- Record: found
- Abstract: found
- Article: found
Heterosubtypic Immunity to Influenza A Virus Infections in Mallards May Explain Existence of Multiple Virus Subtypes
research-article
Author(s):
Neus Latorre-Margalef
1 ,
Vladimir Grosbois
2 ,
John Wahlgren
3 ,
Vincent J. Munster
4
,
5 ,
Conny Tolf
1 ,
Ron A. M. Fouchier
4 ,
Albert D. M. E. Osterhaus
4 ,
Björn Olsen
6 ,
Jonas Waldenström
1
,
*
Publication date (Electronic):
20 June 2013
Journal:
PLoS Pathogens
Publisher:
Public Library of Science
Read this article at
There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.
Abstract
Wild birds, particularly duck species, are the main reservoir of influenza A virus (IAV) in nature. However, knowledge of IAV infection dynamics in the wild bird reservoir, and the development of immune responses, are essentially absent. Importantly, a detailed understanding of how subtype diversity is generated and maintained is lacking. To address this, 18,679 samples from 7728 Mallard ducks captured between 2002 and 2009 at a single stopover site in Sweden were screened for IAV infections, and the resulting 1081 virus isolates were analyzed for patterns of immunity. We found support for development of homosubtypic hemagglutinin (HA) immunity during the peak of IAV infections in the fall. Moreover, re-infections with the same HA subtype and related prevalent HA subtypes were uncommon, suggesting the development of natural homosubtypic and heterosubtypic immunity ( p-value = 0.02). Heterosubtypic immunity followed phylogenetic relatedness of HA subtypes, both at the level of HA clades ( p-value = 0.04) and the level of HA groups ( p-value = 0.05). In contrast, infection patterns did not support specific immunity for neuraminidase (NA) subtypes. For the H1 and H3 Clades, heterosubtypic immunity showed a clear temporal pattern and we estimated within-clade immunity to last at least 30 days. The strength and duration of heterosubtypic immunity has important implications for transmission dynamics of IAV in the natural reservoir, where immune escape and disruptive selection may increase HA antigenic variation and explain IAV subtype diversity.
Author Summary
Influenza A viruses (IAV) infect a range of hosts, with the largest diversity being found in waterfowl, particularly dabbling ducks. In these hosts, IAV causes only mild disease, while viruses that infect other hosts, such as poultry, horses or humans, can cause fatal infections. In fact, all known pandemic flu viruses have contained gene segments that originated in the wild bird reservoir. We sampled a wild population of Mallards over eight seasons and characterized infection histories in 7728 birds. For hemagglutinin (HA) the subtype recoveries indicated that once a Mallard has been infected, re-infection with the same HA subtype is uncommon within the next month, clearly indicating homosubtypic immunity. Moreover, we found evidence for natural heterosubtypic immunity, where phylogenetically related HA subtypes at clade and group levels were less common in re-infections than expected. On the contrary no specific patterns of immunity was found for neuraminidase subtypes. IAVs exist in numerous antigenic subtypes that co-circulate. The strength of heterosubtypic immunity in natural infections provides evidence that HA subtypes compete over hosts and that immune escape may result in positive selection for HA antigenic variation in the virus, and thus explain IAV subtype diversity.
Related collections
Most cited references36
- Record: found
- Abstract: found
- Article: not found
Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls.
Ron Fouchier, Vincent Munster, Anders Wallensten … (2005)
In wild aquatic birds and poultry around the world, influenza A viruses carrying 15 antigenic subtypes of hemagglutinin (HA) and 9 antigenic subtypes of neuraminidase (NA) have been described. Here we describe a previously unidentified antigenic subtype of HA (H16), detected in viruses circulating in black-headed gulls in Sweden. In agreement with established criteria for the definition of antigenic subtypes, hemagglutination inhibition assays and immunodiffusion assays failed to detect specific reactivity between H16 and the previously described subtypes H1 to H15. Genetically, H16 HA was found to be distantly related to H13 HA, a subtype also detected exclusively in shorebirds, and the amino acid composition of the putative receptor-binding site of H13 and H16 HAs was found to be distinct from that in HA subtypes circulating in ducks and geese. The H16 viruses contained NA genes that were similar to those of other Eurasian shorebirds but genetically distinct from N3 genes detected in other birds and geographical locations. The European gull viruses were further distinguishable from other influenza A viruses based on their PB2, NP, and NS genes. Gaining information on the full spectrum of avian influenza A viruses and creating reagents for their detection and identification will remain an important task for influenza surveillance, outbreak control, and animal and public health. We propose that sequence analyses of HA and NA genes of influenza A viruses be used for the rapid identification of existing and novel HA and NA subtypes.
- Record: found
- Abstract: found
- Article: not found
The Evolutionary Genetics and Emergence of Avian Influenza Viruses in Wild Birds
Vivien Dugan, Rubing Chen, David Spiro … (2008)
Introduction Low pathogenic (LP), antigenically diverse influenza A viruses are widely distributed in wild avian species around the world. They are maintained by asymptomatic infections, most frequently documented in aquatic birds of the orders Anseriformes and Charadriformes. As such, wild birds represent major natural reservoirs for influenza A viruses [1]–[11] and at least 105 species of the more than 9000 species of wild birds have been identified as harboring influenza A viruses [8],[12],[13]. These influenza A viruses, commonly referred to as avian influenza viruses (AIV), possess antigenically and genetically diverse hemagglutinin (HA) [14] and neuraminidase (NA) subtypes, which includes all known influenza A virus HA (H1–H16) and NA (N1–N9) subtypes. At least 103 of the possible 144 type A influenza A virus HA-NA combinations have been found in wild birds [8],[15]. AIV maintained in wild birds have been associated with stable host switch events to novel hosts including domestic gallinaceous poultry, horses, swine, and humans leading to the emergence of influenza A lineages transmissible in the new host. Adaptation to domestic poultry species is the most frequent [16]–[26]. Sporadically, strains of poultry-adapted H5 or H7 AIV evolve into highly pathogenic (HP) AIV usually through acquisition of an insertional mutation resulting in a polybasic amino acid cleavage site within the HA [15],[25]. The current panzootic of Asian-lineage HP H5N1 AIV appears to be unique in the era of modern influenza virology, resulting in the deaths of millions of poultry in 64 countries on three continents either from infection or culling. There are also significant zoonotic implications of this panzootic, with 379 documented cases in humans, resulting in 239 deaths in 14 countries since 2003 (as of April 2008 [27]). The Asian lineages of HP H5N1 AIV have also caused symptomatic, even lethal, infections of wild birds in Asia and Europe, suggesting that migratory wild birds could be involved in the spread of this avian panzootic [28]–[31]. Concern is heightened since wild birds are also likely to be the reservoir of influenza A viruses that switch hosts and stably adapt to mammals including horses, swine, and humans [3]. The last three human influenza pandemic viruses all contained two or more novel genes that were very similar to those found in wild birds [16],[20],[32],[33]. Despite the recent expansion of AIV surveillance [7],[8],[10],[34],[35] and genomic data [5], [36]–[38], fundamental questions remain concerning the ecology and evolution of these viruses. Prominent among these are: (i) the structure of genetic diversity of AIV in wild birds, including the role played by inter-hemispheric migration, (ii) the frequency and pattern of segment reassortment, and (iii) the evolutionary processes that determine the antigenic structure of AIV, maintained as discrete HA and NA subtypes. Herein, we address these questions using the largest data set of complete AIV genomes compiled to date. Results/Discussion Global Genome Diversity of AIV The complete genomes of 167 influenza A viruses isolated from 14 species of wild Anseriformes in 4 locations in the U.S. (Alaska, Maryland, Missouri, and Ohio) were sequenced; viral isolates consisted of 29 HA and NA combinations, including 11 HA subtypes (H1–H8, H10–H12) and all 9 neuraminidase subtypes (N1–N9). These sequences were collected as part of an ongoing AIV surveillance project at The Ohio State University and collaborators in other states (1986–2005) using previously described protocols [39], and more than double the number of complete U.S.