A Bayesian phylogeographical analysis of type 1 porcine reproductive and respiratory syndrome virus (PRRSV)

Abstract

Understanding viral transmission is an important factor for the effective prevention one of the most devastating swine diseases, porcine reproductive and respiratory syndrome. Focusing on molecular epidemiology of type 1 PRRSV, this study analysed a large ORF5 dataset collected worldwide from 1991 to 2012 using a coalescent-based Bayesian Markov chain Monte Carlo approach. The results suggested that the virus diversified into unique subpopulations in Russia & Belarus and Italy approximately 100 years ago. Previously unreported consecutive diffusions of the virus were identified, which showed that some countries, such as Spain and Germany, acted as distribution sources to some extent. This study also provided statistical evidence for the existence of an ORF5-based phylogeographical structure of type 1 PRRSV, in which the virus tended to cluster by geographical locations more tightly than expected by chance. In contrast to this tight geographical structure, the evolution of the ORF5 gene, based on mapping of non-synonymous/synonymous substitutions, was best described by a non-homogeneous process that could be implicated as a mechanism for viral immune evasion.