Comparing pandemic H1N1 influenza viruses at the molecular level yields key insights about pathogenesis
Jessica A. Belser and Terrence M. Tumpey
Jessica A. Belser is a Microbiologist and Terrence M. Tumpey is a Team Leader in the Influenza Division at the Centers for Disease Control and Prevention, Atlanta, Ga.
Author Profile: Tumpey: Influenza Viruses, Plagues on Screen, but No Dr. Seuss
• Studying the reconstructed 1918 influenza virus is leading to a fuller understanding of other strains within this family, including the 2009 H1N1 virus.
• Particular genes from the 1918 influenza strain, specifically those encoding the hemagglutinin (HA), neuraminidase (NA), and polymerase basic-1 (PB1) proteins, significantly contributed to its replication efficiency and virulence.
• Two genes from the 1918 influenza strain, HA and PB2, confer heightened transmissibility of the virus in ferrets.
• Based on insights from studying the reconstructed 1918 pandemic virus, researchers are designing novel agents to inhibit influenza polymerase proteins; such studies will better prepare us for future pandemics.
The 1918 H1N1 pandemic led to as many as 50 million deaths worldwide, making it the deadliest human influenza virus in recorded history. Poor control measures and the erroneous belief that the culprit was a bacterium called Pfeiffer's bacillus, a gram-negative bacterium now recognized as Haemophilus influenzae, contributed to that dreadful outcome. Subsequent influenza pandemics in 1957 with an H2N2 strain and in 1968 with an H3N2 strain were less severe, offering little to explain the extraordinary virulence of the 1918 strain. As the influenza virus was not isolated in a laboratory until 1930, experts long thought that the 1918 pandemic virus was lost to history.
While intact wild-type H1N1 viruses from 1918 are not available, advances in plasmidbased reverse genetics enabled virologists to reconstruct that virus based on genomic data from RNA fragments from formalin-fixed lung tissues from one patient and from tissues of another victim buried in the Alaskan permafrost. After nearly a decade of effort, the nucleotide sequence of the 1918 influenza virus genome was completed in 2005. The subsequently reconstructed 1918 virus offered researchers the first opportunity to begin identifying genetic determinants that might explain the heightened pathogenicity and transmissibility of this strain.
Last year, another H1N1 influenza subtype of swine origin- called 2009 H1N1- emerged in April, amid reports of influenzalike illnesses, pneumonia cases, and deaths. Laboratory-confirmed cases rose rapidly as the virus spread beyond North America, and a pandemic was declared on 11 June, by which time 74 countries reported more than 28,000 cases with 144 deaths (Fig. 1). When the World Health Organization (WHO) declared the pandemic over in August 2010, the virus had caused more than 18,000 deaths worldwide- considered an underestimate-with confirmed cases in at least 214 countries. Similar to the 1918 pandemic, a majority of 2009 H1N1 infections occurred among young, previously healthy individuals.
Separated by 91 years, the availability of these two H1N1 viruses provides an unparalleled opportunity to better recognize those properties associated with virulent pandemic viruses, thus allowing for a more comprehensive assessment of emerging influenza viruses with pandemic potential.
Molecular Features of Viral Pathogenesis
With the reconstructed 1918 virus, it is possible to compare its molecular features with seasonal and other influenza strains. The main approach to understanding the exceptional virulence of the 1918 virus has been to reconstruct and analyze reassortant viruses, in which genes of the 1918 virus are replaced with genes of less virulent viruses. Although the hemagglutinin (HA), neuraminidase (NA), and polymerase basic protein 1 (PB1) genes appear to contribute significantly to its extraordinary virulence, the unique constellation of all viral genes comprising the reconstructed 1918 virus nonetheless results in a more virulent strain compared with other pandemic viruses.
Initially, the analysis focused on the genes encoding the HA and NA proteins, which protrude from virions as spikes and by which influenza viruses attach to and release from host cells, respectively (Fig. 2). Among the eight 1918 gene segments studied, only theHAgene confers a virulent phenotype in mice when introduced within the genetic background of a contemporary human influenza virus.
Further studies of 1918 reassortant viruses, including in cells derived from human pharyngeal epithelium, revealed that the 1918 HA gene is essential for maximum virus replication of this pandemic strain. Because the influenza HA has multiple functions in the viral life cycle, we can only speculate on how this protein served its function as the primary virulence factor of this pandemic strain.
The HA from the 1918 virus binds efficiently to α2-6-linked sialic acids, the receptor to which human influenza viruses preferentially bind. Avian influenza viruses, by contrast, preferentially bind to α2-3-linked sialic acids. Further, the binding strength of 1918 HA to the human α2-6 glycan receptor appears to be higher than that observed for other influenza strains, including the pandemic 2009 H1N1 virus. This distinction is noteworthy because there is a predominance of α2-6-linked sialic acids in the human upper respiratory tract. Influenza experts generally agree that a flu strain must preferentially bind α2-6-linked sialic acids to spread efficiently among humans.