-origin avian influenza virus genomes available in GenBank. In total, 1340 viral gene segment sequences (2,226,085 nucleotides) were determined (Table S1) and are listed on the Influenza Virus Resource website (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/shipment.cgi). Cloacal samples from wild birds frequently show evidence of mixed infections with influenza viruses of different subtypes by serologic analysis [39]–[41]. Therefore, the isolates chosen for sequence analysis were subjected to sequential limiting dilutions (SLD) [39]. The amplification and sequencing pipeline employed a ‘universal’ molecular subtyping strategy in which every sample was amplified with sets of overlapping primers representing all HA and NA subtypes. In this manner, samples without clear prior subtype information, and/or mixed samples, could be accurately analyzed. Despite performing SLD, 4 samples were shown by sequence analysis to represent a mixed infection (yielding sequence with more than one HA and/or NA subtype. In addition 40 samples had mismatches between the initial antigenic subtyping results (determined on first- or second-egg-passage isolates prior to SLD) and the subtype determined by sequence analysis of cDNA (following one SLD of low-egg-passage isolates) which suggests the possibility of minor populations of antigenically distinct viruses in the low-passage isolate that outgrew the dominant antigenic population in a foreign host system during the SLD or that mixed infections in first egg passage stock caused difficulty in initial subtyping and a dominant strain emerged during SLD (see table of viral isolates at http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/shipment.cgi to examine the discordant results observed). Thus, up to 44 of 167 (26%) of isolates potentially represent mixed infections in the initial cloacal sample. Given the SLD procedure, the true rate of mixed infection, as defined by the presence of >1 HA and/or NA subtype, was likely to be even higher, although mis-serotyping cannot also be ruled out. Sequencing viral genomes directly from primary cloacal material would be the only way to assess the mixed infection frequency, in a manner unbiased by culture, but no such studies have yet been attempted to our knowledge. For a more comprehensive analysis of AIV diversity, the AIV genomes from this study were compared to other AIV genomes available on GenBank [38]. In total, 452 HA sequences and 473 NA sequences, representative of the global diversity of AIV, were used in phylogenetic analyses. For the internal protein genes (PB2, PB1, PA, NP, M, NS), a subset of 407 complete globally-sampled AIV genomes was used to assess the degree of linkage among gene segments. Phylogenetic trees for the HA alignment (Figures 1a and S1) and NA alignment (Figure 1b and S2) are shown here. Phylogenetic trees for the six other gene segments are presented in Figures S3, S4, S5, S6, S7 and S8. 10.1371/journal.ppat.1000076.g001 Figure 1 Maximum likelihood trees of HA and NA genes. (a) Maximum likelihood tree of the HA gene segment of 452 isolates of avian influenza A virus, including representatives of all 16 subtypes. For clarity, all branches within individual subtypes have been collapsed and color-coded to signify individual subtypes. Bootstrap values above 70% are shown next to relevant branches. Branch lengths are scaled according to the number of nucleotide substitutions per site. See Figure S1 for an expanded form of this tree. (b) Maximum likelihood tree of the NA gene segment of 473 isolates of avian influenza A virus, including representatives of all 9 subtypes. The mix of HA subtypes (color-coded according to Figure 1a) observed within each NA type is shown, highlighting the frequency of reassortment. For clarity, all branches within individual subtypes have been collapsed. Bootstrap values above 65% are shown next to the relevant branches. Branch lengths are scaled according to the number of nucleotide substitutions per site. See Figure S2 for an expanded form of this tree, in which individual viral isolates are marked. The topology of the HA phylogeny reflects the antigenically defined subtypes, with some higher-order clustering among them (e.g., H1, H2, H5 and H6; H7, H10 and H15; Figures 1a and S1), as seen previously in smaller studies [14],[42]. Although most subtypes are found in numerous avian species and occupy wide global distributions, this phylogenetic structure indicates that HA subtypes did not originate in a single radiation. More striking was the high level of genetic diversity between different subtypes; the average amino acid identity of 120 inter-subtype comparisons of full-length HA was 45.5%. As expected, inter-subtype comparisons of the HA1 domain exhibited more diversity, with an average inter-subtype identity of 38.5%. In contrast, intra-subtype identity is high (averaging >92%), even when comparing sequences from different hemispheres. Hence, the genetic structure of the AIV HA is characterized by highly divergent subtypes that harbor relatively little internal genetic diversity. However, 4 subtype comparisons show considerably less divergence (76–79% identity); H4 vs. H14, H7 vs. H15, H13 vs. H16, and H2 vs. H5, indicating that they separated more recently (Figure 1; see below). A similar phylogenetic structure was seen in the NA (Figure 1b and S2), again with evidence for higher-order clustering (e.g., N6 and N9; N1 and N4). In contrast to the HA, however, levels of genetic divergence among the NA types are more uniform, with the 9 subtypes exhibiting an average inter-subtype identity of 43.6% (with an average intra-subtype identity of >89%) and no clear outliers. Hence, no new (detected) NA types have been created in the recent evolutionary past. This correlates with the more uniform distribution of NA than HA subtypes in wild bird AIV isolates [43]. The topology of the NS segment phylogeny was also of note, being characterized by the deep divergence among the A and B alleles as described [44] (Figure S8). Almost every HA and NA subtype of AIV contain both the A and B NS alleles, without evidence of ‘intermediate’ lineages expected under random genetic drift, strongly suggesting that the two alleles are subject to some form of balancing selection. The NS1 protein has pleiotropic functions during infection in mammalian cells, and plays an important role in down-regulating the type I interferon response [45]. Supporting these results are the elevated rates of nonsynonymous to synonymous substitution per site (ratio dN/dS) observed for the NS1 gene in both avian and human influenza viruses [46] suggesting that the NS1 protein has experienced adaptive evolution in both host types. Whether this selection relates to the role the NS1 protein plays in its interaction in the type I interferon pathway is currently unclear. Far less genetic diversity is observed in the 5 remaining AIV gene segments (PB2, PB1, PA, NP, and M - Figures S3, S4, S5, S6 and S7). Indeed, the extent of diversity in these genes is less than that within a single HA or NA subtype, with average pairwise identities ranging from 95–99%. Our phylogenetic analysis also revealed a clear separation of AIV sequences sampled from the Eastern and Western Hemispheres, as previously noted (3,19), indicating that there is relatively little gene flow between overlapping Eastern and Western Hemisphere flyways. However, despite this strong biogeographic split, mixing of hemispheric AIV gene pools clearly occurs at a low level (see below). Abundant reassortment in AIV To assess the frequency and pattern of reassortment in AIV, we compared the extent of topological similarity (congruence) among phylogenetic trees of each internal segment. This analysis revealed a remarkably frequent occurrence of reassortment, supporting previous studies on smaller data sets [37],[47]. For example, 5 H4N6 AIV isolates were recovered from mallards sampled at the same location in Ohio on the same morning and in the same trap (Figure 2). For the internal genes, these viruses contained 4 different genome ‘constellations’, with only 1 pair of viruses sharing the same constellation. In the data set as a whole, the large number of different subtype combinations recovered highlights the frequency of reassortment (Figures 1b and S2), and provides little evidence for the elevated fitness of specific HA/NA combinations in AIV isolates from wild birds. That the majority of HA/NA combinations have been recovered [8],[15] also strongly supports the high frequency of reassortment involving these surface protein genes. 