The 2009 H1N1 influenza outbreak in its historical context

Journal of Clinical Virology 45 (2009) 174–178

Contents lists available at ScienceDirect

Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv

Review

The 2009 H1N1 influenza outbreak in its historical context Derek Gatherer ∗ MRC Virology Unit, Institute of Virology, Church Street, Glasgow G11 5JR, UK

a r t i c l e

i n f o

Article history: Received 4 June 2009 Accepted 5 June 2009 Keywords: Influenza A Swine flu Porcine influenza H1N1 Pandemic Zoonotic

a b s t r a c t Of the 16 known serotypes of influenza A haemagglutinin, 6 have been isolated from humans at the molecular level (H1, H2, H3, H5, H7, H9). 3 of these have been involved in past pandemics (H1, H2, H3). Traditional pandemic surveillance has focussed on monitoring antigenic shift, meaning the re-assortment of novel haemagglutinins into seasonal human influenza A viruses during rare events of double infection with seasonal and zoonotic strains. H5, from avian H5N1 influenza, has been the major cause for concern in recent years. However, the 2009 H1N1 zoonotic event demonstrates that even serotypes already encountered in past human pandemics may constitute new pandemic threats. The protein sequence divergence of the 2009 zoonotic H1 from human seasonal influenza H1 is around 20–24%. A similar level of divergence is found between the 2009 H1 and European swine flu. By contrast, its divergence from North American swine flu strains is around 1–9%. Given that the divergence between H1 and its nearest serotype neighbour H2 is around 40–46%, the 2009 H1 may be broadly considered as halfway towards a new serotype. The current situation is one of antigenic pseudo-shift. © 2009 Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Introduction: the basic biology of influenza . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The lessons of past pandemics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Porcine influenza in pigs and humans: (pseudo)-serotypes and (pseudo)-pandemics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction: the basic biology of influenza Influenza A (Family Orthomyxoviridae, Genus Influenzavirus A) is currently the greatest pandemic disease threat to humankind. Its rivals for this title (HIV-1, Ebola, SARS, pneumonic plague) have higher mortality if untreated, but either lack influenza’s rapid inter-personal transmission (HIV-1) or its widespread seasonal distribution (Ebola, SARS, pneumonic plague). Influenza A is unique among the major pandemic threats in that it could potentially infect 30% of the world’s population within a matter of months. Even at a conservative overall mortality rate of 2%, it would result in around 135 million deaths worldwide within the first year of a new pandemic outbreak. This is about 4 times the total mortality attributed to HIV-1 in the last 30 years.

∗ Tel.: +44 141 330 6268. E-mail address: [email protected]. 1386-6532/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2009.06.004

174 175 176 178 178

Influenza’s endemic reservoir is in aquatic wildfowl, many of which are migratory. Containment of the disease outside of mammals is therefore virtually impossible. Once a novel strain of influenza has crossed the species barrier from birds into a mammalian host, it may persist in that new host species for many decades. Molecular evidence exists for the presence of influenza A in at least 18 mammalian species (Table 1). It is also capable of transmission between mammals, and the current influenza A H1N1 2009 outbreak is now known to have originated in pigs.2,3 The orthomyxoviruses are single-stranded RNA viruses of the negative strand variety, and include the Isavirus and Thogotovirus genera. The first of these causes anaemia in fish, and the latter includes several tick-borne mammalian encephalitis viruses which can be clinically serious. In addition to these there are several unclassified orthomyxoviruses and three genera of influenza viruses: A, B and C. The last two are clinically milder than influenza A, but influenza B is nevertheless severe enough to be included in seasonal vaccination programmes. The absence of life-long

D. Gatherer / Journal of Clinical Virology 45 (2009) 174–178

175

Table 1 Species distribution of haemagglutinin serotypes from the NCBI Influenza Resource (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database, accessed 3rd June 2009). Note that serotypes 1–5, 7, 9, 10 and 13 have all been found at least once in mammals. Species Avian Blow fly Camel Canine Cat Civet Equine Ferret Giant anteater Human Leopard Mink Muskrat Plateau pika Racoon dog Seal Stone marten Swine Tiger Whale Total

