HLA-B*27 subtype specificity determines the targeting of HLA-B*27 restricted HCV-specific CD8+ T cell epitopes

The HLA class I allele B*27 is associated with spontaneous viral clearance in Hepatitis C virus (HCV) genotype 1 infection. This protective effect of HLA-B*27 is linked to an immunodominant epitope as well as a subdominant epitope, both located in the HCV polymerase NS5B. Viral escape mutations occur consistently in these two epitopes in chronically infected patients and can abolish recognition of these epitopes by CD8+ T cells.

Additionally, a third HLA-B*27 epitope has been described that is also located in the NS5B protein (NS5B-2820, ARHTPVNSW). However, neither recognition of NS5B-2820 by CD8+ T cells nor viral escape mutations in this epitope have been observed in a substantial number of HLA-B*27+ patients. Here, we analyzed why NS5B-2820 is not dominantly recognized in patients carrying HLA-B*27. Four digit HLA class I typing, sequence analyses of the autologous viral sequences and immunological analyses of CD8+ T cell responses to this epitope were performed in a cohort of 55 HLA-B*27+ patients with chronic HCV genotype 1 infection. As expected, CD8+ T cell responses against the NS5B-2820 epitope were detectable in a minority of patients only. Additionally, only 9 out of the 55 HLA-B*27+ patients displayed viral escape mutations in the HLA-B*27 binding anchor at the second amino acid of the epitope (R2821K). Next we correlated these findings with the HLA-B*27 subtypes of the patients. We found that the vast majority (9 out of 10) of patients who were HLA-B*27:02+ showed the R2821K mutation while none of the patients carrying the more prevalent HLA-B*27:05 subtype or other rare HLA-B*27 subtypes displayed escape mutations within the NS5B-2820 epitope (p<0.0001). These results are in line with the previous observation that tryptophan (W), the amino acid present at the C terminus of the NS5B-2820 epitope, is allowed as C terminus of HLA-B*27:02 peptide ligands, but not as C terminus of peptides binding to other HLA-B*27 subtypes.

In conclusion, our data clearly indicate that the NS5B-2820 epitope is immunodominant in the context of HLA-B*27:02, but is not restricted by other HLA-B*27 subtypes. This finding clearly indicates the importance of HLA subtypes in the restriction of virus-specific CD8+ T cell responses. It also indicates that immunodominant HCV-specific CD8+ T cell responses may vary in different global populations, not only due to predominance of different HLA groups (e.g., HLA-A2, HLA-B27, etc.) but also due to different HLA subtypes (e.g., HLA-B*27:02, HLA-B*27:05 etc.). This limitation has important implications for the selection of epitopes for global vaccine design.