10.1371/journal.ppat.1000076.g002 Figure 2 The genome constellations (internal gene segments only) of 5 H4N6 viruses collected from mallards at the same location in Ohio, USA on the same day. The different colors reflect segments whose sequences fall into different major clades – defined by strong bootstrap support (>80%) – in each internal gene segment tree. For example, all 6 internal gene segments from isolates A/Mall/OH/655/2002 and A/Mall/OH/657/2002 have the same, shared phylogenetic position (shaded red), but exhibit a significantly different phylogenetic pattern, indicative of reassortment, with A/Mall/OH/667/2002 in the PB1 and PA gene segments (individual trees presented in Figures S4 and S5). Similarly, isolate A/Mall/OH/668/2002 shows phylogenetic evidence of reassortment in 5 of 6 internal gene segments compared to A/Mall/OH/655/2002. Thus, while there is strong evidence of frequent reassortment between HA and NA, we also sought to assess the extent of reassortment among the less commonly studied internal gene segments. A maximum likelihood test of phylogenetic congruence [48] revealed that although the topologies of the internal segment trees are more similar to each other than expected by chance, so that the segments are not in complete linkage equilibrium (in which case they would be no more similar in topology than two random trees), the difference among them is extensive, indicative of extremely frequent reassortment and with little clear linkage among specific segments (Figure 3). Of the 6 internal segments, NS exhibited the least linkage to other genes, falling closest to the random distribution (i.e. possessed the greatest phylogenetic incongruence). This is compatible with the deep A and B allelic polymorphism in this segment. In contrast, the M segment showed the greatest phylogenetic similarly, albeit slight, to the other segments. Overall, however, the relationships between segments are better described by their dissimilarity than their congruence. 10.1371/journal.ppat.1000076.g003 Figure 3 Maximum likelihood analysis of congruence among the internal gene segments of 407 isolates of avian influenza virus. Each column represents the difference in log likelihood (Δ-lnL) between the ML trees of each gene (shown by colored dots). In every case, the ML tree estimated for the reference gene has the highest likelihood, while lower likelihoods (greater Δ-lnL values) are observed when the ML trees for the other genes are fitted to the sequence data from the reference gene and branch lengths re-optimized. To assess the extent of similarity in topology among genes, 500 random trees were created for each data set and their likelihoods assessed for each gene in turn using the same procedure (indicated by horizontal bars). In every case, and most notably for NS, the trees inferred for each gene have likelihoods closer to the random set than to the ML tree for the reference gene, indicative of extensive incongruence. Occasional AIV isolates demonstrated hemispheric mixing with reassortment. As reported previously, the majority of such mixing occurs in shorebirds and gulls [36] (with the exception of Eurasian lineage H6 HA genes distributed widely in North American Anseriformes [5] as also revealed in this study). Interestingly, no completely Eurasian-lineage AIV genome has been reported in North America, or vice versa [9],[49]. This suggests that birds initially carrying AIV between the hemispheric flyways have not been identified in surveillance efforts. Most mixed isolates possess only one gene segment derived from the other hemisphere, indicating that there is little or no survival advantage for such hemispheric crossovers in the new gene pool. Since Asian lineage HP H5N1 AIV have been isolated from wild birds in Eurasia [50], concern has been raised over the importation of the virus into North America via migratory birds. Our analyses suggest that enhanced surveillance in gulls and other shorebirds may be warranted, and that with frequent reassortment (see below), entire Asian HP H5N1 AIV isolate genome constellations may not be detected in these surveys. Overall, 25 of 407 (6%) AIV genomes show evidence of hemispheric mixing, with the phylogenies suggesting a general pattern of viral gene flow from Eurasia to North America: 5 North American isolates possessed two Eurasian-lineage internal gene segments, and 20 carried a single segment. North American isolates possessing a Eurasian-lineage M segment were the most common, seen in 18 isolates (Figure S7), followed by 8 with a Eurasian PB2 segment (Figure S3), four with a Eurasian PB1 segment (Figure S4), and 1 with a Eurasian PA segment (Figure S5). The 18 Eurasian M segments and the 8 Eurasian PB2 segments each form monophyletic groups, suggesting single introductions to North America. In each case, sequences from domestic ducks in China and turkeys in Europe were the closest relatives. It is therefore theoretically possible that some of these introductions may have been derived from imported poultry rather than migratory birds. In contrast, 3 of the 4 Eurasian PB1 and the single Eurasian PA segment in North American AIV contained genes whose closest relatives were in viruses found in red-necked stints from Australia. These small waders are widely migratory, with a range from Siberia to Australasia, and occasionally in Europe and North America. Interestingly, 23 of 25 such mixed genomes were observed in shorebirds along the U.S. Atlantic coast. Unfortunately, no complete AIV genomes are available from shorebirds on the U.S. Pacific coast for comparison. The Evolutionary Genetics of AIV In theory, two evolutionary models can explain the global pattern of AIV diversity, analogous to the allopatric and sympatric models of speciation. Under the allopatric model, the HA and NA subtypes correspond to viral lineages that became geographically isolated, resulting in a gradual accumulation of amino acid changes among them. Because of physical separation through geographical divergence, there is no requirement for natural selection to reinforce the partition of HA and NA diversity into discrete subtypes by preferentially favoring mutations at antigenic sites. In contrast, under the sympatric model, the discrete HA and NA subtypes originate within the same spatial population, such that natural selection must have reinforced speciation; subtypes that were too antigenically similar would be selected against because of cross-protective immune responses. Therefore, mutations would accumulate first at key antigenic sites, allowing subtypes to quickly diversify in the absence of herd immunity. The AIV genomic data available here suggest a complex interplay of evolutionary processes. That discrete HA and NA subtypes, as well as the 2 divergent NS alleles, are maintained in the face of frequent reassortment strongly suggests that each represents a peak on a fitness landscape shaped by cross-immunity (Figure 4a). Under this hypothesis, ‘intermediate’ HA/NA/NS alleles would be selected against because they generate more widespread herd immunity, corresponding to fitness valleys. Indeed, it is the likely lack of immunological cross-protection at the subtype level that allows the frequent mixed infections described here (although mixed infections may also occur in young, immunologically naïve birds). Further, in most cases these divergent HA, NA and NS alleles circulate in the same bird species in the same geographical regions, compatible with their divergence under sympatry. In addition, 3 of the most closely related pairs of HA subtypes contain an HA that is rarely isolated or limited geographically or by host species restriction, implying that their dispersion is inhibited by existing immunity; H14 has only been isolated rarely in Southern Russia, H15 only in Australia, and H16 has only been described in gulls. The possible exception is H2–H5, where both subtypes have been isolated from a variety of bird species in a global distribution. Although these may represent more recent occurrences of allopatric speciation, antigenic cross-reactivity between the H2–H5, H7–H15, H4–H14 pairs was recently demonstrated [51], again compatible with the sympatric model. Further support for possible cross-immunity between these subtypes would require experimental challenge studies. 