Rapid fixation of a distinctive sequence motif in the 3' noncoding region of the clade of West Nile virus invading North America

Abstract

Phylogenetic analysis of complete genomes of West Nile virus (WNV) by a variety of methods supported the hypothesis that North American isolates of WNV constitute a monophyletic group, together with an isolate from Israel and one from Hungary. We used ancestral sequence reconstruction in order to obtain evidence for evolutionary changes that might be correlated with increased virulence in this clade (designated the N.A. clade). There was one amino acid change (I-->T at residue 356 of the NS3 protein) that occurred in the ancestor of the N.A. clade and remained conserved in all N.A. clade genomes analyzed. There were four changes in the upstream portion of the 3' noncoding region (the AT-enriched region) that occurred in the ancestor of the N.A. clade and remained conserved in all N.A. clade genomes analyzed, changes predicted to alter RNA secondary structure. The AT-enriched region showed a higher rate of substitution in the branch ancestral to the N.A. clade, relative to polymorphism, than did the remainder of the noncoding regions, synonymous sites in coding regions, or nonsynonymous sites in coding regions. The high rate of occurrence of fixed nucleotide substitutions in this region suggests that positive Darwinian selection may have acted on this portion of the 3'NCR and that these fixed changes, possibly in concert with the amino acid change in NS3, may underlie phenotypic effects associated with increased virulence in North American WNV.

Figure 1

NJ tree of WNV polyprotein amino acid sequences (based on 3413 aligned sites). Branches with less than 50% bootstrap support are condensed, and topology only is shown. Accession numbers and geographic origin (including abbreviations for states in the case of isolates from the United States) are given. Lineages are identified following Bakonyi et al. (2006). Selected JEV isolates were used as an outgroup to root the tree.

Figure 2

NJ tree of coding nucleotide sequences of WNV lineage 1a (based on 10286 aligned sites). Branches with less than 50% bootstrap support are condensed, and topology only is shown. Following Figure 1, AY603654 (Ethiopia), AF260968 (Egypt), and AY490240 (China) were used as an outgroup to root the tree.

Figure 3

NJ tree of noncoding nucleotide sequences of WNV lineage 1a (based on 467 aligned sites). Branches with less than 50% bootstrap support are condensed, and topology only is shown. Following Figure 1, AY603654 (Ethiopia), AF260968 (Egypt), and AY490240 (China) were used as an outgroup to root the tree.

Figure 4

Predicted secondary structures of WNV 3′ NCRs: (A) AF260968 (a representative member of lineage 1a outside the N.A. clade); (B) DQ164195 (a representative member of the N.A. clade); (C) common ancestor of the N.A. clade; (D) immediate ancestor of the common ancestor of the N.A. clade (ancestor of the N.A. clade and its nearest sister group); (E) sequence differing from that in (D) by only the four unique nucleotide substitutions conserved in the N.A. clade. The arrow (B and C) indicates a stem-loop structure in which position 46 of the 3′NCR contributes to the loop.