) RNA was analyzed by agarose gel electrophoresis. RNA integrity and quality were validated by the presence of sharp clear bands of 28S and 18S ribosomal RNA, with a 28S:18S ratio of 2:1, along with the absence of genomic DNA and degraded RNA. As part of the initial QC and as an indication of consistent H5N1 infection, parallel quantitative real-time reverse transcriptase PCR (qRT-PCR) using the same RNA samples used for HiSeq analysis was performed in duplicate as previously described 6 to measure mRNA expression of IL6, IFNB1, CXCL10, CCL5, TNF, SOCS1 and SOCS3, all of which are known to be upregulated following HPAI H5N1 infection of A549 cells 6 Sequencing analysis and annotation. After confirming checksums and assessing raw data quality of the FASTQ files with FASTQC, RNA-Seq reads were processed according to standard Tuxedo pipeline protocols 56 , using the annotated human genome (GRCh37, downloaded from Illumina iGenomes) as a reference. Briefly, raw reads for each sample were mapped to the human genome using TopHat2, sorted and converted to SAM format using Samtools and then assembled into transcriptomes using Cufflinks. Cuffmerge was used to combine transcript annotations from individual samples into a single reference transcriptome, and Cuffquant was used to obtain per-sample read counts. Cuffdiff was then used to conduct differential expression analysis. All programs were run using recommended parameters. It is important to note that the reference gtf file provided to cuffmerge was first edited using a custom python script to exclude lines containing features other than exon/cds, and contigs other than chromosomes 1-22, X, Y. GO term and KEGG enrichment. Official gene IDs for transcripts that were differentially modulated following HPAI H5N1 infection with or without apocynin treatment were compiled into six target lists from pairwise comparisons ("ND vs. HD Up", "ND vs. HD Down", "ND vs. HA Up", "ND vs. HA Down", "HD vs. HA Up", "HD vs. HA Down"). Statistically significant differentially expressed transcripts were defined as having ≥2-fold change with a Benjamini-Hochberg adjusted P value < 0.01. A background list of genes was compiled by retrieving all gene IDs identified from the present HiSeq analysis with FPKM > 1. Biological process GO enrichment was performed using Gorilla, comparing unranked background and target lists 57 . Redundant GO terms were removed using REVIGO 58 . Target lists were also subjected to KEGG pathway analysis using a basic KEGG pathway mapper 59 and DAVID Bioinformatics Resources Functional Annotation Tool 60,61 . Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR). mRNA concentrations of genes of interest were assessed and analyzed using qRT-PCR performed in duplicate as previously described 6 . Briefly, after total RNA extraction from influenza-infected cells, cDNA was SCIEntIfIC RepoRtS | (2018) 8:15468 | DOI:10.1038/s41598-018-33605-6 prepared using SuperScript ™ III First-Strand Synthesis SuperMix (Invitrogen). Gene expression of various cytokines was assessed using TaqMan Gene Expression Assays (Applied Biosystems) with commercial TaqMan primers and probes, with the exception of the influenza Matrix (M) gene (forward primer 5′-CTTCTAACCGAGGTCGAAACGTA-3′; reverse primer 5′-GGTGACAGGATTGGTCTTGTCTTTA-3′; probe 5′-FAM-TCAGGCCCCCTCAAAGCCGAG-NFQ-3′) 62 . Specific primers 63 (Table S5) were designed to estimate the expression of CEACAM1-4L, -4S, -3L and -3S in ATII and A549 cells using iTaq Universal SYBR Green Supermix (Bio-Rad) according to manufacturer's instruction. The absence of nonspecific amplification was confirmed by agarose gel electrophoresis of qRT-PCR products (15 μL) (data not shown). Gene expression was normalized to β-actin mRNA using the 2 −ΔΔCT method where expression levels were determined relative to uninfected cell controls. All assays were performed in duplicate using an Applied Biosystems ® StepOnePlus TM Real-Time PCR System. Western blot analysis. Protein expression of CEACAM1 was determined using Western blot analysis as previously described 6 . Protein concentrations in cell lysates were determined using EZQ ® Protein Quantitation Kit (Molecular Probes TM , Invitrogen). Equal amounts of protein were loaded on NuPAGE 4-12% Bis-Tris gels (Invitrogen), resolved by SDS/PAGE and transferred to PVDF membranes (Bio-Rad). Membranes were probed with rabbit anti-human CEACAM1 monoclonal antibody EPR4049 (ab108397, Abcam) followed by goat anti-rabbit HRP-conjugated secondary antibody (Invitrogen). Proteins were visualized by incubating membranes with Pierce enhanced chemiluminescence (ECL) Plus Western Blotting Substrate (Thermo Scientific) followed by detection on a Bio-Rad ChemiDoc ™ MP Imaging System or on Amersham ™ Hyperfilm ™ ECL (GE Healthcare). To use β-actin as a loading control, the same membrane was stripped in stripping buffer (1.5% (w/v) glycine, 0.1% (w/v) SDS, 1% (v/v) Tween-20, pH 2.2) and re-probed with a HRP-conjugated rabbit anti-β-actin monoclonal antibody (Cell Signaling). In some cases, two SDS/PAGE were performed simultaneously with equal amounts of protein loaded onto each gel for analysis of CEACAM1 and β-actin protein expression in each sample, respectively. Protein band density was quantified using Fiji software (version 1.49J10) 64 . CEACAM1 protein band density was normalized against that of β-actin and expressed as fold changes compared to controls. Knockdown of endogenous CEACAM1. ATII and A549 cells were grown to 80% confluency in 6-well plates then transfected with small interfering RNA (siRNA) targeting the human CEACAM1 gene (siCEACAM1; s1976, Silencer ® Select Pre-designed siRNA, Ambion ® ) or siRNA control (siNeg; Silencer ® Select Negative Control No. 1 siRNA, Ambion ® ) using Lipofetamine 3000 (ThermoFisher Scientific) according to manufacturer's instructions. Transfection and silencing efficiency were evaluated after 48 hours by Western blot analysis of CEACAM1 protein expression and by qRT-PCR analysis of CEACAM1 variants. In parallel experiments, virus replication and cytokine/chemokine production was analyzed in siCEACAM1-or siNeg-transfected cells infected with H5N1 virus (MOI = 0.01) at 24 hpi. Statistical analysis. Differences between two experimental groups were evaluated using a Student's unpaired, two-tailed t test. Fold-change differences of mRNA expression (qRT-PCR) between three experimental groups was evaluated using one-way analysis of variance (ANOVA) followed by a Bonferroni multiple-comparison test. Differences were considered significant with a p value of <0.05. The data are shown as means ± standard error of the mean (SEM) from three or four individual experiments. Statistical analyses were performed using GraphPad Prism for Windows (v5.02). All data generated or analyzed during this study are included in this published article or the supplementary information file. The raw and processed HiSeq data has been deposited to GEO (GSE119767; https://www.ncbi. nlm.nih.gov/geo/).

How do influenza viruses escape binding by the natural killer cell activating receptors?

