Extensive Mammalian Ancestry of Pandemic (H1N1) 2009 Virus

We demonstrate that the novel pandemic influenza (H1N1) viruses have human virus–like receptor specificity and can no longer replicate in aquatic waterfowl, their historic natural reservoir. The biological properties of these viruses are consistent with those of their phylogenetic progenitors, indicating longstanding adaptation to mammals.

Pandemic (H1N1) 2009 virus strains were recently reported to be reassortants of the North American and European swine lineages (6). Phylogenetic evidence suggests that this reassortment event occurred 10-17 years ago (7). These data suggest that the current pandemic (H1N1) 2009 virus strains should have receptor specifi city typically found in the HA of mammalian viruses (Neu5Acα2,6Gal). In addition, they may have lost the ability to replicate in avian hosts, the natural reservoir species. To test these hypotheses, we examined the biological properties of pandemic (H1N1) 2009 virus, including receptor specifi city, erythrocyte binding, and ability to replicate in avian species.

The Study
We fi rst tested species-specifi c erythrocyte agglutination by the pandemic (H1N1) 2009 isolates A/ California/04/2009 and A/Tennessee/1-560/2009 and by other isolates from humans, swine, and birds ( Table 1). The pandemic (H1N1) 2009 isolates showed reduced or absent agglutination of goose and chicken erythrocytes. Human and swine H1N1 viruses were agglutinated by turkey, guinea pig, chicken, and goose erythrocytes, and all erythrocytes we tested except those of swine were agglutinated by avian isolates (Table 1). We next measured the receptor binding of the 2 pandemic (H1N1) 2009 isolates to sialic substrates, both natural (fetuin) and synthetic (3′-sialyllactose [3′SL] and 6′-sialyllactosamine [6′SLN] attached to a polyacrylic carrier) (Figure). The binding pattern to fetuin was identical among all isolates tested (association constant K ass ≈ 5.8 ± 0.5, 1/μM sialic acid). The currently circulating human and pandemic infl uenza (H1N1) viruses showed a preference for 6′SLN and negligible binding to the avian-type 3′SL. A similar pattern was observed for 2 recent swine viruses, which bound only to 6′SLN receptors with nearly equal affi nity as pandemic (H1N1) 2009 isolates. As expected, the 2 avian H1 viruses bound strongly only to 3′SL (Figure).
To assess the infectivity and pathogenicity of pandemic (H1N1) 2009 virus strain A/California/04/2009 in terrestrial (chickens, quails) and aquatic (domestic and wild ducks) avian species, we inoculated 10 birds of each species by intranasal, intraocular, and intratracheal instillation with ≈10 6.0 of the 50% egg infectious dose (EID 50 ) of the virus. We then observed the birds for the next 2 weeks for death and for viral shedding and signs of illness. No birds showed obvious clinical signs of disease. Virus was detected only on postinoculation day 1 in infected chickens and ducks and only in tracheal samples at low titers (<1.7 log 10 of the EID 50 /mL [8]) (Table 2). However, no later shedding of virus was observed, indicating that the virus detected on postinoculation day 1 could have been caused by residual virus particles after inoculation. In contrast, our results revealed that the A/California/04/2009 strain efficiently infected quails with signifi cantly higher titers (<3.4 log 10 EID 50 /mL until postinoculation day 5; p<0.05) in both oropharyngeal and cloacal swab specimens ( Table 2). The virus was detected in the trachea (1.7 log 10 EID 50 /g), lungs (2.3 log 10 EID 50 /g), and cecal tonsil (0.8 log 10 EID 50 /g) of quails on postinoculation day 5.
The potential bird-to-bird intraspecies transmission of the A/California/04/2009 pandemic (H1N1) 2009 virus strain in avian species was also examined by introducing 3 contact birds to the inoculated birds' cages on postinoculation day 1. There was no subsequent evidence of viral shedding through the upper respiratory tract or fecal-oral route in any group of birds except 1 of 3 contact quails ( Table  2). Oropharyngeal virus titers in this quail were l.7 and 1.5 log 10 EID 50 /mL on postinoculation days 3 and 5, indicating that productive viral replication was occurring.

