Background:  Infection with hepatitis C virus (HCV) is an important cause of liver disease both nationally and internationally, disproportionately affecting persons with HIV and AIDS. Advancement in understanding HCV pathogenesis has been hampered by inefficient amplification of large segments of the HCV genome, heterogeneous natural infection, and heterogeneity of host HLA alleles. We have overcome these limitations by applying a long-amplicon RT-PCR method to specimens from a rare common-source outbreak, in persons selected for the most common HLA alleles. These resources allow us to compare sequence changes 20 years after infection with a known, relatively homogeneous inoculum, and compare these changes with predicted HLA class I epitopes.

Methods:  Serum was obtained from a cohort of Irish women 20±2 years after accidental exposure to HCV-contaminated anti-D immunoglobulin, which occurred in 1977. On the basis of common A locus HLA alleles (A1, A2, or A3), 22 subjects were selected, including all available subjects who were homozygous for A1 or A2. RNA was isolated from cryopreserved serum specimens, and RT-PCR and cDNA cloning were performed. Ten clones per specimen were analyzed to identify 2 representative clones from each subject for sequencing of 5 kb spanning the Core, E1, E2, p7, NS2, and NS3 genes. Sequences were analyzed using PAUP*, SimPlot, and VarPlot.

Results:  Phylogenetic analysis of the E1-E2 region spanning HVR1 revealed that all 220 sequences obtained 20 years after acute infection, plus 10 each from 2 specimens of inoculum source serum, clustered monophyletically against a background of nearly 100 subtype 1b reference sequences. Despite this shared ancestry, remarkable diversity was observed, particularly in the first hypervariable region (HVR1), which varied in length from 24 to 28 amino acids. With 2 exceptions, all of the sequences from each study subject clustered monophyletically, indicating individualized pathways of sequence variation. Amino acid substitutions frequently occurred in predicted HLA class I epitopes.

Conclusions:  HCV sequence variation in a common-source cohort of young women reveals evidence of nonrandom substitution correlated with predicted HLA class I epitopes. These results suggest that HCV sequence variation is driven by the host immune response, and therefore represents a biological probe for host-virus interaction.

Keywords: HCV; Evolution; Immunology