Spread from the Sink to the Patient: In Situ Study Using Green Fluorescent Protein (GFP)-Expressing Escherichia coli To Model Bacterial Dispersion from Hand-Washing Sink-Trap Reservoirs https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5377511/ SHA: 615071c8c959f24857b1bad521cc432b59719bfb Authors: Kotay, Shireen; Chai, Weidong; Guilford, William; Barry, Katie; Mathers, Amy J. Date: 2017-03-31 DOI: 10.1128/aem.03327-16 License: cc-by Abstract: There have been an increasing number of reports implicating Gammaproteobacteria as often carrying genes of drug resistance from colonized sink traps to vulnerable hospitalized patients. However, the mechanism of transmission from the wastewater of the sink P-trap to patients remains poorly understood. Herein we report the use of a designated hand-washing sink lab gallery to model dispersion of green fluorescent protein (GFP)-expressing Escherichia coli from sink wastewater to the surrounding environment. We found no dispersion of GFP-expressing E. coli directly from the P-trap to the sink basin or surrounding countertop with coincident water flow from a faucet. However, when the GFP-expressing E. coli cells were allowed to mature in the P-trap under conditions similar to those in a hospital environment, a GFP-expressing E. coli-containing putative biofilm extended upward over 7 days to reach the strainer. This subsequently resulted in droplet dispersion to the surrounding areas (<30 in.) during faucet operation. We also demonstrated that P-trap colonization could occur by retrograde transmission along a common pipe. We postulate that the organisms mobilize up to the strainer from the P-trap, resulting in droplet dispersion rather than dispersion directly from the P-trap. This work helps to further define the mode of transmission of bacteria from a P-trap reservoir to a vulnerable hospitalized patient. IMPORTANCE Many recent reports demonstrate that sink drain pipes become colonized with highly consequential multidrug-resistant bacteria, which then results in hospital-acquired infections. However, the mechanism of dispersal of bacteria from the sink to patients has not been fully elucidated. Through establishment of a unique sink gallery, this work found that a staged mode of transmission involving biofilm growth from the lower pipe to the sink strainer and subsequent splatter to the bowl and surrounding area occurs rather than splatter directly from the water in the lower pipe. We have also demonstrated that bacterial transmission can occur via connections in wastewater plumbing to neighboring sinks. This work helps to more clearly define the mechanism and risk of transmission from a wastewater source to hospitalized patients in a world with increasingly antibiotic-resistant bacteria that can thrive in wastewater environments and cause infections in vulnerable patients. Text: D espite early reports (1) (2) (3) (4) (5) , the premise that hand-wash sink traps can act as reservoirs of bacteria that cause nosocomial infections has been frequently overlooked. There has recently been an alarming increase in sink-related outbreaks worldwide, with many reports establishing an observational link (6) (7) (8) (9) (10) (11) (12) (13) . A sink often operates as an open conduit to wastewater in a patient care area that is often in the same room as the patient. Health care establishments often invest in desperate interventions to deal with nosocomial outbreaks. The preferred method for addressing most of the environmentrelated transmission is to employ enhanced cleaning using chemical and physical agents (14, 15) . Unfortunately, routine approaches are inefficient in completely eliminating drug-resistant Gammaproteobacteria in an inaccessible microbiologically active area such as a sink trap (6, (16) (17) (18) (19) (20) . The wet, humid, and relatively protected environment in a sink trap favors the formation of rich stable microbial communities (16, 21, 22) . These communities will be exposed to liquids and waste that are discarded in a sink and may include antimicrobials, discarded beverages, soap, presumably pathogenic bacteria from health care workers' hands, and other items. In short, sink traps could serve as a breeding ground for opportunistic and highly antimicrobial-resistant bacteria that cannot be easily cleaned or removed (23) (24) (25) (26) (27) (28) . There are many reports of a genetic association between pathogens found in sink traps and those found in patients (29, 30) . However, surprisingly little work has been done to understand the microscale transmission dynamics. It was previously demonstrated using a suspension of fluorescent particles (Glo Germ; Glo Germ Co., Moab, UT) that material injected into the P-trap gets dispersed around a hand-washing sink (6) . This result however has not been replicated hitherto in the follow-up studies. Dispersion has never been investigated with living organisms. Ultimately, many details remain unaddressed surrounding the spread of Enterobacteriaceae in sink-trap wastewater systems: (i) can organisms grow retrograde from the P-trap water to the sink strainer, (ii) can organisms spread from one sink to another along the internal surfaces of pipes with shared drainage systems, and (iii) which portion of a colonized drain pipe results in dispersion into the sink bowl during a hand-washing event? We aim to better understand the dispersion dynamics of Gammaproteobacteria living in the wastewater of a sink strainer and P-trap into an area where patients and health care workers could be exposed. To study this dynamic, we used a surrogate organism that could be easily tracked while remaining in the Enterobacteriaceae family, where some of the most concerning threats in antimicrobial resistance are developing (30) . Growth and colonization of GFP-expressing E. coli in the P-trap. In the first 14 days following the installation of the P-trap with established green fluorescent protein (GFP)-expressing Escherichia coli and just water running from the faucet, GFPexpressing E. coli was not detected in the tailpipe beyond 1.5 in. above the liquid level in the P-trap. GFP-expressing E. coli, however, was found to be viable in the P-trap without any nutrients added. A nutrient regimen was then instituted to understand the influence of nutrients on mobility and upward growth. The addition of tryptic soy broth (TSB) promoted GFP-expressing E. coli growth as early as day 1, with growth observed in the tailpipe 2 in. above the liquid surface in the P-trap (Table 1) . On day 7, the strainer (ϳ8 in. above the liquid in the P-trap) was found to be colonized with GFP-expressing E. coli. This translates to an average growth rate of 1 in./day along the length of the tailpipe with the addition of nutrients and without faucet operation. GFP-expressing E. coli was not detected in the faucet water. Sink-to-sink transmission of bacteria. In these experiments, a flanking sink (sink 5) was the only P-trap inoculated with GFP-expressing E. coli and therefore was the sole source for transmission to the connected sinks. Starting with a lower inoculum concentration (10 3 CFU/ml) in sink 5, on day 7, GFP-expressing E. coli was detected in the sink 2 and sink 3 P-traps (Fig. 1a) . With inoculum concentrations of 10 6 CFU/ml and Ͼ10 10 CFU/ml in sink 5, all of the sink P-traps in the sink gallery with the exception of sink 1 were found to be colonized with GFP-expressing E. coli after 7 days (Fig. 1b and c) . Faucet water and aerators tested negative for GFP-expressing E. coli. Irrespective of the starting inoculum concentration, on day 7 the highest level of colonization was recorded in the sink 3 P-trap. After day 7, when the nutrient regimen (described previously) was followed for an additional 7 days in each of the sinks in the sink gallery with an inoculum concentration of Ͼ10 10 CFU/ml, GFP-expressing E. coli was detected in the strainers of sinks 2 and 3 on day 14. This finding validated the upward growth and growth rate in the tailpipe when nutrients were added. Nonfluorescent colonies were occasionally observed in the P-trap water samples collected from the sinks, which were subsequently identified to be Pseudomonas sp. or Stenotrophomonas maltophilia, and fluorescent colonies were confirmed to be E. coli. Dispersion of microspheres from sinks. In the first dispersion experiment, when fluorescent microspheres were inoculated into the offset drain tailpiece only 4 in. below the strainer, no microspheres were detected on the polyester sheets placed on the counter space. However, when the sink bowl was coated with the microspheres, polyester sheets overlaid on the counter space captured the dispersed microspheres caused by the faucet operation. Dispersion was observed on almost all zones of the sink counter space (Fig. 2) . Relatively higher levels of dispersion were observed along the major and minor axes of the elliptical sink bowl (zones 2, 5, 6, 9, 11, and 12) . Anterior corners of the sink counter space (zones 4 and 7), which were most distant from the impact of water in the sink bowl, received the lowest dispersion. Dispersion of GFP-expressing E. coli from sinks. Initially the P-trap alone was inoculated with GFP-expressing E. coli and carefully installed, keeping the tailpipe and strainer free of GFP-expressing E. coli before operating the faucets. No fluorescent CFU were observed on the plates placed on the counter or attached to the bowl surface after faucet operation. Similarly, no fluorescent CFU were detected when GFPexpressing E. coli was inoculated into the offset drain tailpiece only 4 in. below the strainer. Interestingly, when there was conspicuous water backup over the strainer as a result of a higher water flow rate from the faucet than the drainage rate from the P-trap, dispersal was detected on the plates attached to the bowl surface. The dispersion pattern recorded when the sink bowl was coated with GFPexpressing E. coli was comparable to the pattern recorded when the sink bowl was coated with fluorescent microspheres (Fig. 2) . Dispersion was significantly higher along the axes (zones 6, 9, 11, and 12) and lower at the corners of the sink counter space (zones 4, 7, and 10). In contrast, dispersion of GFP-expressing E. coli caused by the faucet operation was much more extensive when the strainer was allowed to be colonized with GFPexpressing E. coli prior to the dispersion experiment. In addition to the sink counter space, we measured dispersion to the sink bowl, faucet, faucet handles, splatter shields, and the extended counter surface. Dispersion of GFP-expressing E. coli was highest on the plates attached to the sink bowl (Fig. 3b) . Further, dispersion was greater along the Ϫ" and "ϩ" denote the absence and presence of GFP-expressing E. coli, respectively. minor axis of the sink bowl (Fig. 3b , zones B3, B4, and B10) than along the major axis of the sink bowl, associated with a shorter distance from the strike point of the faucet water to the bowl along this axis. The next highest CFU count from the dispersal was recorded on the counter area near the faucets (Fig. 3a , zones 12 and 11). A similar pattern of higher