H1

H2

93

105

H3 265

H4

H5

H6

H7

H8

H9

H10

H11

H12

H13

H14

H15

H16

212

1918 1

370

509

26

904

86

88

44

39

3

6

14

4,682 1 1 8 7 1 181 1 1 6,758 6 3 1 1 1 5 1 701 7 8

88

44

40

3

6

14

12,375

1 8 7 1 164 1 1 1802

106

17

4610

224 6 1

1

8

8 1

1 1 1 2

1

419

1 2

2

238

2

12 7 7

2325

214

5280

216

269

28 1 370

543

immunity is a major contributor to influenza A’s pandemic potential, and features of the influenza virus make design of a vaccine that would confer immunity against all strains, highly problematic.4 The first difficulty is that the majority of the cellular immune response to an influenza virus infection is directed against two proteins on the virus’s surface, haemagglutinin and neuraminidase. Influenza strains are therefore usually broadly categorised according to their serotype for these two proteins; for instance the current outbreak is H1N1, and previous pandemics have involved H3N2 and H2N2. There are in total 16 serotypes of haemagglutinin, and 9 serotypes of neuraminidase. Haemagglutinin functions in the binding of the virus to its putative target cell, whereas neuraminidase plays a role in exit of the virus from the cell in preparation for another round of infection. Haemagglutinin in particular has a high amino acid substitution rate in its epitope regions. The ratio of nonsynonymous substitutions to synonymous substitutions (omega) significantly exceeds 1 at several amino acid positions, indicating a selective advantage for novel amino acid sequences in this region.6,8 This appears to be directly related to evasion of the host immune system. This process of evolution of a single haemagglutinin serotype is often referred to in the influenza literature as antigenic drift,4 although this is something of a misnomer given that drift is used by evolutionary biologists to refer to neutral evolution, meaning evolution in the absence of selective pressures. On the contrary, influenza A haemagglutinin evolution is a mixture of functional constraint and positive selection for variability in epitope regions. Antigenic “drift” can render immunity acquired in one influenza season, either by infection with the virus or by vaccination, useless in a short period of time. Repeated seasonal vaccination for vulnerable groups has therefore become a necessary public heath policy. A second and more serious problem is a result of the fact that the influenza genome, although small – encoding just 11 proteins – is divided into 8 segments. In the rare event of a double infection with two different strains of influenza into a single host, re-assortment of the genome segments can occur, producing a series of completely novel combinations of genome segments in the progeny viruses. When the double infection is with viruses originating in different species, for instance if a patient has a human seasonal influenza simultaneous with a zoonotic influenza such as porcine or avian influenza, such re-assorted strains may be the source of new pandemic influenza variants incorporating haemagglutinin and/or

