The phylogenetics and ecology of the orthopoxviruses endemic to North America

Abstract

The data presented herein support the North American orthopoxviruses (NA OPXV) in a sister relationship to all other currently described Orthopoxvirus (OPXV) species. This phylogenetic analysis reaffirms the identification of the NA OPXV as close relatives of "Old World" (Eurasian and African) OPXV and presents high support for deeper nodes within the Chordopoxvirinae family. The natural reservoir host(s) for many of the described OPXV species remains unknown although a clear virus-host association exists between the genus OPXV and several mammalian taxa. The hypothesized host associations and the deep divergence of the OPXV/NA OPXV clades depicted in this study may reflect the divergence patterns of the mammalian faunas of the Old and New World and reflect a more ancient presence of OPXV on what are now the American continents. Genes from the central region of the poxvirus genome are generally more conserved than genes from either end of the linear genome due to functional constraints imposed on viral replication abilities. The relatively slower evolution of these genes may more accurately reflect the deeper history among the poxvirus group, allowing for robust placement of the NA OPXV within Chordopoxvirinae. Sequence data for nine genes were compiled from three NA OPXV strains plus an additional 50 genomes collected from Genbank. The current, gene sequence based phylogenetic analysis reaffirms the identification of the NA OPXV as the nearest relatives of "Old World" OPXV and presents high support for deeper nodes within the Chordopoxvirinae family. Additionally, the substantial genetic distances that separate the currently described NA OPXV species indicate that it is likely that many more undescribed OPXV/NA OPXV species may be circulating among wild animals in North America.

Figure 1. Poxvirus phylogenetic tree inferred from Bayesian analysis of the DNA sequences from 9 conserved genes.

Four Metropolis-coupled Markov chain Monte Carlo simulations using the general time reversible (GTR) + γ + I model were run over 5 million generations, with trees and parameters being sampled every 100 generations. The consensus tree was derived from 30,000 of the trees sampled. All but 2 nodes had posterior probabilities of 1.00; these two nodes are labeled in the figure. A close-up of the Orthopoxvirus lineage is presented in the inset. GC-rich genomes are indicated by the yellow box.