dispersion near the faucets and lower dispersion at the corners of the counter space (Fig. 3a , zones 4, 7, and 10) was also observed using microspheres. Dispersion was also recorded in other zones of the counter space, on the Plexiglas splatter shields, faucets, faucet handles, and extended surface (Fig. 3c ). There were no GFP-expressing E. coli CFU recorded on plates placed beyond 30 in. from the strainer, demarcating the range of dispersion under these experimental conditions. Table 2 gives a summary of the total distribution loads recorded using fluorescent microspheres and GFP-expressing E. coli across each experiment. The loads of dispersion on the sink counter were comparable when the sink bowl was coated with microspheres or GFP-expressing E. coli before the faucet operation. Although the dispersion load on the sink counter was lower when the sink strainer was colonized, it is interesting to note that the sink bowl received the highest dispersion. To mimic dispersion in a hospital setting, we first investigated whether GFPexpressing E. coli would establish consistent colonization in a sink trap as many other Gammaproteobacteria implicated in nosocomial outbreaks have done (6, 28) . Many recent reports demonstrate that P-traps become colonized with highly consequential Gammaproteobacteria, which then results in nosocomial transmission (29, 31, 32) . The retained water in a sink P-trap is present to provide a water barrier to prevent off-gassing of sewer smell, but it may inadvertently provide favorable conditions for pathogenic and opportunistic antibiotic-resistant microorganisms to survive and develop resilient biofilms (3, 33) . However, the mechanism of dispersal of the bacteria in the P-trap to patients or the surrounding health care area had not been fully elucidated. We began with the hypothesis that the bacteria originate from the P-trap via droplet creation when the water from the faucet hits the P-trap water, thus contaminating the sink bowl and the surrounding area. The finding supporting this theory had been previously reported using Glo Germ particles (6) . However, in the present study with careful attention to avoid strainer and tail piece contamination, the dispersal directly from the sink P-trap with either microspheres or GFP-expressing E. coli could not be reproduced as previously reported (6) . Rather this work demonstrates a different, more staged mode of transmission from a P-trap reservoir to the sink and surrounding environment. GFP-expressing E. coli in the P-trap alone was sustained for 14 days but did not grow or mobilize up the tailpipe to the strainer with just intermittent water exposure. However, when nutrients were subsequently added to the system, the organisms rapidly grew up the tailpipe to the strainer at approximately an inch per day. In a real-world setting, motility of bacteria inside the tailpipe is restricted to relatively sporadic and brief wetting events in which swimming is an opportunity to colonize new surfaces. It is assumed that once established, the biofilm promotes the upward growth of GFP-expressing E. coli in the tailpipe at an accelerated rate. The nutrient regimen that promoted colonization in our model reflects our and others' observations of items commonly disposed of in hospital sinks (intravenous fluids, feeding supplements, and leftover beverages) (5, 32) . Transmission of bacteria between sinks via a common pipe was a key finding in this study as this highlights the concept that premise plumbing may be a more continuous system with shared microbiology than a single isolated sink. The sink gallery used in this study provided a unique in situ advantage to investigate sink-to-sink transmission of bacteria through common drains. The two possible mechanisms for P-trap strainers becoming colonized are seeding of organisms from above and retrograde spread of organisms along common pipes in a hospital wastewater infrastructure. Here we demonstrate that it is possible for GFP-expressing E. coli to contaminate adjacent P-traps with just time and water given a standard U.S. code piping rise of 0.25 ft/ft. Sink-to-sink or retrograde transmission may explain the recurrence of pathogen colonization following intervention strategies like disinfection or replacement of plumbing (23) . Sink 3 was lowest on the slope in the drain line (see Fig. 4 ) with arguably the most opportunity for reflux and retrograde wetting. Sink 1, on the other hand, was farthest away from the source (sink 5), and its P-trap had the greatest incline in the drain line connecting the sinks, which could perhaps contribute to the reasons there was no GFP-expressing E. coli colonization detected in it after 7 days. There has been more investigation about microbiologic dynamics of infectious viral particles such as those of severe acute respiratory syndrome (SARS) and Ebola viruses through premise plumbing systems (34) (35) (36) . However, the microbiology, sustainability, and dynamics might be very different, although the backflow and inoculation issues could have some parallels when comparing viruses to bacteria. As Enterobacteriaceae can either multiply or remain viable for long periods of time in biofilms coating the interior of P-traps and the connected plumbing, it may not be sustainable to target any intervention limited to a single isolated sink as a source of a particular pathogen. Data from different dispersal experiments suggest that although P-traps can act as the source or the reservoir of pathogens, the physical presence of the organism in the sink bowl or colonization of the strainer is necessary for the dispersal to occur. Colonization of strainers or drains reported in earlier studies (7, 10, 13, 24, 37) was perhaps a result of ascending biofilm growth from the P-trap to the strainer or introduction through contaminated fluids. Many of the studies used swab samples, which likely sampled the strainer rather than P-trap water (17, 20) . Once the strainer was colonized, the water from the faucet resulted in GFP-expressing E. coli dispersion in the bowl and to the surrounding surfaces of up to 30 in. The range of dispersal (6) . Greater dispersal near the faucet may be attributed to the specific designs of the sink bowl and faucet in this study, which determine the contact angle of water impact. This is an important finding since many sinks in hospitals are similar in design, with faucet handles representing a high-touch surface for the sink users (38) . It can also be concluded from the dispersion experiments that secondary and successive dispersals would likely increase the degree and the scope of dispersion. There are several limitations to this work. First the use of similar sink bowls across these sinks only examines the dispersion pattern of this particular sink design. Similarly the sink-to-sink transmission may not be applicable to all wastewater plumbing systems as the fixtures on the pipe are very close together, unlike most layouts in health care settings. However, we speculate that transmission could occur on larger systems over greater time scales, especially if heavy nutrient and contamination loads were also included. GFP-expressing E. coli is a laboratory surrogate, and the putative biofilms established in the short time frame of our experiments are unlikely to be as complex or stable as biofilms developed in a hospital wastewater system over many years. However, to address the monomicrobial dominance of the GFP-expressing E. coli added to the system, we kept the system open, and other environmental organisms were able to cocolonize in an attempt to mimic the hospital system. Another limitation was the need to add nutrients to the drain to ensure rapid and robust colonization. We are not clear how widespread the practice of disposing of dextrose-containing intravenous fluids or leftover beverages in the hand-wash sinks is; however, we have observed this practice, and anecdotally it appears to be relatively common in the United States. We also did not completely characterize the droplet sizes, nor do we demonstrate air sampling to understand if the dispersion is only droplet or if there are also aerosols that contain GFP-expressing E. coli. This would require additional testing and is planned as future work. In summary, this work for the first time better models the mechanisms of spread of multidrug-resistant pathogens arising from the sink drain and infecting patients. Droplet dispersion from the P-trap does not happen directly. Rather it is a multistage process: dispersal originates from the strainer and/or the bowl after growth of the biofilm up from the microbial reservoir of the P-trap. We also demonstrate sink-to-sink transmission via a common sanitary pipe. This work could have implications for patient safety, infection control, and interventions as well as the design of future hospital plumbing systems to eliminate this mode of transmission to vulnerable hospitalized patients. Sink gallery design. A dedicated sink gallery was set up to simulate hospital hand-washing sinks. The gallery was comprised of five sink modules assembled next to each other (Fig. 4) . The five hand-wash sink stations were identical in bowl designs and dimensions and were modeled from the most common intensive care unit hand-washing sink type in the acute care hospital at the University of Virginia Medical Center. Partitions made of 24-in.-high Plexiglas sheet were installed between the sinks to prevent splatter and cross contamination. Each sink module was built with Corian integrated sink/countertops without an overflow and fitted with an 8-in. centerset 2-handle gooseneck faucet (Elkay, Oak Brook, IL). The drain line (Dearborn Brass-Oatey, Cleveland, OH). All of the fixtures were made of brass with chrome plating. Each of the sink P-traps was connected to a 3-in. common cast-iron pipe sloping into a T-joint leading into the building sanitary line located behind sink 3 (Fig. 4) . Inoculation, growth, and establishment of GFP-expressing E. coli in sink P-traps. For the GFP-expressing E. coli strain (ATCC 25922GFP), the green fluorescent protein (GFP) gene is contained on a plasmid that also contains an ampicillin resistance gene. A single isolated colony of GFP-expressing E. coli grown from a Ϫ80°C stock was inoculated into 5 ml tryptic soy broth (TSB) (Becton, Dickinson and Company, Sparks, MD) containing 100 g/ml ampicillin (ATCC medium 2855). The inoculum concentration and method varied for each experiment. For establishment of GFP-expressing E. coli in sink P-traps, new autoclaved P-traps were filled with 100 ml 0.1ϫ TSB and inoculated with ϳ10 3 CFU/ml GFPexpressing E. coli. Following inoculation, both ends of the P-traps were covered with perforated Parafilm (Bemis, Inc., Oshkosh, WI) and allowed to incubate at room temperature (22 Ϯ 2°C) for 14 days to facilitate adherent bacterial growth. The medium in the P-trap was decanted and replaced with fresh 0.1ϫ TSB every 48 h. An aliquot of decanted medium and a swab sample from the inner surface of the P-trap were plated on tryptic soy agar (Becton, Dickinson and Company, Sparks, MD) plates containing 100 g/ml