Divergent Human-Origin Influenza Viruses Detected in Australian Swine Populations

Abstract

Global swine populations infected with influenza A viruses pose a persistent pandemic risk. With the exception of a few countries, our understanding of the genetic diversity of swine influenza viruses is limited, hampering control measures and pandemic risk assessment. Here we report the genomic characteristics and evolutionary history of influenza A viruses isolated in Australia from 2012 to 2016 from two geographically isolated swine populations in the states of Queensland and Western Australia. Phylogenetic analysis with an expansive human and swine influenza virus data set comprising >40,000 sequences sampled globally revealed evidence of the pervasive introduction and long-term establishment of gene segments derived from several human influenza viruses of past seasons, including the H1N1/1977, H1N1/1995, H3N2/1968, and H3N2/2003, and the H1N1 2009 pandemic (H1N1pdm09) influenza A viruses, and a genotype that contained gene segments derived from the past three pandemics (1968, reemerged 1977, and 2009). Of the six human-derived gene lineages, only one, comprising two viruses isolated in Queensland during 2012, was closely related to swine viruses detected from other regions, indicating a previously undetected circulation of Australian swine lineages for approximately 3 to 44 years. Although the date of introduction of these lineages into Australian swine populations could not be accurately ascertained, we found evidence of sustained transmission of two lineages in swine from 2012 to 2016. The continued detection of human-origin influenza virus lineages in swine over several decades with little or unpredictable antigenic drift indicates that isolated swine populations can act as antigenic archives of human influenza viruses, raising the risk of reemergence in humans when sufficient susceptible populations arise.IMPORTANCE We describe the evolutionary origins and antigenic properties of influenza A viruses isolated from two separate Australian swine populations from 2012 to 2016, showing that these viruses are distinct from each other and from those isolated from swine globally. Whole-genome sequencing of virus isolates revealed a high genotypic diversity that had been generated exclusively through the introduction and establishment of human influenza viruses that circulated in past seasons. We detected six reassortants with gene segments derived from human H1N1/H1N1pdm09 and various human H3N2 viruses that circulated during various periods since 1968. We also found that these swine viruses were not related to swine viruses collected elsewhere, indicating independent circulation. The detection of unique lineages and genotypes in Australia suggests that isolated swine populations that are sufficiently large can sustain influenza virus for extensive periods; we show direct evidence of a sustained transmission for at least 4 years between 2012 and 2016.

Keywords: antigenicity; influenza surveillance; pandemic risk; phylogenetic analysis; reassortment; swine influenza.

FIG 1

(a and b) Evolutionary relationships of the H1-HA (a) and H3-HA (b) genes of Australian swine influenza viruses (2012 to 2016) (red). Swine viruses isolated elsewhere are colored gray, while human lineages are colored black. Nth American, North American. (c to f) The human origins of each of the Australian swine HA gene lineages are shown with a representative set of closely related human or swine viruses: H1N1/pdm09-like (c), H1N1/1995-like (d), H1N1/1977-like (e), and H3N2/1995-like (f). The scale bar represents the number of nucleotide substitutions per site. Branch support values (>70 UFBoot) are shown at the nodes. Blue branches in Fig. 1c reveal the phylogenetic relationships of previously published H1N1pdm09-like viruses from Australian swine.

FIG 2

(a and b) Evolutionary relationships of the N2-NA (a) and N1-NA (b) genes of Australian swine influenza viruses (2012 to 2016) (red). Swine viruses isolated elsewhere are colored gray, while human lineages are colored black. (c to f) The human origins of each of the Australian swine NA gene lineages are shown with a representative set of closely related human or swine viruses: H3N2/2003-like (c), H3N2/1995-like (d), H3N2/1968-like (e), and H1N1/pdm09-like (f). The scale bar represents the number of nucleotide substitutions per site. Branch support values (>70 UFBoot) are shown at the nodes.

FIG 3

(a) Gene sources of Australian swine influenza viruses. (b and c) Maximum likelihood phylogenetic trees of the PB2 (b) and PB1 (c) data sets are shown to illustrate the internal gene sources of Australian swine viruses. Branch support values (>70 UFBoot) are shown at the nodes. The scale bar represents the number of nucleotide substitutions per site.

FIG 4

Comparison of HA1 amino acid substitutions between 2012 and 2016 isolates of Western Australian swine influenza H3N2 (a) and H1N2 (b) viruses and antigenically cross-reactive influenza A virus antigens. Known HA1 antigenic sites are boxed in gray and numbered according to the H3 (47) and H1 (32) numbering systems. Dots denote positions with amino acid residues identical to those of sw/Western Australia/2577766G/2012(H3N2) and sw/Western Australia/2577896X/2012(H1N2).