10.1371/journal.ppat.1000076.g004 Figure 4 The fitness landscapes of avian influenza virus. (a) The fitness landscapes observed in HA, NA and NS, and represented here by NA. Each colored cone represents an individual subtype. These subtypes are connected by a bifurcating tree. The lack of ‘intermediate’ subtypes – those falling below the pink disc – reflects major valleys in fitness, such that any virus falling in this area will experience a major reduction in fitness, most likely due to an elevated cross-protective immune response. Occasionally, individual subtypes jump species barriers and spread in new hosts (such as humans), where they experience a continued selection pressure and hence accumulate amino acid substitutions in a progressive manner, as shown. (b) The fitness landscapes observed in the remaining internal protein segments of avian influenza virus – PB2, PB1, PA, NP and M (represented by different colors). In this case, there is little functional difference among the genetic variants of each segment, so that the fitness landscape is flat. This equivalence in fitness among genome constellations also means that reassortment is frequent among them (as reassortants suffer no fitness cost), represented by the horizontal lines connected each internal gene segment. In contrast to the extensive genetic diversity seen in HA, NA and NS, the 5 remaining internal gene segments encode proteins that are highly conserved at the amino acid level, indicating that they are subject to widespread purifying selection. The fitness landscape for these genes is therefore not determined by cross-immunity, but by functional viability, with less selective pressure to fix advantageous mutations (Figure 4b). Further, given such strong conservation of amino acid sequence, large-scale reassortment is permitted as it will normally involve the exchange of functionally equivalent segments, with little impact on overall fitness. These data also suggest that the cross-immunity provided by these proteins is minimal. Together, these global genomic data provide new insight into the different evolutionary dynamics exhibited by influenza A viruses in their natural wild bird hosts and in those viruses stably adapted to novel species (e.g., domestic gallinaceous poultry, horses, swine, and humans). Based on these analyses, we hypothesize that AIV in wild birds exists as a large pool of functionally equivalent, and so often inter-changeable, gene segments that form transient genome constellations, without the strong selective pressure to be maintained as linked genomes. Rather than favoring successive changes in single subtypes, geographic and ecologic partitioning within birds, particularly within the different flyways, coupled with complex patterns of herd immunity, has resulted in an intricate fitness landscape comprising multiple fitness peaks of HA, NA and NS alleles, interspersed by valleys of low fitness which prevent the generation of intermediate forms (Figure 4a). In contrast, stable host switching involves the acquisition of a number of (as yet) poorly characterized mutations [24],[33],[52],[53] that serve to separate an individual, clonally derived influenza virus strain from the large wild bird AIV gene pool. Because adaptation to a new host likely limits the ability of these viruses to return to the wild bird AIV gene pool [24],[54], these emergent viruses must evolve as distinct eight-segment genome configurations within the new host. The ability of recent HP H5N1 AIV to cause spillover infections in wild birds is an unprecedented exception. Further, because humans represent a large and spatially mixed population, natural selection is able to act efficiently on individual subtypes [55]. Hence, a limited number of subtypes circulate within humans and evolve by antigenic drift to escape population immunity. Notably, the recent Asian lineage HP H5N1 AIV strains are intermediate between these two contrasting influenza ecobiologies; a combination of large poultry populations allows natural selection to effectively drive rapid antigenic and genetic change within a single subtype [46],[56], while reassortment with the wild bird AIV gene pool facilitates the generation of new genome constellations [57]–[59]. Similar patterns have also been observed with the widely circulating H9N2 and H6N1 viruses in gallinaceous poultry in Eurasia [60],[61]. Previous analyses have also shown that recent HP H5N1 viruses had the highest evolutionary rates and selection pressures (dN/dS ratios) as compared to other AIV lineages [46]. Consequently, these results underscore the importance of determining the mechanistic basis of how H5N1 has spread so successfully among a diverse range of both domestic and wild bird species. Materials and Methods Sample collection and virus isolation The genomes of 167 influenza A virus isolates recovered from 14 species of wild Anseriformes located in four U.S. states (Alaska, Maryland, Missouri, Ohio) were sequenced for this study; viral isolates consisted of 29 hemagglutinin (HA) and neuraminidase (NA) combinations, including H1N1, H1N6, H1N9, H2N1, H3N1, H3N2, H3N6, H3N8, H4N2, H4N6, H4N8, H5N2, H6N1, H6N2, H6N5, H6N6, H6N8, H7N3, H7N8, H8N4, H10N7, H10N8, H11N1, H11N2, H11N3, H11N6, H11N8, H11N9, H12N5. Cloacal swabs were collected as previously described [39] from 1986–2005 as part of The Ohio State University's ongoing influenza A virus surveillance activities and in collaboration with many researchers in other states since 2001. A table listing the details of each isolate are available from the Influenza Virus Resource page (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/shipment.cgi). Avian influenza viruses were originally isolated using standard viral isolation procedures after 1–2 passages in 10-day-old embryonated chicken eggs (ECEs) [62]. Type A influenza virus was confirmed using commercially available diagnostic assays (Directigen Flu A Assay, Becton Dickinson Microbiology Systems, Cockeysville, MD) and isolates were subtyped at the National Veterinary Services Laboratories (NVSL), Animal and Plant Health Inspection Service, United States Department of Agriculture, Ames, Iowa, using standard hemagglutinin inhibition and neuraminidase inhibition testing procedures [51]. Sequential Limiting Dilutions Isolates for this investigation were generally selected from the viral archives based on antigenic diversity, clustering of recoveries, no evidence of antigenically mixed subtypes, and distribution over time. First- or second-egg-passage isolates in chorioallantoic fluid (CAF) were rapidly thawed from −80°C to room temperature, vortexed for 30 seconds and centrifuged at 1500 rpm for 10 minutes. Approximately 0.5 ml of CAF was drawn from the vial using a 26-gauge needle and subsequently passed through a 25 mm, 0.2 µm filter. Following filtration, a 10−1 CAF stock dilution was obtained by adding 0.2 ml filtered CAF to 1.8 ml Brain Heart Infusion Broth containing penicillin and streptomycin and vortexed for 30 seconds. Serial dilutions (10−6 maximum) were performed and 0.1 ml of each dilution was inoculated into each of four 10-day-old ECEs. After approximately 48 hours of incubation at 35°C/60% humidity, the inoculated eggs were chilled overnight and CAF was harvested from each egg and tested for hemagglutinating activity. The CAF from the last dilution positive for hemagglutinating activity was tested for the presence of type A influenza virus using the Directigen Flu A or Synbiotics Flu Detect Antigen Capture Test Strips™ (Synbiotics Corp., San Diego, CA). Hemagglutination titer assays were performed and CAF aliquots from the most dilute influenza A positive samples were stored at −80°C. If no endpoint titer was determined, the 10−6 CAF dilution was stored at −80°C and the procedure repeated utilizing 10−4 to 10−9 sequential dilutions. Preliminary molecular testing Viral RNA was isolated from allantoic fluid using Trizol® Reagent (Invitrogen Corp., Carlsbad, CA) and transcribed into 20 µl of cDNA for a subset of samples [63]. Segment-specific universal primers designed to amplify partial and/or full-segments were initially used in RT-PCR assays to assess vRNA quality and RT-PCR primer specificity and sensitivity. Additionally, M13 