Novel reassortant H9N2 viruses in pigeons and evidence for antigenic diversity of H9N2 viruses isolated from quails in Egypt

Abstract

The endemicity of avian influenza viruses (AIVs) among Egyptian poultry represents a public health risk. Co-circulation of low pathogenic AIV H9N2 subtype with highly pathogenic AIV H5N1 subtype in Egyptian farms provides a possibility to generate novel reassortant viruses. Here, the genetic characteristics of surface glycoproteins of 59 Egyptian H9N2 viruses, isolated between 2013 and 2015, were analysed. To elucidate the potential of genetic reassortment, 10 H9N2 isolates were selected based on different avian hosts (chickens, ducks, pigeons and quails) and phylogenetic analyses of their full genome sequences were conducted. Additionally, we performed antigenic analysis to further investigate the antigenic evolution of H9N2 viruses isolated during 2011-2015. Different viral characteristics including receptor-binding affinity and drug resistance of representative Egyptian H9N2 viruses were further investigated. The surface glycoproteins of current Egyptian H9N2 viruses were closely related to viruses of the G1-like lineage isolated from Egypt. Several genetic markers that enhance virulence in poultry and transmission to humans were detected. Analysis of the full genome of 10 H9N2 isolates indicated that two pigeon isolates inherited five internal genes from Eurasian AIVs circulating in wild birds. Antigenic conservation of different Egyptian H9N2 isolates from chickens, pigeons and ducks was observed, whereas quail isolates showed antigenic drift. The Egyptian H9N2 viruses preferentially bound to the human-like receptor rather than to the avian-like receptor. Our results suggest that the endemic H9N2 viruses in Egypt contain elements that may favour avian-to-human transmission and thus represent a public health risk.

Fig. 1.

Schematic illustration of the reassortment process of H9N2 viruses isolated from pigeons in Egypt.

Fig. 2.

Phylogenetic tree of the nucleotide sequences of HA and NA of H9N2 viruses isolated from different hosts in Egypt. H9N2 isolates sequenced specifically for this study are labelled with red circles. H9N2 vaccine strains are labelled with blue squares. Ancestor H9N2 viruses are labelled with red colour. Phylogenetic analysis was performed by using mega version 6.

Fig. 3.

Phylogenetic trees of the nucleotide sequences of the PB2, PB1, PA, NP, M and NS genes of H9N2 viruses isolated from different hosts in Egypt. H9N2 isolates sequenced specifically for this study are labelled with red circles. Ancestor H9N2 viruses are labelled with red colour. Non-H9N2 subtypes are labelled with blue colour. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown at the dendrogram nodes. Phylogenetic analysis was performed by using mega version 6.

Fig. 3.

Phylogenetic trees of the nucleotide sequences of the PB2, PB1, PA, NP, M and NS genes of H9N2 viruses isolated from different hosts in Egypt. H9N2 isolates sequenced specifically for this study are labelled with red circles. Ancestor H9N2 viruses are labelled with red colour. Non-H9N2 subtypes are labelled with blue colour. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown at the dendrogram nodes. Phylogenetic analysis was performed by using mega version 6.

Fig. 3.

Phylogenetic trees of the nucleotide sequences of the PB2, PB1, PA, NP, M and NS genes of H9N2 viruses isolated from different hosts in Egypt. H9N2 isolates sequenced specifically for this study are labelled with red circles. Ancestor H9N2 viruses are labelled with red colour. Non-H9N2 subtypes are labelled with blue colour. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown at the dendrogram nodes. Phylogenetic analysis was performed by using mega version 6.

Fig. 4.

Antigenic cartography representation of the HI data generated by using a panel of mAbs (a) and polyclonal chicken antisera (b). The maps were generated by using AntigenMap (http://sysbio.cvm.msstate.edu/AntigenMap). One grid represents a twofold change in the HI assay results. Three drifted isolates from quails are enclosed by an oval.

Fig. 5.

Receptor-binding specificity of Egyptian H9N2 viruses. Direct binding of Egyptian H9N2 viruses to different concentrations of biotinylated sialylglycopolymers (X axis) containing 3′-sialyllactose (α2,3-SL), 6′-sialyllactose (α2,6-SL) or 6-sialyl-N-acetyllactosamine (6′-SLN) was measured at 490 nm (Y axis). Influenza A/chicken/Egypt/M7217B/2013(H5N1), A/Hong Kong/1073/99(H9N2) and A/duck/Hong Kong/365/78(H4N6) were used as controls for binding assay.