ampicillin (TSA) to monitor the growth of GFP-expressing E. coli in the P-traps. TSA plates were incubated overnight at 37°C, and CFU fluorescing under UV light were enumerated. All preparatory culturing of GFP-expressing E. coli took place in a separate room from the sink gallery to avoid unintentional contamination. Installation of P-traps colonized with GFP-expressing E.coli. After the 14-day incubation, P-traps were fastened into the plumbing of the sinks (Fig. 5a) . The remainder of the drain line was either autoclaved (strainer, tailpipe, and trap arms) prior to installation or surface disinfected (sink bowl, countertop, and faucets) with Caviwipes-1 (Meterx Research, Romulus, MI), maintaining at least 1 min of contact time. After the P-trap was installed, a daily regimen was followed in which 25 ml of TSB followed by 25 ml of 0.9% NaCl solution (saline) were added in the ratio 1:3 via the strainer (Fig. 5b) to mimic the potential nutrient exposure in the hospital. Sampling and enumeration of GFP-expressing E. coli. To monitor the growth of GFP-expressing E. coli in the plumbing, sampling ports were drilled along the length of the tailpiece (between the P-trap and the strainer) and the trap arm (between the P-trap and the common line). These holes were fitted with size 00 silicone stoppers (Cole-Parmer, Vernon Hills, IL) (Fig. 5a) . Sterile cotton swabs (Covidien, Mansfield, MA) presoaked in saline were inserted through these sampling ports, and samples were collected by turning the swab in a circular motion on the inner surface (ϳ20 cm 2 ) of the tailpipes. Sample swabs were pulse-vortexed in 3 ml saline, and serial dilutions were plated on TSA. The strainer, faucet aerator, and bowl surface were sampled with presoaked swabs and processed as described earlier. Sink-to-sink transmission of bacteria. To investigate sink-to-sink transmission of bacteria, a distal sink (sink 5) (Fig. 4) was fitted with a P-trap inoculated with GFP-expressing E. coli. The effects of different inoculum concentrations of GFP-expressing E. coli-10 3 , 10 6 , and Ͼ10 10 CFU/ml (colonized for 14 days)-were investigated. Identification to the species level of fluorescent and nonfluorescent colonies identified from mixed pipe cultures was performed using a matrix-assisted laser desorption-ionization (MALDI)-time of flight (MALDI-TOF) mass spectrometer (Vitek-MS; bioMérieux, Durham, NC). The wastewater paths of sinks 1 to 4 were either autoclaved (strainer, tailpipe, P-traps, and trap arms) prior to installation or surface disinfected (sink bowl, countertop, and faucets) with Caviwipes-1 (Meterx Research, Romulus, MI). Faucets on each of the five sinks were turned on simultaneously for 1 min, supplying water at a flow rate of 8 liters/min, once every 24 h for 7 days. No additional feed to any of the sinks was added during this period of 7 days. On days 0 and 7, P-traps on each of the five sinks were unfastened, and swab samples from the P-trap were collected and processed as described earlier. Dispersion measured using fluorescent microspheres. Fluoresbrite YO carboxylate microspheres (Polysciences, Inc.) which had a 1-m diameter and maximum excitation and emission of 529 nm and 546 nm, respectively, were used as a tracer in the preliminary experiments to understand droplet dispersion from the hand-wash sinks. To test whether microspheres could be dispersed from below the sink strainer, a 1-ml suspension of microspheres (ϳ10 10 particles) was injected through a strainer attached to a Hert 4.5-in. offset drain tailpiece typically used for wheelchair-accessible sinks (American Standard, model 7723018.002) (Fig. 5c) . The vertical distance between the strainer and microsphere suspension injected into the tailpipe was ϳ4 in. Counter space around the sink bowl was thoroughly wiped with alcohol wipes (Covidien Webcol 6818; Kendall), and polyester sheets precut to appropriate shapes were placed on the counter to cover the entire sink counter and labeled according to position (Fig. 6a) . The faucet was turned on for 5 min at a water flow rate of 1.8 to 3.0 liters/min. Polyester sheets were harvested and immediately analyzed using a ChemiDoc MP system (Bio-Rad Laboratories, Inc.) with an exposure time of 5 s. Fluorescent microspheres were enumerated from the digital micrographs using the Image Lab Software (Bio-Rad Laboratories, Inc.). To test whether microspheres could be dispersed from the surface of the sink bowl, the sink bowl was evenly coated with a 20-ml microsphere suspension (ϳ10 10 particles/ml) using a disposable swab (Sage Products, Inc., Cary, IL), and the dispersion experiment was repeated following the protocol described above. To ascertain there was no nonspecific background fluorescence in the sink and/or the water from the faucet, a control using the same protocol but without the fluorescent microspheres was performed before each experiment. Dispersion measured using GFP-expressing E.coli. Dispersion using GFP-expressing E. coli was investigated in three experiments. To test whether live organisms in the P-trap could be dispersed by running water, ϳ10 10 CFU/ml GFP-expressing E. coli in saline was added to an autoclaved P-trap and fitted into the drain line that was preautoclaved (strainer, tailpipe, and trap arms). Similarly, to test whether live organisms could be dispersed from the tailpieces of wheelchair-accessible sinks, a suspension of ϳ10 10 CFU/ml GFP-expressing E. coli was added with a syringe through the strainer into the Hert 4.5-ft offset drain tailpiece (Fig. 5c ). Just as in the microsphere dispersion experiment, the vertical distance between the strainer and GFP-expressing E. coli suspension injected into the tailpipe was ϳ4 in. We next tested whether live organisms from the surface of the sink bowl could be dispersed by running water. The sink bowl surface was evenly coated with an approximately 20-ml suspension of 10 10 CFU/ml GFP-expressing E. coli. Finally, to mimic all of these conditions, a P-trap colonized with GFP-expressing E. coli (for 14 days) was installed, and a nutrient regimen (Fig. 5b) was followed for 14 days to intentionally promote the GFP-expressing E. coli colonization in the attached tailpipe and strainer. On day 15, the dispersion experiment was performed. Before each of the GFP-expressing E. coli dispersion experiments, the counter space was thoroughly disinfected with Caviwipes-1. TSA plates were then positioned on the sink counter surrounding the bowl and an extension platform (Fig. 6b) . Additional plates were attached to the sink bowl, faucets, Plexiglas partitions, and faucet handles using adhesive tape. TSA plates were also placed 3 m away from the sink as negative controls. The faucet was turned on for 5 min with a water flow rate of 1.8 to 3.0 liters/min. Lids of the TSA plates were removed only during faucet operation. Swab samples from the faucet aerators before and after operation were collected and plated on TSA. Prior to each dispersion experiment, 50 ml water from the faucet was also collected, and aliquots were plated to assess for the presence of GFP-expressing E. coli in source water and ensure cross contamination of GFP-expressing E. coli had not occurred. A control dispersion experiment was also performed using the same protocol prior to GFP-expressing E. coli inoculation in each case. Dispersion per defined area (CFU per square centimeter) was deduced by dividing the CFU counts in the TSA plate by the surface area of the TSA plate.

Which type of bacteria are implicated in carrying genes of drug resistance?

Spread from the Sink to the Patient: In Situ Study Using Green Fluorescent Protein (GFP)-Expressing Escherichia coli To Model Bacterial Dispersion from Hand-Washing Sink-Trap Reservoirs https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5377511/ SHA: 615071c8c959f24857b1bad521cc432b59719bfb Authors: Kotay, Shireen; Chai, Weidong; Guilford, William; Barry, Katie; Mathers, Amy J. Date: 2017-03-31 DOI: 10.1128/aem.03327-16 License: cc-by Abstract: There have been an increasing number of reports implicating Gammaproteobacteria as often carrying genes of drug resistance from colonized sink traps to vulnerable hospitalized patients. However, the mechanism of transmission from the wastewater of the sink P-trap to patients remains poorly understood. Herein we report the use of a designated hand-washing sink lab gallery to model dispersion of green fluorescent protein (GFP)-expressing Escherichia coli from sink wastewater to the surrounding environment. We found no dispersion of GFP-expressing E. coli directly from the P-trap to the sink basin or surrounding countertop with coincident water flow from a faucet. However, when the GFP-expressing E. coli cells were allowed to mature in the P-trap under conditions similar to those in a hospital environment, a GFP-expressing E. coli-containing putative biofilm extended upward over 7 days to reach the strainer. This subsequently resulted in droplet dispersion to the surrounding areas (<30 in.) during faucet operation. We also demonstrated that P-trap colonization could occur by retrograde transmission along a common pipe. We postulate that the organisms mobilize up to the strainer from the P-trap, resulting in droplet dispersion rather than dispersion directly from the P-trap. This work helps to further define the mode of transmission of bacteria from a P-trap reservoir to a vulnerable hospitalized patient. IMPORTANCE Many recent reports demonstrate that sink drain pipes become colonized with highly consequential multidrug-resistant bacteria, which then results in hospital-acquired infections. However, the mechanism of dispersal of bacteria from the sink to patients has not been fully elucidated. Through establishment of a unique sink gallery, this work found that a staged mode of transmission involving biofilm growth from the lower pipe to the sink strainer and subsequent splatter to the bowl and surrounding area occurs rather than splatter directly from the water in the lower pipe. We have also demonstrated that bacterial transmission can occur via connections in wastewater plumbing to neighboring sinks. This work helps to more clearly define the mechanism and risk of transmission from a wastewater source to hospitalized patients in a world with increasingly antibiotic-resistant bacteria that can thrive in wastewater environments and cause infections in vulnerable patients. Text: D espite early reports (1) (2) (3) (4) (5) , the premise that hand-wash sink traps can act as reservoirs of bacteria that cause nosocomial infections has been frequently overlooked. There has recently been an alarming increase in sink-related outbreaks worldwide, with many reports establishing an observational link (6) (7) (8) (9) (10) (11) (12) (13) . A sink often operates as an open conduit to wastewater in a patient care area that is often in the same room as the patient. Health care establishments often invest in desperate interventions to deal with nosocomial outbreaks. The preferred method for addressing most of the environmentrelated transmission is to employ enhanced cleaning using chemical and physical agents (14, 