sequencing tags (F primer: GTAAAACGACGGCCAG; R primer: CAGGAAACAGCTATGAC) were added to each primer set for ease of sequencing RT-PCR products in both forward and reverse directions. Primer design For initiation of a high-throughput sequencing pipeline, a universal strategy for primer design was employed to ensure detection of multiple viral infections within a single sample. Primers were designed to semi-conserved areas of the six internal segments. For the segments encoding the external proteins, primers were designed from alignments of subsets of the 16 HA and 9 NA avian subtypes. Alignments were generated with MUSCLE [64] and visualized with BioEdit [65]. An M13 sequence tag was added to the 5′ end of each primer to be used for sequencing. Four sequencing reactions per run were analyzed on an agarose gel for quality control purposes. The sequence success rate of each primer pair was analyzed relative to the HA and NA subtype. Primers that did not perform well were altered or replaced. All primers and RT-PCR assay cycling conditions are available upon request. cDNA Synthesis and Sequencing Influenza A virus isolates were amplified with the OneStep RT-PCR kit (Qiagen, Inc., Valencia, CA). Amplicons were sequenced in both the forward and reverse directions. Each amplicon was sequenced from each end using M13 primers (F primer: TGTAAAACGACGGCCAGT; R primer: CAGGAAACAGCTATGACC). Sequencing reactions were performed using Big Dye Terminator chemistry (Applied Biosystems, Foster City, CA) with 2 µl of template cDNA. Additional RT-PCR and sequencing was performed to close gaps and to increase coverage in low coverage or ambiguous regions. Sequencing reactions were analyzed on a 3730 ABI sequencer and sequences were assembled in a software pipeline developed specifically for this project. Sequence trimming and assembly Once genomic sequence was obtained for an individual sample, reads for each segment were downloaded, trimmed to remove amplicon primer-linker sequence and low quality sequence, and assembled. A small genome assembly suite called Elvira (http://elvira.sourceforge.net/), based on the open-source Minimus assembler, was developed to automate these tasks. The Elvira software delivers exceptions including failed reads, failed amplicons, and insufficient coverage to a reference sequence (as obtained from GenBank), ambiguous consensus sequence calls, and low coverage areas. The avian influenza A sequences (with GenBank Accession numbers) produced from this ongoing study are available at http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/shipment.cgi. The first 167 avian influenza genomes from this collection were submitted to GenBank and included in this study. Evolutionary analysis The genomes of avian influenza virus newly determined here were combined with those already available on GenBank, particularly from recent large-scale surveys of viral biodiversity [38]. Sequences from viruses isolated before 1970, which may have been subjected to extensive laboratory passage, were excluded as were the large numbers of H5N1 sequences collected in recent years (a sample of H5N1 genomes, 1997–2005, were included for analysis). In total, 452 HA sequences and 473 NA sequences were used in analyses. For the internal protein-encoding segments (PB2, PB1, PA, NP, M, NS), a total of 407 genomes were analyzed (by considering a common data set we were able to investigate patterns of segment linkage, see below). For each data set, sequence alignments of the coding regions were created using MUSCLE [64] and adjusted manually using Se-Al [66] according to their amino acid sequence. In the case of HA and NA, some regions of the inter-subtype sequence alignment were extremely divergent such that they could not be aligned with certainty (HA signal peptide and cleavage site insertions in HPAI H5 or H7, and variable small stalk deletions in NA). Because of their potential to generate phylogenetic error, these small regions of ambiguity were deleted. This resulted in the following sequence alignments used for evolutionary analysis: PB2 = 2277 nt; PB1 = 2271 nt; PA = 2148 nt; HA = 1683 nt; NP = 1494 nt; NA = 1257 nt; M = 979 nt; NS = 835 nt. All sequence alignments are available from the authors on request. Nucleotide and amino acid identity was calculated using Megalign (Lasergene 7.2, DNAStar, Madison, WI). Using these alignments, maximum likelihood (ML) trees were inferred using PAUP* [67], based on the best-fit models of nucleotide substitution models determined by MODELTEST [68]. In most cases, the preferred model of nucleotide substitution was GTR+I+Γ4, or a close relative. For each of these trees, the reliability of all phylogenetic groupings was determined through a bootstrap resampling analysis (1000 pseudo-replicates of neighbor-joining trees estimated under the ML substitution model). We employed a maximum likelihood method to assess the extent of phylogenetic congruence, indicative of reassortment [48]. To reduce any bias in phylogenetic structure caused by geographic segregation, only isolates from North American flyways were used in analyses of the internal gene segments. Briefly, ML trees for each internal gene segment were estimated as described above. Next, the log likelihood (-LnL) of each of the ML trees was estimated on each gene segment data set in turn, optimizing branch lengths under the ML substitution model in every case. The topological similarity between each gene segment tree on each data set was then determined by compared the difference in likelihood among them (Δ-LnL). Clearly, the greater the similarity in topology (congruence) among the trees for each segment, the closer their likelihood scores and so the more likely they are to be linked. To put the distribution of Δ-LnL values in context, we constructed 500 random trees for each data set and optimized their branch lengths in the same manner. If any of the Δ-LnL values among the ML trees falls within the random distribution then we can conclude that the gene segments in question are in complete linkage equilibrium. All these analyses were conducted using PAUP* package [67]. Supporting Information Figure S1 Maximum likelihood tree of the HA gene of 452 isolates of avian influenza A virus, including representatives of all 16 subtypes. Sequences are color-coded according to HA subtype (see Figure 1). Internal branches are color-coded to reflect the flyway from which the viruses were sampled; North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to the relevant branches. Branch lengths are scaled according to the number of nucleotide substitutions per site. (1.16 MB EPS) Click here for additional data file. Figure S2 Maximum likelihood tree of the NA gene of 473 isolates of avian influenza A virus, including representatives of all 9 subtypes. Sequences are color-coded according to HA subtype (see Figure 1), with the mix of colors highlighting the frequency of reassortment. Internal branches are color-coded to reflect the flyway from which the viruses were sampled; North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to the relevant branches. Branch lengths are scaled according to the number of nucleotide substitutions per site. (1.02 MB EPS) Click here for additional data file. Figure S3 Maximum likelihood tree of the PB2 gene of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.84 MB EPS) Click here for additional data file. Figure S4 Maximum likelihood tree of the PB1 gene of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.84 MB EPS) Click here for additional data file. Figure S5 Maximum likelihood tree of the PA gene of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.83 MB EPS) Click here for additional data file. Figure S6 Maximum likelihood tree of the NP gene of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.79 MB EPS) Click here for additional data file. Figure S7 Maximum likelihood tree of the M genes of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.79 MB EPS) Click here for additional data file. Figure S8 Maximum likelihood tree of the NS genes of avian influenza A viruses. Sequences are color-coded according to HA subtype. Internal branches are color-coded to reflect the flyway from which the viruses were samples: North American flyway in red, Eurasian flyway in blue. Bootstrap values above 70% are shown next to relevant branches. (0.83 MB EPS) Click here for additional data file. Table S1 Sequencing results for 167 complete genomes of 29 subtypes of avian influenza A viruses. (0.06 MB DOC) Click here for additional data file.
