Abstract
The presence of a hypervariable (HVR) region within the genome of hepatitis E virus (HEV) remains unexplained. Previous studies have described the HVR as a proline-rich spacer between flanking functional domains of the ORF1 polyprotein. Others have proposed that the region has no function, that it reflects a hypermutable region of the virus genome, that it is derived from the insertion and evolution of host sequences or that it is subject to positive selection. This study attempts to differentiate between these explanations by documenting the evolutionary processes occurring within the HVR. We have measured the diversity of HVR sequences within acutely infected individuals or amongst sequences derived from epidemiologically linked samples and, surprisingly, find relative homogeneity amongst these datasets. We found no evidence of positive selection for amino acid substitution in the HVR. Through an analysis of published sequences, we conclude that the range of HVR diversity observed within virus genotypes can be explained by the accumulation of substitutions and, to a much lesser extent, through deletions or duplications of this region. All published HVR amino acid sequences display a relative overabundance of proline and serine residues that cannot be explained by a local bias towards cytosine in this part of the genome. Although all published HVRs contain one or more SH3-binding PxxP motifs, this motif does not occur more frequently than would be expected from the proportion of proline residues in these sequences. Taken together, these observations are consistent with the hypothesis that the HVR has a structural role that is dependent upon length and amino acid composition, rather than a specific sequence.