15) . Unfortunately, routine approaches are inefficient in completely eliminating drug-resistant Gammaproteobacteria in an inaccessible microbiologically active area such as a sink trap (6, (16) (17) (18) (19) (20) . The wet, humid, and relatively protected environment in a sink trap favors the formation of rich stable microbial communities (16, 21, 22) . These communities will be exposed to liquids and waste that are discarded in a sink and may include antimicrobials, discarded beverages, soap, presumably pathogenic bacteria from health care workers' hands, and other items. In short, sink traps could serve as a breeding ground for opportunistic and highly antimicrobial-resistant bacteria that cannot be easily cleaned or removed (23) (24) (25) (26) (27) (28) . There are many reports of a genetic association between pathogens found in sink traps and those found in patients (29, 30) . However, surprisingly little work has been done to understand the microscale transmission dynamics. It was previously demonstrated using a suspension of fluorescent particles (Glo Germ; Glo Germ Co., Moab, UT) that material injected into the P-trap gets dispersed around a hand-washing sink (6) . This result however has not been replicated hitherto in the follow-up studies. Dispersion has never been investigated with living organisms. Ultimately, many details remain unaddressed surrounding the spread of Enterobacteriaceae in sink-trap wastewater systems: (i) can organisms grow retrograde from the P-trap water to the sink strainer, (ii) can organisms spread from one sink to another along the internal surfaces of pipes with shared drainage systems, and (iii) which portion of a colonized drain pipe results in dispersion into the sink bowl during a hand-washing event? We aim to better understand the dispersion dynamics of Gammaproteobacteria living in the wastewater of a sink strainer and P-trap into an area where patients and health care workers could be exposed. To study this dynamic, we used a surrogate organism that could be easily tracked while remaining in the Enterobacteriaceae family, where some of the most concerning threats in antimicrobial resistance are developing (30) . Growth and colonization of GFP-expressing E. coli in the P-trap. In the first 14 days following the installation of the P-trap with established green fluorescent protein (GFP)-expressing Escherichia coli and just water running from the faucet, GFPexpressing E. coli was not detected in the tailpipe beyond 1.5 in. above the liquid level in the P-trap. GFP-expressing E. coli, however, was found to be viable in the P-trap without any nutrients added. A nutrient regimen was then instituted to understand the influence of nutrients on mobility and upward growth. The addition of tryptic soy broth (TSB) promoted GFP-expressing E. coli growth as early as day 1, with growth observed in the tailpipe 2 in. above the liquid surface in the P-trap (Table 1) . On day 7, the strainer (ϳ8 in. above the liquid in the P-trap) was found to be colonized with GFP-expressing E. coli. This translates to an average growth rate of 1 in./day along the length of the tailpipe with the addition of nutrients and without faucet operation. GFP-expressing E. coli was not detected in the faucet water. Sink-to-sink transmission of bacteria. In these experiments, a flanking sink (sink 5) was the only P-trap inoculated with GFP-expressing E. coli and therefore was the sole source for transmission to the connected sinks. Starting with a lower inoculum concentration (10 3 CFU/ml) in sink 5, on day 7, GFP-expressing E. coli was detected in the sink 2 and sink 3 P-traps (Fig. 1a) . With inoculum concentrations of 10 6 CFU/ml and Ͼ10 10 CFU/ml in sink 5, all of the sink P-traps in the sink gallery with the exception of sink 1 were found to be colonized with GFP-expressing E. coli after 7 days (Fig. 1b and c) . Faucet water and aerators tested negative for GFP-expressing E. coli. Irrespective of the starting inoculum concentration, on day 7 the highest level of colonization was recorded in the sink 3 P-trap. After day 7, when the nutrient regimen (described previously) was followed for an additional 7 days in each of the sinks in the sink gallery with an inoculum concentration of Ͼ10 10 CFU/ml, GFP-expressing E. coli was detected in the strainers of sinks 2 and 3 on day 14. This finding validated the upward growth and growth rate in the tailpipe when nutrients were added. Nonfluorescent colonies were occasionally observed in the P-trap water samples collected from the sinks, which were subsequently identified to be Pseudomonas sp. or Stenotrophomonas maltophilia, and fluorescent colonies were confirmed to be E. coli. Dispersion of microspheres from sinks. In the first dispersion experiment, when fluorescent microspheres were inoculated into the offset drain tailpiece only 4 in. below the strainer, no microspheres were detected on the polyester sheets placed on the counter space. However, when the sink bowl was coated with the microspheres, polyester sheets overlaid on the counter space captured the dispersed microspheres caused by the faucet operation. Dispersion was observed on almost all zones of the sink counter space (Fig. 2) . Relatively higher levels of dispersion were observed along the major and minor axes of the elliptical sink bowl (zones 2, 5, 6, 9, 11, and 12) . Anterior corners of the sink counter space (zones 4 and 7), which were most distant from the impact of water in the sink bowl, received the lowest dispersion. Dispersion of GFP-expressing E. coli from sinks. Initially the P-trap alone was inoculated with GFP-expressing E. coli and carefully installed, keeping the tailpipe and strainer free of GFP-expressing E. coli before operating the faucets. No fluorescent CFU were observed on the plates placed on the counter or attached to the bowl surface after faucet operation. Similarly, no fluorescent CFU were detected when GFPexpressing E. coli was inoculated into the offset drain tailpiece only 4 in. below the strainer. Interestingly, when there was conspicuous water backup over the strainer as a result of a higher water flow rate from the faucet than the drainage rate from the P-trap, dispersal was detected on the plates attached to the bowl surface. The dispersion pattern recorded when the sink bowl was coated with GFPexpressing E. coli was comparable to the pattern recorded when the sink bowl was coated with fluorescent microspheres (Fig. 2) . Dispersion was significantly higher along the axes (zones 6, 9, 11, and 12) and lower at the corners of the sink counter space (zones 4, 7, and 10). In contrast, dispersion of GFP-expressing E. coli caused by the faucet operation was much more extensive when the strainer was allowed to be colonized with GFPexpressing E. coli prior to the dispersion experiment. In addition to the sink counter space, we measured dispersion to the sink bowl, faucet, faucet handles, splatter shields, and the extended counter surface. Dispersion of GFP-expressing E. coli was highest on the plates attached to the sink bowl (Fig. 3b) . Further, dispersion was greater along the Ϫ" and "ϩ" denote the absence and presence of GFP-expressing E. coli, respectively. minor axis of the sink bowl (Fig. 3b , zones B3, B4, and B10) than along the major axis of the sink bowl, associated with a shorter distance from the strike point of the faucet water to the bowl along this axis. The next highest CFU count from the dispersal was recorded on the counter area near the faucets (Fig. 3a , zones 12 and 11). A similar pattern of higher dispersion near the faucets and lower dispersion at the corners of the counter space (Fig. 3a , zones 4, 7, and 10) was also observed using microspheres. Dispersion was also recorded in other zones of the counter space, on the Plexiglas splatter shields, faucets, faucet handles, and extended surface (Fig. 3c ). There were no GFP-expressing E. coli CFU recorded on plates placed beyond 30 in. from the strainer, demarcating the range of dispersion under these experimental conditions. Table 2 gives a summary of the total distribution loads recorded using fluorescent microspheres and GFP-expressing E. coli across each experiment. The loads of dispersion on the sink counter were comparable when the sink bowl was coated with microspheres or GFP-expressing E. coli before the faucet operation. Although the dispersion load on the sink counter was lower when the sink strainer was colonized, it is interesting to note that the sink bowl received the highest dispersion. To mimic dispersion in a hospital setting, we first investigated whether GFPexpressing E. coli would establish consistent colonization in a sink trap as many other Gammaproteobacteria implicated in nosocomial outbreaks have done (6, 28) . Many recent reports demonstrate that P-traps become colonized with highly consequential Gammaproteobacteria, which then results in nosocomial transmission (29, 31, 32) . The retained water in a sink P-trap is present to provide a water barrier to prevent off-gassing of sewer smell, but it may inadvertently provide favorable conditions for pathogenic and opportunistic antibiotic-resistant microorganisms to survive and develop resilient biofilms (3, 33) . However, the mechanism of dispersal of the bacteria in the P-trap to patients or the surrounding health care area had not been fully elucidated. We began with the hypothesis that the bacteria originate from the P-trap via droplet creation when the water from the faucet hits the P-trap water, thus contaminating the sink bowl and the surrounding area. The finding supporting this theory had been previously reported using Glo Germ particles (6) . However, in the present study with careful attention to avoid strainer and tail piece contamination, the dispersal directly from the sink P-trap with either microspheres or GFP-expressing E. coli could not be reproduced as previously reported (6) . Rather this work demonstrates a different, more staged mode of transmission from a P-trap reservoir to the sink and surrounding environment. GFP-expressing E. coli in the P-trap alone was sustained for 14 days but did not grow or mobilize up the tailpipe to the strainer with just intermittent water exposure. However, when nutrients were subsequently added to the system, the organisms rapidly grew up the tailpipe to the strainer at approximately an inch per day. In a real-world setting, motility of bacteria inside the tailpipe is restricted to relatively sporadic and brief wetting events in which swimming is an opportunity to colonize new surfaces. It is assumed that once established, the biofilm promotes the upward growth of GFP-expressing E. coli in the tailpipe at an accelerated rate. The nutrient regimen that promoted colonization in our model reflects our and others' observations of items commonly disposed of in hospital sinks (intravenous fluids, feeding supplements, and leftover beverages) (5, 32) . Transmission of bacteria between sinks via a common pipe was a key finding in this study as this highlights the concept that premise plumbing may be a more continuous system with shared microbiology than a single isolated sink. The