- Record: found
- Abstract: found
- Article: found
H5N1 Outbreaks and Enzootic Influenza
Robert G. Webster, Malik Peiris, Honglin Chen … (2006)
Influenza is an ancient disease that has infected humans at irregular intervals throughout recorded history ( 1 ). While the 1918 "Spanish" influenza is the best recorded catastrophic influenza pandemic, similarly severe pandemics occurred earlier, when the human population of the world was much smaller, and they will occur again. Our challenge is to understand all aspects of the influenza virus, the hosts and their response, and the virus' global impact so that we may be better prepared to face the inevitable next influenza pandemic. The influenza virus that appears most threatening is the avian H5N1 strain that since 2003 has infected >130 persons in Vietnam, Thailand, and Cambodia and has killed more than half of them. Nonetheless, the H5N1 influenza threat is viewed with disturbing complacency; a frequently heard statement is "since the virus has not adapted to continuing human-to-human transmission by now, it is unlikely to do so in the future." Such complacency is akin to living on a geologic fault line and failing to take precautions against earthquakes and tsunamis. The Source Influenza A viruses are perpetuated in the wild birds of the world, predominantly in waterfowl, in which the 16 subtypes (which differ by 30% in their hemagglutinin [HA] nucleotide homology) coexist in perfect harmony with their hosts ( 2 , 3 ) (Figure 1). In these natural hosts, the viruses remain in evolutionary stasis, showing minimal evolution at the amino acid level over extended periods. This fact indicates that the influenza-bird association is ancient; this lack of change is surprising because influenza viruses are segmented, negative-stranded RNA viruses that have no quality-control mechanisms during replication and are highly prone to variation. After transfer to a new type of host, either avian or mammalian, influenza viruses undergo rapid evolution. However, all 16 HA subtypes, including H5 and H7, have until recently been considered to be benign in their natural hosts. This benign equilibrium between the influenza virus and its host may have changed. Figure 1 Emergence of H5N1 influenza virus and control options. A nonpathogenic H5 influenza virus is believed to have spread to domestic ducks and geese, then to domestic chickens. In chickens, the H5 virus became highly pathogenic before it was transferred back to domestic ducks and geese. The highly pathogenic H5 virus reassorted its genome with those of other influenza viruses in aquatic birds, and the resulting viruses spread to domestic poultry farms, humans, and occasionally to pigs. These viruses acquired mutations in their PB2, HA, NA, and NS genes that made them lethal to domestic and wild waterfowl and humans. Solid lines, transmission demonstrated; dotted lines, transmission postulated but not demonstrated. Multiple opportunities exist for control of highly pathogenic avian influenza: 1) prevent contact between wild and domestic poultry by use of screened poultry houses and treated water; 2) prevent contact between domestic waterfowl and gallinaceous poultry by use of screened houses and treated water and by exclusion of waterfowl from "wet markets"; 3) eradicate H5/H7 influenza viruses from gallinaceous poultry by culling or the use of vaccines that prevent disease and transmission; 4) prevent contact between poultry, pigs, and humans and make vaccines and antiviral drugs available. Genesis of the H5N1 Virus Before 1997, no evidence had indicated that H5 influenza viruses could infect humans and cause fatal disease. The H7 influenza viruses were known to cause conjunctivitis in humans, and serologic studies provided evidence of subclinical human infection with the subtypes prevalent in avian live poultry markets ( 4 ). The precursor of the H5N1 influenza virus that spread to humans in 1997 was first detected in Guangdong, China, in 1996, when it caused a moderate number of deaths in geese and attracted very little attention ( 5 ). This goose virus acquired internal gene segments from influenza viruses later found in quail (A/Quail/HK/G1/97 [H9N2]) and also acquired the neuraminidase gene segment from a duck virus (A/Teal/HK/W312/97 [H6N1]) before the goose virus became widespread in live poultry markets in Hong Kong and killed 6 of 18 infected persons ( 6 , 7 ). This H5N1 virus was eradicated by culling all domestic poultry in Hong Kong, and the genotype has not been detected since that time. However, different reassortants continued to emerge from goose and duck reservoirs ( 8 ) that contained the same H5 HA glycoprotein but had various internal genes. The H5N1 viruses continued to evolve, and in late 2002, a single genotype was responsible for killing most wild, domestic, and exotic waterfowl in Hong Kong nature parks ( 9 , 10 ). This genotype of H5N1 spread to humans in Hong Kong in February 2002, killing 1 of 2 infected persons ( 11 ), and was the precursor of the Z genotype that became dominant. The Z genotype spread in an unprecedented fashion across Southeast Asia, affecting Vietnam, Thailand, Indonesia, Cambodia, Laos, Korea, Japan, China, and later Malaysia. Further analysis showed that the H5N1 influenza viruses that caused outbreaks in poultry in Japan and Korea were genetically different from those in the other countries (the V genotype) ( 12 , 13 ). The phylogeny of the recent Z genotype viruses showed that viruses isolated in Vietnam and Thailand formed a cluster that remained distinct from those isolated in Indonesia. To date, >140 million domesticated birds have been killed by the virus or culled to stem its spread; as of December 2005, >130 persons have been infected in Vietnam, Thailand, Indonesia, Cambodia, and China, and 70 have died (42 in Vietnam, 14 in Thailand, 8 in Indonesia, 4 in Cambodia, and 2 in China). These recent H5N1 influenza viruses have acquired the unprecedented and disturbing capability to infect humans; to cause neurotropic disease and a high proportion of deaths in waterfowl in nature; to cause death in and be transmitted among felid species, including domestic cats ( 14 ); and to cause neurotropic disease and death in ferrets and mice ( 15 ). These incremental changes intensify concern about this H5N1 virus' pandemic potential. These traits are likely to have been acquired initially by reassortment in 2001 and 2002, when a plethora of different genotypes were detected in poultry markets and later in farms in Hong Kong ( 13 ). These genes were presumably acquired from viruses found in waterfowl in Southeast Asia, but the actual gene donors have not yet been identified. Since late 2002, the Z genotype has become dominant, but phylogenetically distinguishable viruses have continued to cocirculate in Indonesia and western China. These characteristics have been acquired mainly through mutations in the RNA polymerase (PB2) gene, insertions in the HA gene, and deletions in the NA and nonstructural (NS) genes. Thus, the H5N1 viruses continue to evolve, initially by reassortment and more recently by mutation and deletion ( 16 , 17 ). While most H5N1 influenza viruses isolated from avian species in Asia since 1997 are highly pathogenic in gallinaceous poultry, they show heterogeneous pathogenicity in other species. In domestic ducks, the pathogenicity of the H5N1 viruses varies from high to nonpathogenic. In ferrets, most avian isolates replicate and cause respiratory tract infection, while a few strains are highly pathogenic and neurotropic (causing hind leg paralysis), and the virus has been isolated from the brain ( 15 ). In contrast, all isolates from humans are highly pathogenic to ferrets. A similar pattern is found in experimental infection of mice, in which most avian isolates cause respiratory infection. Mechanisms of Spread Were the highly pathogenic H5N1 viruses transferred within and between countries by persons, poultry, or fomites? In previous outbreaks of highly pathogenic H5 and H7 infection in multiple countries, the spread was directly attributable to humans. The main way influenza virus is spread in poultry is by movement of poultry and poultry products; establishing good biosecurity measures on poultry farms is therefore an important defense. The poultry industry is a huge, integrated complex in Asia, and a number of firms have branches in China, Vietnam, Thailand, and Indonesia. Nonetheless, the involvement of multiple lineages of H5N1 argues against human-mediated spread from a single source. Live poultry markets are an amplifier and reservoir of infection ( 18 ) and probably play a role in the maintenance and spread of the virus in the region. However, a number of other factors