Recent studies have gone beyond the simple paradigm of avian influenza viruses binding α2-3 linkages and human influenza viruses binding α2-6 linkages. Human adaptation of influenza A viruses appear to be governed by the binding specificity of HA to either short or long (chain length) α2-6 sialylated glycan receptors on respiratory tissues. For instance, the 1918 HAprotein binds to long-chain α2-6 glycans on specific host cells, such as airway goblet (mucinsecreting epithelial) cells. This additional specificity needs to be further explored.
Mixing of other genes between the reconstructed 1918 virus and contemporary H1N1 viruses indicates that the gene encoding NA proteins is also important for optimal 1918 virus replication. For instance, substituting the 1918 NA with the NA from a contemporary human influenza virus moderately affects how the virus replicates in mouse lungs and human airway cells.
The NA protein of influenza A virus is a sialidase that removes sialic acid residues from both viral and host cellular surfaces, promoting spread of virions. It is generally believed that there must be a close cooperation between the HA and NA surface proteins to optimize receptor binding and release of new viruses. Thus, influenza viruses with the right balance between the avidity of HA for terminal sialic acid-linked host glycoproteins and the efficiency of NA in removing sialic acid residues from host cells likely promotes greater progeny virion release. In fact, human airway cells release 50 times more 1918 progeny particles than do contemporary H1N1 viruses, suggesting efficient functional cooperation between the HA and NA surface proteins of this pandemic strain.
The surface glycoproteins are not the only proteins from the reconstructed 1918 virus that influence virulence. The influenza virus RNA polymerase complex consists of three subunits: the PB1, PB2, and polymerase acid protein (PA). Together, these three subunits catalyze sequential additions of nucleotides to elongate RNA chains. According to studies with reassortant viruses, the PB1 gene was essential for maximal replication and virulence of the 1918 pandemic virus. All 20th-century pandemic viruses acquired an avian PB1 gene. Although avian PB1 genes apparently possess greater transcriptional activity compared with PB1 genes from seasonal influenza viruses, proof of these differences awaits direct measurements.
Alternatively, a functional PB1-F2 protein from an avian source-one that is not in many contemporary seasonal viruses, which code for truncated proteins- could contribute to this advantage. The PB1-F2 protein, which is encoded by an alternative open reading frame within the PB1 gene, contributes to increased pathogenicity of the 1918 virus, possibly by inducing apoptosis in host antigen-presenting cells that trigger adaptive immune responses. In particular, a serine at position 66 in the 1918 PB1-F2 protein is linked with enhanced virulence, and changing this amino acid to asparagine reduces morbidity and mortality when this virus infects mice.
Despite its role in blocking type 1 interferon (IFN)-mediated host antiviral responses, the NS1 influenza protein, which is a splice product from the influenza nonstructural (NS) gene, is not a crucial virulence factor for the 1918 virus. However, the C-terminal end of the 1918 NS1 protein contains a PDZ ligand domain, not present in viruses isolated during the 1957 or 1968 pandemics. It may be a virulence determinant for this virus that functions via IFN-independent mechanisms.
Any influenza virus must meet several criteria to cause a pandemic. The virus must infect humans and cause illness, bear a HA subtype to which the human population is immunologically naive, and be capable of sustained human-tohuman transmission. For instance, although highly pathogenic avian influenza viruses of the H5 and H7 subtypes can cause severe disease and death in humans, they meet only the first two of those conditions. These avian viruses are not efficiently transmitted through the air-a requirement to sustain a pandemic.
In contrast, the 2009 H1N1 pandemic virus transmitted from humans in a localized area to all populated continents in a matter of weeks (Fig. 1). The Centers for Disease Control and Prevention (CDC) now estimate that approximately 50 million Americans were infected with this virus, resulting in 200,000 hospitalizations and 10,000 deaths. Despite its global spread, household transmissions of the 2009 H1N1 virus were lower than those of the 1957 or 1968 pandemic viruses. Tests in ferrets further support this finding (see box).
