Viral sequence evolution in acute hepatitis C virus infection

Abstract

CD8(+)-T-cell responses play an important role in the containment and clearance of hepatitis C virus (HCV) infection, and an association between viral persistence and development of viral escape mutations has been postulated. While escape from CD8+ -T-cell responses has been identified as a major driving force for the evolution of human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV), a broader characterization of this relationship is needed in HCV infection. To determine the extent, kinetics, and driving forces of HCV sequence evolution, we sequenced the entire HCV genome longitudinally in four subjects monitored for up to 30 months after acute infection. For two subjects the transmission sources were also available. Of 53 total non-envelope amino acid substitutions detected, a majority represented forward mutations away from the consensus sequence. In contrast to studies in HIV and SIV, however, only 11% of these were associated with detectable CD8+ T-cell responses. Interestingly, 19% of non-envelope mutations represented changes toward the consensus sequence, suggesting reversion in the absence of immune pressure upon transmission. Notably, the rate of evolution of forward and reverse mutations correlated with the conservation of each residue, which is indicative of structural constraints influencing the kinetics of viral evolution. Finally, the rate of sequence evolution was observed to decline over the course of infection, possibly reflective of diminishing selection pressure by dysfunctional CD8+ T cells. Taken together, these data provide insight into the extent to which HCV is capable of evading early CD8+ T-cell responses and support the hypothesis that dysfunction of CD8+ T cells may be associated with failure to resolve HCV infections.

FIG. 1.

Clinical course of acute HCV infection. Four study subjects were monitored longitudinally from acute HCV infection. Alanine aminotransferase (ALT) levels, HCV viral loads, bilirubin levels of >1 mg/dl, and the presence of anti-HCV antibodies are shown. Arrows indicate time points from which viral sequences were generated.

FIG. 2.

Close phylogenetic relationship between viral sequences in transmission pairs. Using the neighbor-joining method, HCV sequences 2,960 amino acids in length were compared to 38 HCV genotype 1 sequences arbitrarily chosen from the Los Alamos HCV database and an additional 11 full-length GT1a sequences generated in-house (Timm et al., unpublished). Sequences of the earliest time points in the recipients (BR554 first, BR1427 first) are more closely related to sequences from the sources (BR601 and BR1430) than to consecutive sequences in the recipients (BR554 last, BR1427 last) or any other GT1 sequence from the database (bootstrap values 93 to 100), confirming transmission of the virus from BR601 to BR554 and from BR1430 to BR1427.

FIG. 3.

Sites of sequence variation and detected CD8⁺ T-cell responses. HCV viral genome sequences at multiple time points are illustrated as horizontal bars with time points shown on the left. Sequence changes are indicated as vertical dashes for each sequenced time point. The number of new mutations and the sum of mutations since the earliest available time point are listed on the right, with mutations considered to have happened during the early phase of the infection being framed in gray boxes. Detected CD8⁺ T-cell responses are depicted as circles and arrows. Asterisks mark targeted epitopes in which sequence variations were observed over time.

FIG. 4.

Rates of nonsynonymous changes decline over the course of infection. Rates of nonsynonymous changes in the dominant viral sequence (per year, normalized per 100 codons) are indicated separately for envelope and for the rest of the HCV genome. Independent of the viral loads, the rate of sequence changes is highest during the acute phase of the infection and decreases thereafter. dN/dS ratios are low upon transmission and fluctuate around individually higher levels thereafter.

FIG. 5.

The rate of forward and reverse mutations correlates with the conservation of amino acid residues on a population level. Conservation scores were plotted for the residues in which forward and reverse mutations were detected, comparing mutations arising between the first six months of infection and the remaining time of follow up. Early arising forward mutations were located within less conserved residues compared to late arising mutations, while reverse mutations tended to arise faster (nonsignificant) within amino acid residues exhibiting a higher degree of conservation.