26

940

87

Total

neuraminidase proteins against which hosts may have virtually no immunity. This sudden introduction of completely new serotypes of haemagglutinin into circulating human viruses is referred to as antigenic shift. Although serotypes H1, H2 and H3 are the only ones known to have been involved in influenza pandemics, molecular evidence exists for the occasional infection of humans with avian viruses containing H5, H7 and H9. The prospect of an antigenic shift involving any of these is of major concern in pandemic surveillance. The current H1N1 outbreak is now known to involve a re-assorted virus produced from two kinds of porcine influenza, one of which is itself already a “triple re-assortant” strain containing segments originating ultimately in human seasonal H3N2 influenza and in avian influenza as well as porcine influenza.2 2. The lessons of past pandemics The “Spanish Flu” of 1918–1920 is the earliest known pandemic for which hard molecular evidence exists for the involvement of influenza A. Work is currently underway to isolate influenza haemagglutinin sequences from clinical material dating from the previous pandemic event of 1889–1890.1 This is widely regarded as an influenza pandemic on clinical, epidemiological and some limited serological grounds. Prior to this date, identification of influenza relies on diagnostic detective work in medical records. The similarity of the symptoms of influenza to other feverish respiratory viruses inevitably renders some of this a little speculative, but nevertheless candidate previous widespread influenza outbreaks have been identified in 1173, 1510, 1580, 1729, 1781 and 1830. The term influenza was only used to describe these pandemics from the 18th century onwards (although it had previous medical use dating back to the 14th century to describe epidemics in general), and the earliest case description that satisfies modern clinicians is that of the French physician Molineux in 1694.5 The 1918–1920 H1N1 Spanish Flu, which killed 40 million people worldwide is informative as a “worst case scenario” for a flu pandemic. The H1 haemagglutinin in this pandemic may have been of avian origin, and the disease was first detected in the USA in prisons and military bases. Troop movements during the First World War appear to have been a major cause of its global distribution. Antigenic drift subsequent to the initial antigenic shift is the likely cause of the several waves of renewed virulence shown by the virus. Spanish Flu illustrates both the potential of influenza for

176

D. Gatherer / Journal of Clinical Virology 45 (2009) 174–178

Fig. 1. Phylogenetic tree of H1 haemagglutinin sequences in influenza A H1N1 strains infecting humans from 1918 to the present day. 3 clades can be seen: (1) the 1918 pandemic strain which is an outlier; (2) strains circulating seasonally since at least the 1930s (lower of 2 major clades); (3) zoonotic strains (upper clade). The positions of the current outbreak and the most recent vaccine strains are indicated. Tree drawn using the NCBI Influenza Virus Resource tools: (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database).

morbidity and mortality and also the tendency of severe pandemics to occur in several waves as the virus adapts to its new human hosts. After 1920, some gaps exist in the molecular sequence record, but H1N1 was certainly a regular seasonal influenza from 1934 until the mid-1950s. It then reappeared in 1977–1978 and current human seasonal H1 proteins belong to the same lineage that has been circulating at least since the 1930s. Fig. 1 shows a phylogenetic tree of the haemagglutinin proteins of all H1N1 strains circulating in humans since the Spanish Flu. The 1918 protein is an outgroup, and the remainder of the sequences are either seasonal H1N1 from the 1934/1977 lineage or zoonotic strains. Between the temporary disappearance of H1 in the 1950s and its reappearance in the 1970s, there were two further pandemics: the 1957–1958 H2N2 “Asian Flu” with 1.5 million fatalities worldwide and the 1968–1969 H3N2 “Hong Kong Flu” with 1 million fatalities worldwide. H2 has been absent from human populations since the late 1960s, but H3, like H1, persists in a seasonal human influenza evolving by antigenic drift. The fact that both H1 and H2 have in the past disappeared – in the former case only temporarily – from human populations demonstrates that previous candidate pandemics of influenza may have involved haemagglutinin serotypes not currently found in humans. Although only H1, H2, H3, H5, H7 and H9 are known to have infected humans, there is no particular reason to exclude the possibility that humans have in the past been, or may in the future be, infected with the remaining 10 serotypes. Any pandemic involving a haemagglutinin serotype not seen in the last century, will almost certainly be very severe.