sink gallery used in this study provided a unique in situ advantage to investigate sink-to-sink transmission of bacteria through common drains. The two possible mechanisms for P-trap strainers becoming colonized are seeding of organisms from above and retrograde spread of organisms along common pipes in a hospital wastewater infrastructure. Here we demonstrate that it is possible for GFP-expressing E. coli to contaminate adjacent P-traps with just time and water given a standard U.S. code piping rise of 0.25 ft/ft. Sink-to-sink or retrograde transmission may explain the recurrence of pathogen colonization following intervention strategies like disinfection or replacement of plumbing (23) . Sink 3 was lowest on the slope in the drain line (see Fig. 4 ) with arguably the most opportunity for reflux and retrograde wetting. Sink 1, on the other hand, was farthest away from the source (sink 5), and its P-trap had the greatest incline in the drain line connecting the sinks, which could perhaps contribute to the reasons there was no GFP-expressing E. coli colonization detected in it after 7 days. There has been more investigation about microbiologic dynamics of infectious viral particles such as those of severe acute respiratory syndrome (SARS) and Ebola viruses through premise plumbing systems (34) (35) (36) . However, the microbiology, sustainability, and dynamics might be very different, although the backflow and inoculation issues could have some parallels when comparing viruses to bacteria. As Enterobacteriaceae can either multiply or remain viable for long periods of time in biofilms coating the interior of P-traps and the connected plumbing, it may not be sustainable to target any intervention limited to a single isolated sink as a source of a particular pathogen. Data from different dispersal experiments suggest that although P-traps can act as the source or the reservoir of pathogens, the physical presence of the organism in the sink bowl or colonization of the strainer is necessary for the dispersal to occur. Colonization of strainers or drains reported in earlier studies (7, 10, 13, 24, 37) was perhaps a result of ascending biofilm growth from the P-trap to the strainer or introduction through contaminated fluids. Many of the studies used swab samples, which likely sampled the strainer rather than P-trap water (17, 20) . Once the strainer was colonized, the water from the faucet resulted in GFP-expressing E. coli dispersion in the bowl and to the surrounding surfaces of up to 30 in. The range of dispersal (6) . Greater dispersal near the faucet may be attributed to the specific designs of the sink bowl and faucet in this study, which determine the contact angle of water impact. This is an important finding since many sinks in hospitals are similar in design, with faucet handles representing a high-touch surface for the sink users (38) . It can also be concluded from the dispersion experiments that secondary and successive dispersals would likely increase the degree and the scope of dispersion. There are several limitations to this work. First the use of similar sink bowls across these sinks only examines the dispersion pattern of this particular sink design. Similarly the sink-to-sink transmission may not be applicable to all wastewater plumbing systems as the fixtures on the pipe are very close together, unlike most layouts in health care settings. However, we speculate that transmission could occur on larger systems over greater time scales, especially if heavy nutrient and contamination loads were also included. GFP-expressing E. coli is a laboratory surrogate, and the putative biofilms established in the short time frame of our experiments are unlikely to be as complex or stable as biofilms developed in a hospital wastewater system over many years. However, to address the monomicrobial dominance of the GFP-expressing E. coli added to the system, we kept the system open, and other environmental organisms were able to cocolonize in an attempt to mimic the hospital system. Another limitation was the need to add nutrients to the drain to ensure rapid and robust colonization. We are not clear how widespread the practice of disposing of dextrose-containing intravenous fluids or leftover beverages in the hand-wash sinks is; however, we have observed this practice, and anecdotally it appears to be relatively common in the United States. We also did not completely characterize the droplet sizes, nor do we demonstrate air sampling to understand if the dispersion is only droplet or if there are also aerosols that contain GFP-expressing E. coli. This would require additional testing and is planned as future work. In summary, this work for the first time better models the mechanisms of spread of multidrug-resistant pathogens arising from the sink drain and infecting patients. Droplet dispersion from the P-trap does not happen directly. Rather it is a multistage process: dispersal originates from the strainer and/or the bowl after growth of the biofilm up from the microbial reservoir of the P-trap. We also demonstrate sink-to-sink transmission via a common sanitary pipe. This work could have implications for patient safety, infection control, and interventions as well as the design of future hospital plumbing systems to eliminate this mode of transmission to vulnerable hospitalized patients. Sink gallery design. A dedicated sink gallery was set up to simulate hospital hand-washing sinks. The gallery was comprised of five sink modules assembled next to each other (Fig. 4) . The five hand-wash sink stations were identical in bowl designs and dimensions and were modeled from the most common intensive care unit hand-washing sink type in the acute care hospital at the University of Virginia Medical Center. Partitions made of 24-in.-high Plexiglas sheet were installed between the sinks to prevent splatter and cross contamination. Each sink module was built with Corian integrated sink/countertops without an overflow and fitted with an 8-in. centerset 2-handle gooseneck faucet (Elkay, Oak Brook, IL). The drain line (Dearborn Brass-Oatey, Cleveland, OH). All of the fixtures were made of brass with chrome plating. Each of the sink P-traps was connected to a 3-in. common cast-iron pipe sloping into a T-joint leading into the building sanitary line located behind sink 3 (Fig. 4) . Inoculation, growth, and establishment of GFP-expressing E. coli in sink P-traps. For the GFP-expressing E. coli strain (ATCC 25922GFP), the green fluorescent protein (GFP) gene is contained on a plasmid that also contains an ampicillin resistance gene. A single isolated colony of GFP-expressing E. coli grown from a Ϫ80°C stock was inoculated into 5 ml tryptic soy broth (TSB) (Becton, Dickinson and Company, Sparks, MD) containing 100 g/ml ampicillin (ATCC medium 2855). The inoculum concentration and method varied for each experiment. For establishment of GFP-expressing E. coli in sink P-traps, new autoclaved P-traps were filled with 100 ml 0.1ϫ TSB and inoculated with ϳ10 3 CFU/ml GFPexpressing E. coli. Following inoculation, both ends of the P-traps were covered with perforated Parafilm (Bemis, Inc., Oshkosh, WI) and allowed to incubate at room temperature (22 Ϯ 2°C) for 14 days to facilitate adherent bacterial growth. The medium in the P-trap was decanted and replaced with fresh 0.1ϫ TSB every 48 h. An aliquot of decanted medium and a swab sample from the inner surface of the P-trap were plated on tryptic soy agar (Becton, Dickinson and Company, Sparks, MD) plates containing 100 g/ml ampicillin (TSA) to monitor the growth of GFP-expressing E. coli in the P-traps. TSA plates were incubated overnight at 37°C, and CFU fluorescing under UV light were enumerated. All preparatory culturing of GFP-expressing E. coli took place in a separate room from the sink gallery to avoid unintentional contamination. Installation of P-traps colonized with GFP-expressing E.coli. After the 14-day incubation, P-traps were fastened into the plumbing of the sinks (Fig. 5a) . The remainder of the drain line was either autoclaved (strainer, tailpipe, and trap arms) prior to installation or surface disinfected (sink bowl, countertop, and faucets) with Caviwipes-1 (Meterx Research, Romulus, MI), maintaining at least 1 min of contact time. After the P-trap was installed, a daily regimen was followed in which 25 ml of TSB followed by 25 ml of 0.9% NaCl solution (saline) were added in the ratio 1:3 via the strainer (Fig. 5b) to mimic the potential nutrient exposure in the hospital. Sampling and enumeration of GFP-expressing E. coli. To monitor the growth of GFP-expressing E. coli in the plumbing, sampling ports were drilled along the length of the tailpiece (between the P-trap and the strainer) and the trap arm (between the P-trap and the common line). These holes were fitted with size 00 silicone stoppers (Cole-Parmer, Vernon Hills, IL) (Fig. 5a) . Sterile cotton swabs (Covidien, Mansfield, MA) presoaked in saline were inserted through these sampling ports, and samples were collected by turning the swab in a circular motion on the inner surface (ϳ20 cm 2 ) of the tailpipes. Sample swabs were pulse-vortexed in 3 ml saline, and serial dilutions were plated on TSA. The strainer, faucet aerator, and bowl surface were sampled with presoaked swabs and processed as described earlier. Sink-to-sink transmission of bacteria. To investigate sink-to-sink transmission of bacteria, a distal sink (sink 5) (Fig. 4) was fitted with a P-trap inoculated with GFP-expressing E. coli. The effects of different inoculum concentrations of GFP-expressing E. coli-10 3 , 10 6 , and Ͼ10 10 CFU/ml (colonized for 14 days)-were investigated. Identification to the species level of fluorescent and nonfluorescent colonies identified from mixed pipe cultures was performed using a matrix-assisted laser desorption-ionization (MALDI)-time of flight (MALDI-TOF) mass spectrometer (Vitek-MS; bioMérieux, Durham, NC). The wastewater paths of sinks 1 to 4 were either autoclaved (strainer, tailpipe, P-traps, and trap arms) prior to installation or surface disinfected (sink bowl, countertop, and faucets) with Caviwipes-1 (Meterx Research, Romulus, MI). Faucets on each of the five sinks were turned on simultaneously for 1 min, supplying water at a flow rate of 8 liters/min, once every 24 h for 7 days. No additional feed to any of the sinks was added during this period of 7 days. On days 0 and 7, P-traps on each of the five sinks were unfastened, and swab samples from the P-trap were collected and processed as described earlier. Dispersion measured using fluorescent microspheres. Fluoresbrite YO carboxylate microspheres (Polysciences, Inc.) which had a 1-m diameter and maximum excitation and emission of 529 nm and 546 nm, respectively, were used as a tracer in the preliminary experiments to understand droplet dispersion from the hand-wash sinks. To test whether microspheres could be dispersed from below the sink strainer, a 1-ml suspension of microspheres (ϳ10 10 particles) was injected through a strainer attached to a Hert 4.5-in. offset drain tailpiece typically used for wheelchair-accessible sinks (American