Role of Position 627 of PB2 and the Multibasic Cleavage Site of the Hemagglutinin in the Virulence of H5N1 Avian Influenza Virus in Chickens and Ducks

Abstract

Highly pathogenic H5N1 avian influenza viruses have caused major disease outbreaks in domestic and free-living birds with transmission to humans resulting in 59% mortality amongst 564 cases. The mutation of the amino acid at position 627 of the viral polymerase basic-2 protein (PB2) from glutamic acid (E) in avian isolates to lysine (K) in human isolates is frequently found, but it is not known if this change affects the fitness and pathogenicity of the virus in birds. We show here that horizontal transmission of A/Vietnam/1203/2004 H5N1 (VN/1203) virus in chickens and ducks was not affected by the change of K to E at PB2-627. All chickens died between 21 to 48 hours post infection (pi), while 70% of the ducks survived infection. Virus replication was detected in chickens within 12 hours pi and reached peak titers in spleen, lung and brain between 18 to 24 hours for both viruses. Viral antigen in chickens was predominantly in the endothelium, while in ducks it was present in multiple cell types, including neurons, myocardium, skeletal muscle and connective tissues. Virus replicated to a high titer in chicken thrombocytes and caused upregulation of TLR3 and several cell adhesion molecules, which may explain the rapid virus dissemination and location of viral antigen in endothelium. Virus replication in ducks reached peak values between 2 and 4 days pi in spleen, lung and brain tissues and in contrast to infection in chickens, thrombocytes were not involved. In addition, infection of chickens with low pathogenic VN/1203 caused neuropathology, with E at position PB2-627 causing significantly higher infection rates than K, indicating that it enhances virulence in chickens.

The Effect of the PB2 Mutation 627K on Highly Pathogenic H5N1 Avian Influenza Virus Is Dependent on the Virus Lineage

“…The finding that this mutation is neutral in ducks may help refine risk assessments, such as those described by Russell et al (2012) (50), and indicates the high probability that this mammalian adaptation motif will be maintained in the avian reservoir. Similarly, Schat et al (2012) (42) concluded that neither disease outcome nor transmission of an HP H5N1 virus in ducks or chickens was altered by PB2 627E or 627K. Their study used a human H5N1 virus isolate of clade 1 (A/Vietnam/1203/2004) which bears 627K.…”

“…6) in chicken cells in either genetic backbone. Schat et al (2012) (42) did not investigate transmission of the LP viruses between chickens or ducks, so the biological relevance of the ob- servations in chickens for the maintenance of 627E genotype virus in LP or HP avian influenza viruses in a natural reservoir is not clear.…”

“…On the other hand, several reports have described the selection of avian influenza viruses with the PB2 627K mutation, particularly during passage in mice, including H9N2 (60, 61) and H7N7 (62). There are also several reports of viruses that have been stably engineered to possess the 627K mutation (29,36,42,63,64). Indeed, in the recent outbreak of a novel H7N9 virus in China in 2013, the PB2 627K mutation is often selected for upon human infection (65,66).…”

“…Polymerase function and growth of PB2 627E/K viruses in avian species-derived cells have demonstrated no host specificity (29,40,41). However, recent work by Schat et al (2012) suggested a small fitness cost associated with H5N1 PB2 627K in the avian host (42). The biological significance of this observation remains unclear, as this single piece of work used a human H5N1 isolate that might have contained other genetic adaptations, and the effect was apparent only in a genetically modified virus background from which the multibasic cleavage site in HA was removed.…”

A Unique Multibasic Proteolytic Cleavage Site and Three Mutations in the HA2 Domain Confer High Virulence of H7N1 Avian Influenza Virus in Chickens

“…Besides the CS, amino acids in its vicinity, mostly in the HA1 domain, were also important in some H5 viruses (7,8). Moreover, virulence markers that extend beyond the HA exist, mainly in the H5N1 subtype (9)(10)(11). Thus, the virulence of H5 AIV is determined multigenically, requiring an optimal gene constellation for full virulence or host adaptation (12).…”