3. Porcine influenza in pigs and humans: (pseudo)-serotypes and (pseudo)-pandemics The association of the origins of the current outbreak with the Mexican pig farming region raised immediate suspicions that porcine influenza was involved, and it was soon demonstrated that the nearest relative of the strains isolated in the latest outbreak was the triple re-assortant porcine influenza that had caused considerable problems for pig farmers for several years.2,3 The new strain incorporated the results of a further re-assortment event, thus generating a quadruple re-assortant virus with genome segments traceable to two major lineages of porcine influenza as well as avian and human influenzas. Porcine influenza demonstrates two major clades in its haemagglutinin protein evolution, associated with what are termed “Classic North American Swine Flu” and “Eurasian Swine Flu” (Fig. 2). In addition there are other porcine influenza haemagglutinins of more uncertain affinities, possibly resulting from independent re-assortment events with avian H1 haemagglutinins. Of all the proven 18 mammalian hosts for influenza, pigs will most readily transmit to humans. Molecular evidence exists for several porcine influenza strains in humans since the 1980s, and all of these have haemagglutinin proteins belonging to the Classic North American Swine Flu lineage. Until the present outbreak however, no cases of porcine influenza were capable of sustained human-tohuman transmission. Table 2 shows that the amino acid divergence between haemagglutinin proteins from classic and Eurasian swine flu strains is 20–25%, approximately the same as either class

D. Gatherer / Journal of Clinical Virology 45 (2009) 174–178

177

Fig. 2. Phylogenetic tree of H1 haemagglutinin sequences in influenza A H1N1 strains infecting pigs from 1930 to the present day. 3 clades can be seen: (1) some early swine flu strains from 1930 to 1945, which are outliers; (2) “Classic” North American swine flu (lower of 2 major clades); (3) atypical swine flu (upper clade, in yellow), including “Eurasian” swine flu and some further atypical strains, including one from 1930 and a group from 2004 to 2006. Tree drawn using the NCBI Influenza Virus Resource tools: (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database).

differ from seasonal human H1, but considerably less than the average 40–46% divergence between H1 and H2 haemagglutinin proteins. At the time of writing (3rd June 2009), it remains unclear if the current outbreak will achieve pandemic status. Although the World Health Organization has reached alert level 5, indicating “pandemic imminent”, and this may rise to level 6, “pandemic underway”, if the rate of infection climbs, several criteria may not yet have met the definition of a pandemic used in much the influenza literature.

For instance, mortality is so far low (113 deaths from 17,436 cases as of 2nd June 2009), and the 1977–1978 Russian Flu is sometimes not recognised as a pandemic on account of its low mortality and restricted preferred host range (predominantly children and young adults). Furthermore, previous pandemics have involved antigenic shift: H1 for Spanish Flu, H2 for Asian Flu and H3 for Hong Kong Flu. The current outbreak is serotype H1 and thus by this criterion not an antigenic shift. However, as a Classic North American Swine Flu H1, it is rather different to the H1 haemagglutinins that have seasonally

Table 2 Number of amino acid substitutions per site over a 561 residue alignment between representative influenza A haemagglutinin proteins. Analyses were conducted using the Poisson correction method in MEGA4.7,9 Divergences of 20–40% are coloured yellow and those of >40% are orange. Representative sequences are (1) 2009 H1N1: ACR09372 A/Mexico/3955/2009, (2) 1976 swine in humans: ACQ99821 A/New Jersey/8/1976, (3) 1977 human pandemic: ABD60933 A/USSR/92/77, (4) 2008 seasonal H1: ACJ73758 A/California/07/2008, (5) 1935 human H1: ABD62781 A/Melbourne/35, (6) 1957 H2 pandemic: BAC43764 A/Kayano/57, (7) 2006 avian H2: CAQ77186 A/wigeon/Norway/10 1783/2006, (8) 2009 classic swine: ACR01025 A/swine/Alberta/OTH-33-8/2009, (9) 1986 classic swine: ABY81426 A/swine/Iowa/1/1986, (10) 1939 Euroswine: BAA00718 A/swine/Cambridge/1939, (11) 1994 Euroswine: AAD05215 A/swine/Scotland/410440/94, and (12) 2006 Euroswine: ACN72617 A/swine/Hungary/19774/2006. Note that the inter-serotype distance (H1 vs. H2) is always greater than 40%, whereas intra-H1 serotype distances range from just 1% between the latest outbreak and recent strains of swine flu, to 25% between classic swine flu and European swine flu and the same between classic swine flu and seasonal human influenza.