What did we learn from studying the 1918 influenza virus in ferrets? First, virus transmissibility is polygenic, influenced by both virus-host interactions and viral factors. Unlike human H1N1 viruses, avian H1N1 viruses do not transmit efficiently by either respiratory droplets or direct contact in ferrets, and therefore provide an ideal set of genes to mix with genes from the highly transmissible 1918 virus. As anticipated, the reconstructed 1918 virus transmitted readily from infected ferrets to naive ferrets via respiratory droplets, similar to seasonal influenza viruses. By mutating key residues in the reconstructed 1918 virus, we learned that an α2-6-linked sialic acid binding preference is critical to maintain the transmissibility of this pandemic virus. For example, amutant 1918 virus with an α2-3 binding preference failed to transmit by respiratory droplets in ferrets and also caused substantially less sneezing in inoculated ferrets. Interestingly, the incidence of sneezing in inoculated ferrets generally correlates with efficiency of virus transmission to influenza virus-naive ferrets, including 2009 H1N1 viruses, which elicit less sneezing during acute infections compared to seasonal H1N1 viruses or the 1918 virus. Such results along with virus binding data suggest that the lack of sustained human-tohuman transmission of avian H5 and H7 subtype viruses is due primarily to their α2-3 binding preference.
With regard to internal proteins, the 1918 PB2 protein appears both necessary and sufficient for a virus expressing the 1918 HA to transmit by respiratory droplets in ferrets. Furthermore, a lysine at residue 627 in PB2 is generally present among human influenza viruses that transmit efficiently, including the three pandemic viruses of the 20th century. The absence of a lysine at this position among 2009 H1N1 viruses suggests that other molecular determinants affect viral transmission, underscoring the complexity of this behavior.
Not surprisingly, the 2009 H1N1 viruses maintain a human receptor binding affinity for α2-6-linked sialic acids. However, unlike pandemic viruses of the 20th century, 2009 H1N1 viruses bear a PB1 protein from the swine triple reassortant lineage instead of a wholly avian source. Furthermore, the majority of viruses isolated from this pandemic do not code for a functional PB1-F2 protein, nor do they possess a lysine at position 627 in PB2. However, introducing a full-length PB1-F2 protein on the 2009 H1N1 backbone does not appreciably enhance the pathogenicity of this virus. The absence of known virulence determinants among 2009 H1N1 viruses underscores the heterogeneity of these viruses and illustrates the scope of work that lies ahead.
Advances in Modern Medicine
The unprecedented mortality associated with the 1918 pandemic provides a window from which to view the potential of influenza virus disease severity. With nearly a century of medical progress to draw upon, it is important to assess how treatment for a virus demonstrating similar virulence would be conducted today. Analysis of clinical and autopsy materials from the 1918 pandemic reveals that most severe cases from this pandemic were associated with secondary bacterial pneumonia following virusinduced tissue damage. While bacterial infections following influenza virus infections are common, the absence of antibiotics surely accounts for much of excess mortality in 1918. By contrast, severe and fatal cases following infections with highly pathogenic avian influenza viruses result typically from either acute respiratory distress syndrome or immune dysregulation, including hypercytokinemia, also known as a "cytokine storm."
Bacterial coinfections occur despite the availability of antibiotics, and they account for some of the fatal 2009 H1N1 cases. However, unlike the 1918 pandemic, they do not markedly correlate with severe cases. Part of the reason for this difference appears to reside in the PB1-F2 protein of the influenza virus. For instance, mice that are infected with viruses expressing the 1918 version of that protein develop a heightened susceptibility to bacterial coinfections compared to mice infected with wild-type strains of the flu virus. Thus, it is possible that a lack of a functional PB1-F2 protein among 2009 H1N1 viruses is contributing to this observation. Another difference between the 1918 and 2009 influenza pandemics lacks a molecular explanation. Although gastrointestinal distress and vomiting were common features in about 40% of 2009 H1N1 cases, these symptoms were not recorded in reports during the 1918 or other 20th-century pandemics.
Scientists and public health experts proved far better prepared to deal with an influenza pandemic in 2009 than they were in 1918, and rapidly implemented a diverse range of control strategies that were not available a century ago. For instance, a vaccine against the 2009 H1N1 virus was developed, manufactured, and distributed within 6 months after the first human cases were detected. Moreover, testing quickly revealed that the 2009 H1N1 virus was sensitive to available neuraminidase inhibitors, leading officials to release stockpiled drugs to meet heightened demand during the pandemic. Other measures to quell the pandemic included use of advanced personal protective equipment, indoor ventilation and filtration systems, and other safety measures to reduce the likelihood of exposure to the virus among health care workers and laboratorians. If those measures were available in 1918, the estimated 2% mortality of that pandemic might have been reduced.
Despite recent progress, many questions about the 1918 pandemic virus remain unanswered. Nonetheless, studies based on reconstruction of this virus illustrate the importance of the HA and polymerase genes for optimizing replication (PB1) and transmission (PB2) of this pandemic strain. Insights from these and other influenza studies continue to guide the development of vaccines and therapeutics to prevent and treat viral infections. For example, researchers are now designing agents to inhibit the viral polymerase proteins, including PB1-F2, by targeting key intersubunit binding sites within the polymerase complex. These and other studies are helping to prepare us for future pandemics.