Standard, model 7723018.002) (Fig. 5c) . The vertical distance between the strainer and microsphere suspension injected into the tailpipe was ϳ4 in. Counter space around the sink bowl was thoroughly wiped with alcohol wipes (Covidien Webcol 6818; Kendall), and polyester sheets precut to appropriate shapes were placed on the counter to cover the entire sink counter and labeled according to position (Fig. 6a) . The faucet was turned on for 5 min at a water flow rate of 1.8 to 3.0 liters/min. Polyester sheets were harvested and immediately analyzed using a ChemiDoc MP system (Bio-Rad Laboratories, Inc.) with an exposure time of 5 s. Fluorescent microspheres were enumerated from the digital micrographs using the Image Lab Software (Bio-Rad Laboratories, Inc.). To test whether microspheres could be dispersed from the surface of the sink bowl, the sink bowl was evenly coated with a 20-ml microsphere suspension (ϳ10 10 particles/ml) using a disposable swab (Sage Products, Inc., Cary, IL), and the dispersion experiment was repeated following the protocol described above. To ascertain there was no nonspecific background fluorescence in the sink and/or the water from the faucet, a control using the same protocol but without the fluorescent microspheres was performed before each experiment. Dispersion measured using GFP-expressing E.coli. Dispersion using GFP-expressing E. coli was investigated in three experiments. To test whether live organisms in the P-trap could be dispersed by running water, ϳ10 10 CFU/ml GFP-expressing E. coli in saline was added to an autoclaved P-trap and fitted into the drain line that was preautoclaved (strainer, tailpipe, and trap arms). Similarly, to test whether live organisms could be dispersed from the tailpieces of wheelchair-accessible sinks, a suspension of ϳ10 10 CFU/ml GFP-expressing E. coli was added with a syringe through the strainer into the Hert 4.5-ft offset drain tailpiece (Fig. 5c ). Just as in the microsphere dispersion experiment, the vertical distance between the strainer and GFP-expressing E. coli suspension injected into the tailpipe was ϳ4 in. We next tested whether live organisms from the surface of the sink bowl could be dispersed by running water. The sink bowl surface was evenly coated with an approximately 20-ml suspension of 10 10 CFU/ml GFP-expressing E. coli. Finally, to mimic all of these conditions, a P-trap colonized with GFP-expressing E. coli (for 14 days) was installed, and a nutrient regimen (Fig. 5b) was followed for 14 days to intentionally promote the GFP-expressing E. coli colonization in the attached tailpipe and strainer. On day 15, the dispersion experiment was performed. Before each of the GFP-expressing E. coli dispersion experiments, the counter space was thoroughly disinfected with Caviwipes-1. TSA plates were then positioned on the sink counter surrounding the bowl and an extension platform (Fig. 6b) . Additional plates were attached to the sink bowl, faucets, Plexiglas partitions, and faucet handles using adhesive tape. TSA plates were also placed 3 m away from the sink as negative controls. The faucet was turned on for 5 min with a water flow rate of 1.8 to 3.0 liters/min. Lids of the TSA plates were removed only during faucet operation. Swab samples from the faucet aerators before and after operation were collected and plated on TSA. Prior to each dispersion experiment, 50 ml water from the faucet was also collected, and aliquots were plated to assess for the presence of GFP-expressing E. coli in source water and ensure cross contamination of GFP-expressing E. coli had not occurred. A control dispersion experiment was also performed using the same protocol prior to GFP-expressing E. coli inoculation in each case. Dispersion per defined area (CFU per square centimeter) was deduced by dividing the CFU counts in the TSA plate by the surface area of the TSA plate.

What may be a likely cause of sink-to-sink spreading of pathogens in the hospital setting?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is disease resilience?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What family of virus does SARS reside in?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What family of virus does MERS reside in?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

When was SARS-CoV first identified?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

How many people did SARS-CoV infect?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What percentage of people infected with MERS-CoV died?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What percentage of people infected with SARS-CoV died?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What was the reservoir for SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What was the primary threatening clinical finding in patients infected with SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the relationship between SARS-CoV and acute lung injury (ALI)?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the relationship between SARS-CoV and acute respiratory distress syndrome (ARDS)?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is required for a person to survive a serious SARS-CoV infection?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

How does cell-entry differ between SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is a major difference in clinical progression between SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

How does transmission differ between SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

How do SARS-CoV and MERS-CoV evade the immune system sensing it's genome?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What role does initial viral titer play in the prognosis of SARS-CoV and MERS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the timeline of the type I interferon (IFN) response in SARS-CoV infection?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

How do SARS-CoV viral proteins interact with the immune response?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What was the role of corticosteroid use in hospitalized patients with SARS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the role of interferon's (IFNs) in the treatment of SARS-CoV?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What are some negative effects of decreasing immunopathology by immunomodulation?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the role of topoisomerase I in improving host resilience in viral lung infections?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the role of complement 5a (C5a) in increasing host resilience to viral lung infection?

Host resilience to emerging coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7079962/ SHA: f7cfc37ea164f16393d7f4f3f2b32214dea1ded4 Authors: Jamieson, Amanda M Date: 2016-07-01 DOI: 10.2217/fvl-2016-0060 License: cc-by Abstract: Recently, two coronaviruses, severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus, have emerged to cause unusually severe respiratory disease in humans. Currently, there is a lack of effective antiviral treatment options or vaccine available. Given the severity of these outbreaks, and the possibility of additional zoonotic coronaviruses emerging in the near future, the exploration of different treatment strategies is necessary. Disease resilience is the ability of a given host to tolerate an infection, and to return to a state of health. This review focuses on exploring various host resilience mechanisms that could be exploited for treatment of severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus and other respiratory viruses that cause acute lung injury and acute respiratory distress syndrome. Text: The 21st century was heralded with the emergence of two novel coronaviruses (CoV) that have unusually high pathogenicity and mortality [1] [2] [3] [4] [5] . Severe acute respiratory syndrome coronavirus (SARS-Cov) was first identified in 2003 [6] [7] [8] [9] . While there was initially great concern about SARS-CoV, once no new cases emerged, funding and research decreased. However, a decade later Middle East respiratory syndrome coronavirus (MERS-CoV), also known as HCoV-EMC, emerged initially in Saudi Arabia [3, 10] . SARS-CoV infected about 8000 people, and resulted in the deaths of approximately 10% of those infected [11] . While MERS-CoV is not as widespread as SARS-CoV, it appears to have an even higher mortality rate, with 35-50% of diagnosed infections resulting in death [3, [12] [13] . These deadly betacoronavirus viruses existed in animal reservoirs [4] [5] 9, [14] [15] . Recently, other CoVs have been detected in animal populations raising the possibility that we will see a repeat of these types of outbreaks in the near future [11, [16] [17] [18] [19] [20] . Both these zoonotic viruses cause a much more severe disease than what is typically seen for CoVs, making them a global health concern. Both SARS-CoV and MERS-CoV result in severe lung pathology. Many infected patients have acute lung injury (ALI), a condition that is diagnosed based on the presence of pulmonary edema and respiratory failure without a cardiac cause. In some patients there is a progression to the more severe form of ALI, acute respiratory distress syndrome (ARDS) [21] [22] [23] . In order to survive a given infection, a successful host must not only be able to clear the pathogen, but tolerate damage caused by the pathogen itself and also by the host's immune response [24] [25] [26] . We refer to resilience as the ability of a host to tolerate the effects of pathogens and the immune response to pathogens. A resilient host is able to return to a state of health after responding to an infection [24, [27] [28] . Most currently available treatment options for infectious diseases are antimicrobials, For reprint orders, please contact: reprints@futuremedicine.com REviEW Jamieson future science group and thus target the pathogen itself. Given the damage that pathogens can cause this focus on rapid pathogen clearance is understandable. However, an equally important medical intervention is to increase the ability of the host to tolerate the direct and indirect effects of the pathogen, and this is an area that is just beginning to be explored [29] . Damage to the lung epithelium by respiratory pathogens is a common cause of decreased resilience [30] [31] [32] . This review explores some of the probable host resilience pathways to viral infections, with a particular focus on the emerging coronaviruses. We will also examine factors that make some patients disease tolerant and other patients less tolerant to the viral infection. These factors can serve as a guide to new potential therapies for improved patient care. Both SARS-CoV and MERS-CoV are typified by a rapid progression to ARDS, however, there are some distinct differences in the infectivity and pathogenicity. The two viruses have different receptors leading to different cellular tropism, and SARS-CoV is more ubiquitous in the cell type and species it can infect. SARS-CoV uses the ACE2 receptor to gain entry to cells, while MERS-CoV uses the ectopeptidase DPP4 [33] [34] [35] [36] . Unlike