178

D. Gatherer / Journal of Clinical Virology 45 (2009) 174–178

circulated in human since the 1930s. Table 2 shows the percentage identity between the sequences in the tree. The divergence between the various H1 proteins and H2 is 40–46% while the divergence between human and porcine H1 is less than 25%. Nevertheless, this is clearly rather more than the 9–15% between seasonal human H1s. A useful comparison may be made with the relationships between haemagglutinin proteins within each of the two lineages, Victoria and Yamagata, of influenza B (data not shown). The influenza B haemagglutinin have intra-lineage similarities at less than 4% divergence, with sequence divergence in inter-lineage comparisons at 6–7% on average. The original Lee strain of influenza B from 1940 is approximately 7% divergent to modern samples of either influenza B lineage. By comparison, influenza A HA sequences are 75–76% divergent from influenza B HA. The arrival of porcine H1 in human influenza viruses may be most accurately described as an antigenic pseudo-shift or para-shift. Therefore the current H1N1 outbreak, although a zoonotic influenza originating in a re-assortment event and exhibiting sustained human-to-human transmission for the first time, may have insufficiently divergent haemagglutinin to be regarded as an antigenic shift, and may be insufficiently virulent ultimately to enter the annals of major pandemics. However, its importance should not be discounted. For the first time, the full repertoire of molecular biology has been applied to a novel form of influenza spreading on a global scale. At the time of writing (3rd June 2009), it appears that rapid molecular analysis and diagnosis coupled to effective clinical and public health methods have at the very least substantially slowed the progress of the outbreak. None of this was possible in 1957 or even in 1977. This provides an important dress rehearsal for potential H5N1 outbreaks. We still cannot stop the arrival of antigenically shifted pandemic influenzas, but in the past we were unaware of their existence until they were already upon us. In future, we are at least likely to see the dust of an approaching pandemic rising in the distance, giving us crucial time to react preventatively.

Finally, the next few months should reveal the answers to several questions: (1) Will antigenic drift sharpen the virulence of the outbreak, as it did in 1918–1920? (2) Will there be double infections with seasonal influenzas increasing the possibility of further reassortant strains incorporating elements of swine flu with seasonal H1 and H3, again potentially increasing virulence? (3) Will swine flu settle into a seasonal pattern in humans, alongside the more established seasonal strains? If so, it may substantially increase the costs of vaccine development. The major legacy of 2009 porcine H1N1 may be economic as much as medical. Conflict of interest statement The author has no conflict of interest. References 1. Altschuler EL, Kariuki YM, Jobanputra A. Extant blood samples to deduce the strains of the 1890 and possibly earlier pandemic influenzas. Med Hypotheses 2009. 2. Dawood FS, and Team. Emergence of a novel Swine-Origin influenza A (H1N1) virus in humans. N Engl J Med 2009. 3. Fraser C, Donnelly CA, Cauchemez S, Hanage WP, Van Kerkhove MD, Hollingsworth TD, et al. Pandemic potential of a strain of influenza A (H1N1): early findings. Science 2009. 4. Hay AJ, Gregory V, Douglas AR, Lin YP. The evolution of human influenza viruses. Phil Trans R Soc Lond B: Biol Sci 2001;356:1861–70. 5. Potter CW. A history of influenza. J Appl Microbiol 2001;91:572–9. 6. Suzuki Y. Positive selection operates continuously on hemagglutinin during evolution of H3N2 human influenza A virus. Gene 2008;427:111–6. 7. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 2007;24:1596–9. 8. Wolf YI, Viboud C, Holmes EC, Koonin EV, Lipman DJ. Long intervals of stasis punctuated by bursts of positive selection in the seasonal evolution of influenza A virus. Biol Direct 2006;1:34. 9. Zuckerkandl E, Pauling L. Evolutionary divergence and convergence in proteins. In: Bryson V, Vogel H, editors. Evolving genes and proteins. New York: Academic Press; 1965. p. 97–166.