unique to affected Asian countries make control difficult. Backyard flocks are common in the region, and these domesticated birds are not subject to any biosecurity measures. Fighting cocks are prized possessions and are often transported long distances. Fighting cocks may also play a role in the spread of infection and in transmission to humans. Many of the affected countries have a weak veterinary infrastructure and are facing highly pathogenic avian influenza outbreaks for the first time. The migrant ducks that commonly wander through rice fields scavenging fallen rice seeds are another potent mechanism for the spread of infection. Role of Domestic Ducks After late 2002, when H5N1 viruses had killed waterfowl in Kowloon Park in Hong Kong, most avian H5N1 isolates isolated in Vietnam, Thailand, and Indonesia were highly pathogenic to chickens and domestic ducks. However, by late 2003 and early 2004, some avian isolates were nonpathogenic to ducks but retained their pathogenicity to chickens ( 19 ). Genetic analysis of these isolates showed evidence of multiple variants within single specimens ( 20 ). On Madin-Darby canine kidney (MDCK) cells, these viruses formed a mixture of small and large plaques that had different biologic properties. Viruses that formed large plaques were usually highly pathogenic to ducks and ferrets, whereas viruses that formed small plaques were usually nonpathogenic to both birds and ferrets. Some virus isolates formed small plaques that were pathogenic to ducks. Thus, plaque size was not a marker of pathogenicity. When ducks were orally infected with the original mixed population of H5N1 viruses, most birds died, but some excreted virus for an extended period (up to 17 days); during this time, viruses that were nonpathogenic to ducks were selected. Serologic testing of these ducks showed hemagglutination inhibition (HI) and neutralizing antibodies against the original dominant virus in the mixture; thus, immune clearance had caused the selection of the minor variants. The viruses shed on day 17 had become nonpathogenic to ducks, although they remained highly pathogenic to chickens. Sequence analysis of the HA showed that these viruses differed from the original dominant virus at multiple amino acids and were antigenically distinguishable in HI tests. Therefore, H5N1 viruses circulating in avian populations in Southeast Asia are clearly heterogeneous. Notably, this phenomenon has repeatedly been reported for other influenza viruses that are in the process of altering their interspecies transmission, including European avian H1N1 viruses that were transmitted to pigs ( 21 ), H9N2 viruses that were transmitted to pigs and humans, and now H5N1 viruses that are transmitted from ducks to humans. How these mixtures of codominant viruses are generated in a quasispecies is unresolved. Suggested mechanisms include mutator mutations or partial heterozygotes, but a satisfactory explanation is not available ( 22 ). A subdominant population of H5N1 viruses is presumably selected in ducks after the immune response clears the dominant virus. The subdominant population appears to be uniformly nonpathogenic to ducks, as if this is the natural situation for influenza in the duck. Whether further selection will occur against the polybasic cleavage site in the HA and the pathogenicity-determining sites in PB2 and NS remains to be seen. These viruses' loss of pathogenicity to ducks, but retention of pathogenicity to chickens and presumably to humans, has been a problem associated with their eradication. In Vietnam, for example, disease signs were used as the criteria for identifying H5N1 infection in ducks. Thus, the duck has become the Trojan horse of highly pathogenic H5N1 influenza in Asia ( 20 ). Role of Migratory Birds Migratory waterfowl are generally believed to be the main reservoir of all 16 subtypes of influenza A viruses, including H5 and H7 subtypes. However, less agreement is found regarding the role of migratory waterfowl in the initial spread of highly pathogenic H5N1 viruses across eastern Asia in 2003. The isolation of highly pathogenic H5N1 from herons, egrets, and peregrine falcons in Hong Kong in 2003 and 2004 leaves no doubt that wild migratory birds can be infected and may spread disease to local poultry flocks. The outbreak in Qinghai Lake ( 16 , 17 ) proves that these highly pathogenic H5N1 influenza viruses are transmissible among migratory waterfowl. The migration route of shorebirds in the east Asian-Australasian flyway does overlap the areas that have had H5N1 outbreaks, although the virus has been notably absent in Taiwan, Malaysia (except for occasional outbreaks near the Thai border), and western Australia (Figure 2). The role of migratory birds in the transmission and spread of highly pathogenic H5N1 viruses is still unclear. However, the recent outbreak of H5N1 infection in bar-headed geese and other species in Qinghai Lake is a cause for concern because these birds migrate southward to the Indian subcontinent, an area that has apparently not been affected by H5N1 avian influenza. If the virus were to become entrenched in India, its geographic range would be substantially extended, and the pandemic threat would increase accordingly ( 17 ). A mutation in the PB2 gene (residue E627K) associated with pathogenicity in mammals ( 16 , 17 ) has been found in viruses isolated from birds in Qinghai Lake; this finding has caused concern that this mutation will be transferred to other migratory birds (e.g., wild ducks) and will be spread because not all infected birds die. Figure 2 igration routes of Asian birds. A) Distribution and migration routes of bar-headed geese (courtesy of P. Leader). B) The Asia-Pacific region contains >240 species of migratory birds. The 3 flyways run primarily in a north-south direction, overlapping and extending from Australia/New Zealand to India, Central Asia, and Siberia. The outbreak of highly pathogenic (HP) H5N1 in migratory waterfowl at Qinghai Lake, China, affected primarily bar-headed geese (Anser indicus); however, other species, including gulls and ducks, were affected ( 16 , 17 ). The outbreak started in early May 2005, and by June >5,000 birds had died. The birds exhibited neurologic signs, inability to stand, diarrhea, and death. Systemic infection was detected in all organs tested. C) Bar-headed goose infected with HP H5N1 influenza virus. D) Immunostain of goose pancreas, using H5 monoclonal antibodies (magnification ×400). (C and D, courtesy of H. Chen). Countries shown in red have had outbreaks of HP H5N1 since 2004. The geographic range of H5N1 may be extended by bar-headed geese or by ducks that are less susceptible to lethal infection. Although culling domestic poultry to contain the spread of highly pathogenic H5N1 virus is considered an acceptable agricultural practice, culling migratory birds is not acceptable to any international authority (Food and Agriculture Organization of the United Nations [FAO], the World Organization for Animal Health [OIE], the World Health Organization [WHO]). The idea of culling migratory birds must be strongly discouraged, for it could have unknown ecologic consequences. Instead, since highly pathogenic H5N1 has been demonstrated in migratory birds, the poultry industries of the world must adapt measures such as increased biosecurity (Figure 1), the use of vaccines, or both. Early detection and aggressive control measures allowed Japan, South Korea, and Malaysia to eradicate H5N1 virus soon after its introduction into those countries' poultry flocks, demonstrating that rapid and determined responses can keep the virus from gaining a foothold. In other countries in Asia, delayed detection and response caused the virus to become entrenched across a wide region, and eradication at this stage has become a formidable undertaking. Agricultural Vaccines The need for H5N1 vaccines for domestic poultry is increasing. Adopting a policy to use vaccines in poultry is an important decision for agricultural authorities in countries such as Thailand (a major poultry exporter) and Vietnam. Both countries are investigating their specific needs. While considerable data exist on the efficacy of influenza vaccines in domestic chickens, little comparable information is available regarding ducks. The pros and cons of the use of vaccines in poultry have been reviewed ( 23 ). Current technologies permit discrimination between vaccinated and naturally infected birds; however, vaccines are not standardized on the basis of antigen content. "Good" and "bad" agricultural vaccines are in use. Good Agricultural Vaccines Good agricultural vaccines provide protection from disease despite lack of a close antigenic match between the vaccine