Conclusions
The A/California/04/2009 pandemic (H1N1) 2009 virus strain showed minimal replication and no transmission in chickens and ducks (domestic and wild), but the virus replicated and had limited transmissibility in quails. Our fi nding is consistent with those of Swayne et al. (9). The inability of the virus to replicate effi ciently in chickens and ducks could very well be linked to its human virus-like receptor recognition.
The ability of infl uenza A viruses to agglutinate erythrocytes from a variety of hosts may refl ect the viruses' receptor specifi city (10,11). We observed similar binding patterns for the mammalian infl uenza (H1N1) viruses, with the exception that the pandemic strains had reduced binding to chicken erythrocytes. This binding pattern was also observed with 1 of the swine isolates, suggesting it might be a trait of swine-adapted viruses. Taken together, a difference in the hemadsorption phenotype observed with erythrocytes from species with either less Neu5Acα2,6Gal or less Neu5Ac linkage overall could be explained by the mammalian origin of the novel pandemic (H1N1) 2009 infl uenza viruses.
To test this possibility, we measured the HA affi nity of H1 infl uenza viruses from various species of origin for synthetic receptor analogues. All mammalian H1 viruses showed a typical human virus-like preference for the Neu5Acα2,6Gal-containing receptor 6′SLN. Compared with the currently circulating H1N1 human viruses, both pandemic (H1N1) 2009 strains and contemporary swine infl uenza virus (H1N1) strains were able to bind substantially more strongly (5-12×) to an α2,6-containing glycopolymer; the currently circulating subtype H1N1 human viruses are strictly adapted to this receptor (12). This feature demonstrated that pandemic H1N1 strains, which have a HA gene To identify substitutions in the HA molecule that could be responsible for the human-like receptor binding phenotype of the pandemic and contemporary swine infl uenza (H1N1) isolates, we compared the H1 HA sequences deposited in the Infl uenza Research Database (www.fl udb.org). We observed that 99.99% of all swine viruses isolated after 1980 have Asp190 or Asn190. HA sequences of swine viruses isolated before 2000 harbor Gly225, whereas 92.3% of more recent classical swine viruses have Asp225. Crystallographic analysis of human and swine H1 HA has shown that Asp225 and Asp190 are responsible for human virus-like receptor specifi city (13). Therefore, the human-like amino acids encoded at HA positions 190 and 225 in the novel pandemic and swine infl uenza (H1N1) viruses may at least partially explain their innate affi nity for the human-type receptor.
Recent phylogenetic analysis showed that each segment of the pandemic (H1N1) 2009 virus is nested within a well-established swine infl uenza lineage for >10 years before the recent outbreak (7). Hence, the ancestors of this virus circulated undetected for about a decade before the virus emerged in humans. Our fi nding that contemporary swine viruses acquired the ability to recognize 6′SLN with at least 5-fold higher affi nity than did human strains and completely lost the ability to bind to Neu5Acα2,3Gal provides clear evidence to support this hypothesis. It is possible that the progenitors of pandemic (H1N1) 2009 virus were accumulating enough mammal-associated changes to allow a refi nement of their receptor-binding properties. Our fi ndings substantiate that strong mammalian-like receptor specifi city is a critical barrier to infection of various hosts with pandemic (H1N1) 2009 virus. Other biological factors associated with their adaptation and tissue tropism in humans will likely be identifi ed in the future. 3.4 0.9 ‡ < 2.0 1.3 ‡ 0.8 0.9 1.3 0.9 1.3 0.9 33 ‡ *dpi, days postinoculation; <, titer below limit of detection (<0.75 log 10 EID 50 /mL). Virus titers were determined in eggs and are expressed as the log 10 50% egg infectious dose (EID 50 )/mL (8). Data are presented as means ± standard deviation of titers of positive samples ( 0.75 log 10 EID 50 /mL). †Percentage of contact birds from which virus was isolated. ‡p<0.05, by 1-way analysis of variance.