Search

H9N2 viruses are the most widespread influenza viruses in poultry in Asia. We evaluated the infection and tropism of human and avian H9 influenza virus in the human respiratory tract using ex vivo respiratory organ culture. H9 viruses infected the upper and lower respiratory tract and the majority of H9 viruses had a decreased ability to release virus from the bronchus rather than the lung. This may be attributed to a weak neuraminidase (NA) cleavage of carbon-6-linked sialic acid (Sia) rather than carbon-3-linked Sia. The modified cleavage of N-acetlylneuraminic acid (Neu5Ac) and N-glycolylneuraminic acid (Neu5Gc) by NA in H9 virus replication was observed by reverse genetics, and recombinant H9N2 viruses with amino acids (38KQ) deleted in the NA stalk, and changing the amino acid at position 431 from Proline-to-Lysine. Using recombinant H9 viruses previously evaluated in the ferret, we found that viruses which replicated well in the ferret did not replicate to the same extent in the human ex vivo cultures. The existing risk assessment models for H9N2 viruses in ferrets may not always have a strong correlation with the replication in the human upper respiratory tract. The inclusion of the human ex vivo cultures would further strengthen the future risk-assessment strategies.

The 701 and 702 positions of influenza PB2 polymerase subunit are previously shown to have roles on host range. Limited polymorphisms at these two residues are identified in natural isolates, thereby limiting the study of their role in the polymerase. In this study, we generated 31 viable viruses by random mutagenesis at this region, indicating that these positions can tolerate a wide range of amino acids. These mutants demonstrated varying polymerase activities and viral replication rates in mammalian and avian cells. Notably, some mutants displayed enhanced polymerase activity, yet their replication kinetics were comparable to the wild-type virus. Surface electrostatic charge predication on the PB2 structural model revealed that the viral polymerase activity in mammalian cells generally increases as this region becomes more positively charged. One of the mutants (701A/702E) showed much reduced pathogenicity in mice while others had a pathogenicity similar to the wild-type level. Distinct tissue tropisms of the PB2-701/702 mutants were observed in infected chicken embryos. Overall, this study demonstrates that the PB2-701/702 region has a high degree of sequence plasticity and sequence changes in this region can alter virus phenotypes in vitro and in vivo.

Novel avian H7N9 virus emerged in China in 2013 resulting in a case fatality rate of around 39% and continues to pose zoonotic and pandemic risk. Amino acid substitutions in PB2 protein were shown to influence the pathogenicity and transmissibility of H7N9 following experimental infection of ferrets and mice. In this study, we evaluated the role of amino acid substitution PB2-627K or compensatory changes at PB2-591K and PB2-701N, on the tropism and replication competence of H7N9 viruses for human and swine respiratory tracts using ex vivo organ explant cultures. Recombinant viruses of A/Shanghai/2/2013 (rgH7N9) and its mutants with PB2-K627E, PB2-K627E + Q591K and PB2-K627E + D701N were generated by plasmid-based reverse genetics. PB2-E627K was essential for efficient replication of rgH7N9 in ex vivo cultures of human and swine respiratory tracts. Mutant rgPB2-K627E + D701N replicated better than rgPB2-K627E in human lung but not as well as rgH7N9 virus. The rgPB2-K627E mutant failed to replicate in human type I-like pneumocytes (ATI) and peripheral blood monocyte-derived macrophages (PMϕ) at 37 °C while the compensatory mutant rgPB2-K627E + Q591K and rgPB2-K627E + D701N had partly restored replication competence in PMϕ. Our results demonstrate that PB2-E627K was important for efficient replication of influenza H7N9 in both human and swine respiratory tracts.

Highly pathogenic avian influenza (HPAI) H5N1 virus continues to pose pandemic threat, but there is a lack of understanding of its pathogenesis. We compared the apoptotic responses triggered by HPAI H5N1 and low pathogenic H1N1 viruses using physiologically relevant respiratory epithelial cells. We demonstrated that H5N1 viruses delayed apoptosis in primary human bronchial and alveolar epithelial cells (AECs) compared to H1N1 virus. Both caspase-8 and -9 were activated by H5N1 and H1N1 viruses in AECs, while H5N1 differentially up-regulated TRAIL. H5N1-induced apoptosis was reduced by TRAIL receptor silencing. More importantly, STAT3 knock-down increased apoptosis by H5N1 infection suggesting that H5N1 virus delays apoptosis through activation of STAT3. Taken together, we demonstrate that STAT3 is involved in H5N1-delayed apoptosis compared to H1N1. Since delay in apoptosis prolongs the duration of virus replication and production of pro-inflammatory cytokines and TRAIL from H5N1-infected cells, which contribute to orchestrate cytokine storm and tissue damage, our results suggest that STAT3 may play a previously unsuspected role in H5N1 pathogenesis.