Evolution and phylogeographic dissemination of endemic porcine picornaviruses in Vietnam

Abstract

Members of the Picornaviridae are important and often zoonotic viruses responsible for a variety of human and animal diseases. However, the evolution and spatial dissemination of different picornaviruses circulating in domestic animals are not well studied. We examined the rate of evolution and time of origin of porcine enterovirus G (EV-G) and porcine kobuvirus species C lineages (PKV-C) circulating in pig farms in Vietnam and from other countries. We further explored the spatiotemporal spread of EV-G and PKV-C in Southwest Vietnam using phylogeographic models. Multiple types of EV-G are co-circulating in Vietnam. The two dominant EV-G types among isolates from Vietnam (G1 and G6) showed strong phylogenetic clustering. Three clades of PKV-C (PKV-C1-3) represent more recent introductions into Vietnam; PKV-C2 is closely related to PKV-C from Southwest China, indicating possible cross-border dissemination. In addition, high virus lineage migration rates were estimated within four districts in Dong Thap province in Vietnam for both EV-G types (G1, G6) and all PKV-C (C1-3) clades. We found that Chau Thanh district is a primary source of both EV-G and PKV-C clades, consistent with extensive pig trading in and out of the district. Understanding the evolution and spatial dissemination of endemic picornaviruses in pigs may inform future strategies for the surveillance and control of picornaviruses.

Keywords: enterovirus; evolution; kobuvirus; phylogeography; porcine picornavirus.

Figure 1.

Simplified Bayesian MCC tree of 155 VP1 sequences from EV-G isolated in pigs. Branches are coloured at the serotype level. The two predominant types in Vietnam: G1 (pink) and G6 (dark blue) are collapsed into triangles. The posterior support values for different EVG types were labelled on the nodes, except for type G13 and G15 (one sequence only). The tree labelled with sequences accession numbers is shown in Supplementary Fig. S1.

Figure 2.

MCC trees from the discrete phylogeography analysis of EV-G. (a) EV-G G1 tree phylogeny. (b) EV-G G6 tree phylogeny. Branches are coloured according to their descendent nodes annotated by the different sampled areas, with the key for colours shown on the left.

Figure 3.

MCC trees from the discrete phylogeography analysis of PKV-C. (a) Simplified Bayesian MCC tree of 240 VP1 sequences of PKV-C. The posterior support for C1 to C3 is labelled on the nodes. The tree labelled with sequences accession numbers is shown in Supplementary Fig. S2. (b) PKV-C Clade 1 tree phylogeny. (c) PKV-C Clade 2 tree phylogeny. (d) PKV-C Clade 3 tree phylogeny. Branches are coloured according to their descendent nodes annotated by the different sampled areas, with the key for colours shown on the left.

Figure 4.

Dissemination networks of co-circulating EV-G and PKV-C in Dong Thap province, Vietnam. Virus lineage migration rates (exchange year⁻¹) between districts and BF support were estimated using an asymmetric discrete trait model on phylogenies of EV-G and PKV-C types separately and jointly. Node size reflects the sample number for each area. Arrows show the direction of transmission between the two states; the edge weight indicates the per capita transmission rate. The colour of edge from pink to red indicates BF support from low to high. Only rates with BF > 3 are shown: (a) EV-G G1; (b) EV-G G6; (c) joint PKV-C C1-3; (d) the location of the four districts in Dong Thap province. The rates and BF support are summarized in Supplementary Tables S3–S5.