and circulating strain and reduce the virus load below the level of transmissibility. They do not provide sterilizing immunity: vaccinated birds may excrete low levels of virus after challenge infection. Sentinel unvaccinated birds are kept in each house to monitor for virus shedding, antigenic drift, or both. Bad Agricultural Vaccines Bad agricultural vaccines prevent disease signs but do not prevent shedding of transmissible levels of virus. They also promote undetected spread of virus on farms and to live poultry markets and promote antigenic drift. China and Indonesia have adopted poultry vaccination to control H5N1, and Vietnam has begun vaccine trials in poultry. However, the resurgence of H5N1 in Indonesian poultry and pigs ( 24 ) and the detection of H5N1 in apparently healthy birds in live poultry markets in China ( 17 ) suggest that some vaccines are of suboptimal quality or that coinfection masks disease. The adoption of a vaccine strategy for H5N2 virus in Mexico in the 1980s reduced disease signs but has not eliminated the H5N2 virus from the region; instead, vaccination may have contributed to the virus' widespread presence in Central America and to its antigenic drift ( 25 ). H9N2 and Cross-protection The clinical signs of infection with highly pathogenic H5N1 virus may be masked by cross-protection by other influenza subtypes, but this fact is often overlooked. During the initial outbreak of highly pathogenic H5N1 in Hong Kong in 1997, chickens in the live poultry markets exhibited no disease signs, yet samples from apparently healthy chickens, ducks, and quail showed highly pathogenic H5N1 in each of the poultry markets surveyed ( 26 ). Surveillance showed that multiple influenza subtypes were cocirculating, including 2 lineages of H9N2, the first represented by the G1 lineage (A/Quail/Hong Kong/G1/97 [H9N2]) and the other by G9 (A/Chicken/Hong Kong/G9/97 [H9N2]). The G1 lineage has the same 6 internal gene segments as the index H5N1 human isolate (A/Hong Kong/156/97 [H5N1]) and is believed to have been the donor of these genes during reassortment that produced the original H5N1 human strain in 1997 ( 27 ). In laboratory studies, chickens previously infected with H9N2 (A/Quail/Hong Kong/G1/97 [H9N2]) were protected from disease signs and death when challenged with highly pathogenic H5N1, but the chickens shed H5N1 virus in their feces ( 28 ). Further studies in inbred chickens established that the cross-protection was due to cell-mediated immunity and that it could be transferred by CD8+ T cells but not by antibodies ( 29 ). The possible effect of cocirculating influenza viruses on the pathogenicity of highly pathogenic H5N1 in Vietnam, Thailand, and elsewhere in Asia has not been resolved. To date, no other subtypes of influenza A viruses have been reported in poultry in Vietnam or Thailand. Surveillance of live poultry in Hong Kong and in Nanchang ( 30 ) suggests that other influenza A viruses are cocirculating in live poultry markets and on duck farms. Definitive information is required to understand the ecology of influenza and the possible masking of disease signs caused by H5N1. Conclusion Conventional wisdom about pandemic influenza holds that a pandemic is inevitable and that the only question remaining is "When?" The H5N1 virus continues to evolve and spread, with additional human infections occurring in Vietnam, Cambodia, Indonesia, China, and Thailand. If this virus acquires human-to-human transmissibility with its present fatality rate of 50%, the resulting pandemic would be akin to a global tsunami. If it killed those infected at even a fraction of this rate, the results would be catastrophic. While the high pathogenicity of the Qinghai bar-headed goose isolate is a continuing threat to poultry and humans, perhaps the most insidious threat comes from unobserved transmission through wild and domestic ducks. The isolation of H5N1 virus from bar-headed geese in Qinghai Lake in southern China in 2005 originated from unobserved infection in poultry markets and suggests that highly pathogenic H5N1 viruses continue to circulate unseen among poultry in China ( 17 ). We cannot afford simply to hope that human-to-human spread of H5N1 will not happen and that, if it does, the pathogenicity of the virus will attenuate. Notably, the precursor of the severe acute respiratory syndrome (SARS)–associated coronavirus ( 31 ) repeatedly crossed species barriers, probably for many years, before it finally acquired the capacity for human-to-human transmission, and its pathogenicity to humans was not attenuated. We cannot wait and allow nature to take its course. SARS was interrupted by early case detection and isolation, but influenza is transmissible early in the course of the disease and cannot be controlled by similar means. Just 1 year before the catastrophic tsunami of December 2004, Asian leaders rejected a proposed tsunami warning system for the Indian Ocean because it was too expensive and the risk was too remote. This mistake must not be repeated in relation to an H5N1 avian influenza pandemic. We must use this window of opportunity to prepare and to begin prepandemic implementation of prevention and control measures.1
Author and article information
Contributors
Neil Ferguson: Role: Editor
Journal
Journal ID (nlm-ta): PLoS Pathog
Journal ID (iso-abbrev): PLoS Pathog
Journal ID (publisher-id): plos
Journal ID (pmc): plospath
Title:
PLoS Pathogens
Publisher:
Public Library of Science
(San Francisco, USA
)
ISSN
(Print):
1553-7366
ISSN
(Electronic):
1553-7374
Publication date Collection:
June
2013
Publication date
(Print):
June
2013
Publication date
(Electronic):
20
June
2013
Volume: 9
Issue: 6
Electronic Location Identifier: e1003443
Affiliations
[1
]Centre for Ecology and Evolution in Microbial Model Systems (EEMiS), Linnæus University,
Kalmar, Sweden
[2
]International Research Center in Agriculture for Development (CIRAD)–UPR AGIRs, Animal
and Integrate Risk Management, Campus International de Baillarguet, Montpellier, France
[3
]Karolinska Institutet, Department of Microbiology, Tumor and Cell Biology (MTC), Stockholm,
Sweden
[5
]Laboratory of Virology, Division of Intramural Research, National Institute of Allergy
and Infectious Diseases, National Institutes of Health, Hamilton, Montana, United
States of America
Author notes
* E-mail:
jonas.waldenstrom@
123456lnu.se
Conceived and designed the experiments: N. Latorre-Margalef, V. Grosbois, V.J. Munster, C. Tolf, R.A.M. Fouchier, B. Olsen, J. Waldenström. Performed the experiments: N. Latorre-Margalef, V. Grosbois, J. Wahlgren, V.J. Munster, C. Tolf, R.A.M. Fouchier, A.D.M.E. Osterhaus, B. Olsen, J. Waldenström. Analyzed the data: N. Latorre-Margalef, V. Grosbois, J. Wahlgren, V.J. Munster, C. Tolf, R.A.M. Fouchier, A.D.M.E. Osterhaus, B. Olsen, J. Waldenström. Contributed reagents/materials/analysis tools: N. Latorre-Margalef, V. Grosbois, J. Wahlgren, V.J. Munster, C. Tolf, R.A.M. Fouchier, A.D.M.E. Osterhaus, B. Olsen, J. Waldenström. Wrote the paper: N. Latorre-Margalef, V. Grosbois, J. Wahlgren, V.J. Munster, C. Tolf, R.A.M. Fouchier, A.D.M.E. Osterhaus, B. Olsen, J. Waldenström.
Article
Publisher ID: PPATHOGENS-D-12-01991DOI: 10.1371/journal.ppat.1003443
PMC ID: 3688562
PubMed ID: 23818849
SO-VID: eeb0ad61-affb-417a-a7a1-02a783c89225
Copyright statement:
Copyright @
2013
License:
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
History
Date
received
: 24
July
2012
Date
accepted
: 6
May
2013
Page count
Pages: 12
Funding
This study was supported by grants from The Swedish Environmental Protection Agency
(V-124-01 and V-98-04), The Swedish Research Council (2008-58, 2010-3067, 2011-48),
The Swedish Research Council Formas (2007-297, 2009-1220) and the Sparbanksstiftelsen
Kronan. The surveillance at Ottenby was part of the European Union wild bird surveillance
and has received support from the Swedish Board of Agriculture and from the EU NP6-funded
New Flubird project. V.J. Munster was partially funded by the Division of Intramural
Research of the National Institute of Allergy and Infectious Diseases, National Institutes
of Health. The funders had no role in study design, data collection and analysis,
decision to publish, or preparation of the manuscript.
Categories
Subject:
Research Article
Subject:
Biology
Subject:
Immunology
Subject:
Immune Response
Subject:
Microbiology
Subject:
Virology
Subject:
Co-Infections
Subject:
Viral Evolution
Subject:
Veterinary Science
Subject:
Animal Types
Subject:
Wildlife
Subject:
Veterinary Diseases
Subject:
Zoonotic Diseases
Subject:
Animal Influenza
Subject:
Veterinary Virology
ScienceOpen disciplines: Infectious disease & Microbiology
Data availability:
ScienceOpen disciplines: Infectious disease & Microbiology