This article is based on a talk presented at the 49th ICAAC meeting in September 2009. We thank Byron Tsang for graphical assistance and our colleagues at the Influenza Division, Centers for Disease Control and Prevention, who have contributed to some of the work discussed in this feature. The findings and conclusions in this report are those of the authors and do not necessarily reflect the views of the funding agency.
Basler, C. F., A. H. Reid, J. K. Dybing, T. A. Janczewski, T. G. Fanning, H. Zheng, M. Salvatore, M. L. Perdue, D. E. Swayne, A. Garcia-Sastre, P. Palese, and J. K. Taubenberger. 2001. 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. USA 98:2746- 2751.
Chandrasekaran, A., A. Srinivasan, R. Raman, K. Viswanathan, S. Raguram, T. M. Tumpey, V. Sasisekharan, and R. Sasisekharan. 2008. Glycan topology determines human adaptation of avian H5N1 virus hemagglutinin. Nature Biotechnol. 26:107-113.
Garten, R. J., C. T. Davis, C. A. Russell, B. Shu, S. Lindstrom, A. Balish, W. M. Sessions, X. Xu, E. Skepner, V. Deyde, M. Okomo-Adhiambo, L. Gubareva, J. Barnes, C. B. Smith, S. L. Emery, M. J. Hillman, P. Rivailler, J. Smagala, M. de Graaf, D. F. Burke, R. A. Fouchier, C. Pappas, C. M. Alpuche-Aranda, H. Lopez-Gatell, H. Olivera, I. Lopez, C. A. Myers, D. Faix, P. J. Blair, C. Yu, K. M. Keene, P. D. Dotson, Jr., D. Boxrud, A. R. Sambol, S. H. Abid, K. St George, T. Bannerman, A. L. Moore, D. J. Stringer, P. Blevins, G. J. Demmler-Harrison, M. Ginsberg, P. Kriner, S. Waterman, S. Smole, H. F. Guevara, E. A. Belongia, P. A. Clark, S. T. Beatrice, R. Donis, J. Katz, L. Finelli, C. B. Bridges, M. Shaw, D. B. Jernigan, T. M. Uyeki, D. J. Smith, A. I. Klimov, and N. J. Cox. 2009. Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science 325:197-201.
Kobasa, D., A. Takada, K. Shinya, M. Hatta, P. Halfmann, S. Theriault, H. Suzuki, H. Nishimura, K. Mitamura, N. Sugaya, T. Usui, T. Murata, Y. Maeda, S. Watanabe, M. Suresh, T. Suzuki, Y. Suzuki, H. Feldmann, and Y. Kawaoka. 2004. Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus. Nature 431:703-707.
Maines, T. R., A. Jayaraman, J. A. Belser, D. A. Wadford, C. Pappas, H. Zeng, K. M. Gustin, M. B. Pearce, K. Viswanathan, Z. H. Shriver, R. Raman, N. J. Cox, R. Sasisekharan, J. M. Katz, and T. M. Tumpey. 2009. Transmission and pathogenesis of swine-origin 2009 A(H1N1) influenza viruses in ferrets and mice. Science 325:484-487.
Matrosovich, M. N., T. Y. Matrosovich, T. Gray, N. A. Roberts, and H. D. Klenk. 2004. Human and avian influenza viruses target different cell types in cultures of human airway epithelium. Proc. Natl. Acad. Sci. USA 101:4620-4624.
Morens, D. M., J. K. Taubenberger, H. A. Harvey, and M. J. Memoli. 2010. The 1918 influenza pandemic: lessons for 2009 and the future. Crit. Care Med. 38:e10-20.
Pappas, C., P. V. Aguilar, C. F. Basler, A. Solorzano, H. Zeng, L. A. Perrone, P. Palese, A. Garcia-Sastre, J. M. Katz, and T. M. Tumpey. 2008. Single gene reassortants identify a critical role for PB1, HA, andNAin the high virulence of the 1918 pandemic influenza virus. Proc. Natl. Acad. Sci. USA 105:3064-3069.
Tumpey, T. M., C. F. Basler, P. V. Aguilar, H. Zeng, A. Solorzano, D. E. Swayne, N. J. Cox, J. M. Katz, J. K. Taubenberger, P. Palese, and A. Garcia-Sastre. 2005. Characterization of the reconstructed 1918 Spanish influenza pandemic virus. Science 310:77-80.
Tumpey, T. M., T. R. Maines, N. Van Hoeven, L. Glaser, A. Solorzano, C. Pappas, N. J. Cox, D. E. Swayne, P. Palese, J. M. Katz, and A. Garcia-Sastre. 2007. A two-amino acid change in the hemagglutinin of the 1918 influenza virus abolishes transmission. Science 315:655-659.