SARS-CoV infection, which causes primarily a severe respiratory syndrome, MERS-CoV infection can also lead to kidney failure [37, 38] . SARS-CoV also spreads more rapidly between hosts, while MERS-CoV has been more easily contained, but it is unclear if this is due to the affected patient populations and regions [3] [4] 39 ]. Since MERS-CoV is a very recently discovered virus, [40, 41] more research has been done on SARS-CoV. However, given the similarities it is hoped that some of these findings can also be applied to MERS-CoV, and other potential emerging zoonotic coronaviruses. Both viral infections elicit a very strong inflammatory response, and are also able to circumvent the immune response. There appears to be several ways that these viruses evade and otherwise redirect the immune response [1, [42] [43] [44] [45] . The pathways that lead to the induction of the antiviral type I interferon (IFN) response are common targets of many viruses, and coronaviruses are no exception. SARS-CoV and MERS-CoV are contained in double membrane vesicles (DMVs), that prevents sensing of its genome [1, 46] . As with most coronaviruses several viral proteins suppress the type I IFN response, and other aspects of innate antiviral immunity [47] . These alterations of the type I IFN response appear to play a role in immunopathology in more than one way. In patients with high initial viral titers there is a poor prognosis [39, 48] . This indicates that reduction of the antiviral response may lead to direct viral-induced pathology. There is also evidence that the delayed type I IFN response can lead to misregulation of the immune response that can cause immunopathology. In a mouse model of SARS-CoV infection, the type I IFN response is delayed [49] . The delay of this potent antiviral response leads to decreased viral clearance, at the same time there is an increase in inflammatory cells of the immune system that cause excessive immunopathology [49] . In this case, the delayed antiviral response not only causes immunopathology, it also fails to properly control the viral replication. While more research is needed, it appears that MERS has a similar effect on the innate immune response [5, 50] . The current treatment and prevention options for SARS-CoV and MERS-CoV are limited. So far there are no licensed vaccines for SAR-CoV or MERS-CoV, although several strategies have been tried in animal models [51, 52] . There are also no antiviral strategies that are clearly effective in controlled trials. During outbreaks several antiviral strategies were empirically tried, but these uncontrolled studies gave mixed results [5, 39] . The main antivirals used were ribavirin, lopinavir and ritonavir [38, 53] . These were often used in combination with IFN therapy [54] . However, retrospective analysis of these data has not led to clear conclusions of the efficacy of these treatment options. Research in this area is still ongoing and it is hoped that we will soon have effective strategies to treat novel CoV [3,36,38,40, [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] . The lack of effective antivirals makes it necessary to examine other potential treatments for SARS-CoV and MERS-CoV. Even if there were effective strategies to decrease viral burden, for these viruses, the potential for new emerging zoonotic CoVs presents additional complications. Vaccines cannot be produced in time to stop the spread of an emerging virus. In addition, as was demonstrated during SARS-CoV and MERS-CoV outbreaks, there is always a challenge during a crisis situation to know which Host resilience to emerging coronaviruses REviEW future science group www.futuremedicine.com antiviral will work on a given virus. One method of addressing this is to develop broad-spectrum antivirals that target conserved features of a given class of virus [65] . However, given the fast mutation rates of viruses there are several challenges to this strategy. Another method is to increase the ability of a given patient to tolerate the disease, i.e., target host resilience mechanisms. So far this has largely been in the form of supportive care, which relies on mechanical ventilation and oxygenation [29, 39, 66] . Since SARS-CoV and MERS-CoV were discovered relatively recently there is a lack of both patient and experimental data. However, many other viruses cause ALI and ARDS, including influenza A virus (IAV). By looking at data from other high pathology viruses we can extrapolate various pathways that could be targeted during infection with these emerging CoVs. This can add to our understanding of disease resilience mechanisms that we have learned from direct studies of SARS-CoV and MERS-CoV. Increased understanding of host resilience mechanisms can lead to future host-based therapies that could increase patient survival [29] . One common theme that emerges in many respiratory viruses including SARS-CoV and MERS-CoV is that much of the pathology is due to an excessive inflammatory response. A study from Josset et al. examines the cell host response to both MERS-CoV and SARS-CoV, and discovered that MERS-CoV dysregulates the host transcriptome to a much greater extent than SARS-CoV [67] . It demonstrates that glucocorticoids may be a potential way of altering the changes in the host transcriptome at late time points after infection. If host gene responses are maintained this may increase disease resilience. Given the severe disease that manifested during the SARS-CoV outbreak, many different treatment options were empirically tried on human patients. One immunomodulatory treatment that was tried during the SARS-CoV outbreak was systemic corticosteroids. This was tried with and without the use of type I IFNs and other therapies that could directly target the virus [68] . Retrospective analysis revealed that, when given at the correct time and to the appropriate patients, corticosteroid use could decrease mortality and also length of hospital stays [68] . In addition, there is some evidence that simultaneous treatment with IFNs could increase the potential benefits [69] . Although these treatments are not without complications, and there has been a lack of a randomized controlled trial [5, 39] . Corticosteroids are broadly immunosuppressive and have many physiological effects [5, 39] . Several recent studies have suggested that other compounds could be useful in increasing host resilience to viral lung infections. A recent paper demonstrates that topoisomerase I can protect against inflammation-induced death from a variety of viral infections including IAV [70] . Blockade of C5a complement signaling has also been suggested as a possible option in decreasing inflammation during IAV infection [71] . Other immunomodulators include celecoxib, mesalazine and eritoran [72, 73] . Another class of drugs that have been suggested are statins. They act to stabilize the activation of aspects of the innate immune response and prevent excessive inflammation [74] . However, decreasing immunopathology by immunomodulation is problematic because it can lead to increased pathogen burden, and thus increase virus-induced pathology [75, 76] . Another potential treatment option is increasing tissue repair pathways to increase host resilience to disease. This has been shown by bioinformatics [77] , as well as in several animal models [30-31,78-79]. These therapies have been shown in cell culture model systems or animal models to be effective, but have not been demonstrated in human patients. The correct timing of the treatments is essential. Early intervention has been shown to be the most effective in some cases, but other therapies work better when given slightly later during the course of the infection. As the onset of symptoms varies slightly from patient to patient the need for precise timing will be a challenge. Examination of potential treatment options for SARS-CoV and MERS-CoV should include consideration of host resilience [29] . In addition to the viral effects, and the pathology caused by the immune response, there are various comorbidities associated with SARS-CoV and MERS-CoV that lead to adverse outcomes. Interestingly, these additional risk factors that lead to a more severe disease are different between the two viruses. It is unclear if these differences are due to distinct populations affected by the viruses, because of properties of the virus themselves, or both. Understanding these factors could be a key to increasing host resilience to the infections. MERS-CoV patients had increased morbidity and mortality if they were obese, immunocompromised, diabetic or had cardiac disease [4, 12] . REviEW Jamieson future science group Risk factors for SARS-CoV patients included an older age and male [39] . Immune factors that increased mortality for SARS-CoV were a higher neutrophil count and low T-cell counts [5, 39, 77] . One factor that increased disease for patients infected with SARS-CoV and MERS-CoV was infection with other viruses or bacteria [5, 39] . This is similar to what is seen with many other respiratory infections. A recent study looking at malaria infections in animal models and human patients demonstrated that resilient hosts can be predicted [28] . Clinical studies have started to correlate specific biomarkers with disease outcomes in ARDS patients [80] . By understanding risk factors for disease severity we can perhaps predict if a host may be nonresilient and tailor the treatment options appropriately. A clear advantage of targeting host resilience pathways is that these therapies can be used to treat a variety of different infections. In addition, there is no need to develop a vaccine or understand the antiviral susceptibility of a new virus. Toward this end, understanding why some patients or patient populations have increased susceptibility is of paramount importance. In addition, a need for good model systems to study responses to these new emerging coronaviruses is essential. Research into both these subjects will lead us toward improved treatment of emerging viruses that cause ALI, such as SARS-CoV and MERS-CoV. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. • Severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus are zoonotic coronaviruses that cause acute lung injury and acute respiratory distress syndrome. • Antivirals have limited effects on the course of the infection with these coronaviruses. • There is currently no vaccine for either severe acute respiratory syndrome coronavirus or Middle East respiratory syndrome coronavirus. • Host resilience is the ability of a host to tolerate the effects of an infection and return to a state of health. • Several pathways, including control of inflammation, metabolism and tissue repair may be targeted to increase host resilience. • The future challenge is to target host resilience pathways in such a way that there are limited effects on pathogen clearance pathways. Future studies should determine the safety of these types of treatments for human patients. Papers of special note have been highlighted as:

What is the role of statins in increasing host resilience to viral lung infections?