Evolution and spread of Venezuelan equine encephalitis complex alphavirus in the Americas

Abstract

Venezuelan equine encephalitis (VEE) complex alphaviruses are important re-emerging arboviruses that cause life-threatening disease in equids during epizootics as well as spillover human infections. We conducted a comprehensive analysis of VEE complex alphaviruses by sequencing the genomes of 94 strains and performing phylogenetic analyses of 130 isolates using complete open reading frames for the nonstructural and structural polyproteins. Our analyses confirmed purifying selection as a major mechanism influencing the evolution of these viruses as well as a confounding factor in molecular clock dating of ancestors. Times to most recent common ancestors (tMRCAs) could be robustly estimated only for the more recently diverged subtypes; the tMRCA of the ID/IAB/IC/II and IE clades of VEE virus (VEEV) were estimated at ca. 149-973 years ago. Evolution of the IE subtype has been characterized by a significant evolutionary shift from the rest of the VEEV complex, with an increase in structural protein substitutions that are unique to this group, possibly reflecting adaptation to its unique enzootic mosquito vector Culex (Melanoconion) taeniopus. Our inferred tree topologies suggest that VEEV is maintained primarily in situ, with only occasional spread to neighboring countries, probably reflecting the limited mobility of rodent hosts and mosquito vectors.

Fig 1. Maximum likelihood phylogeny for VEEV complex.

Relevant internal nodes are identified.

Fig 2. Evidence of branch length underestimation in the VEE complex shown via comparison of tree lengths inferred under Branch-Site REL (BSREL) and GTR+Γ₄ substitution models.

Letters correspond to nodes in maximum likelihood tree. The aBSREL analysis performed with an optimized number of rate classes is shown in black. BSREL analysis performed with a fixed number of three rate classes is shown in gray. Instances in which only a black letter is shown indicate that aBSREL and BSREL produced identical results. The dashed line depicts x = y, an unbiased analysis.

Fig 3. The genetic relationships of the II/ID/IAB/IC subtypes of the VEEV complex.

MCC tree as determined by coalescent analysis for the ID/II strains with the IAB and IC strains included. All branches had posterior probabilities of >0.97 except for the branches marked with a *, which did not have significant support. The time to most recent common ancestors (tMRCA) dates for individual branches as identified from the MCC tree are indicated on significant branches with the highest posterior density (HPD) in brackets, dates of clades supported by BSREL were emboldened. Subtype IAB strains are highlighted by the grey box.

Fig 4. The genetic relationships of the IE subtype of the VEEV complex.

MCC tree as determined by coalescent analysis for the IE strains. All branches had posterior probabilities of >0.97 except for the branches marked with a *, which did not have significant support. Time to most recent common ancestor (tMRCA) as determined by the MCC tree were added to major branches with the HPD in brackets. If the clades were supported by BSREL the tMRCA and HPD were emboldened.