Distinct Escape Pathway by Hepatitis C Virus Genotype 1a from a Dominant CD8+ T Cell Response by Selection of Altered Epitope Processing

Abstract

Antiviral CD8(+) T cells are a key component of the adaptive immune response against HCV, but their impact on viral control is influenced by preexisting viral variants in important target epitopes and the development of viral escape mutations. Immunodominant epitopes highly conserved across genotypes therefore are attractive for T cell based prophylactic vaccines. Here, we characterized the CD8(+) T cell response against the highly conserved HLA-B*51-restricted epitope IPFYGKAI1373-1380 located in the helicase domain of NS3 in people who inject drugs (PWID) exposed predominantly to HCV genotypes 1a and 3a. Despite this epitope being conserved in both genotypes, the corresponding CD8(+) T cell response was detected only in PWID infected with genotype 3a and HCV-RNA negative PWID, but not in PWID infected with genotype 1a. In genotype 3a, the detection of strong CD8(+) T cell responses was associated with epitope variants in the autologous virus consistent with immune escape. Analysis of viral sequences from multiple cohorts confirmed HLA-B*51-associated escape mutations inside the epitope in genotype 3a, but not in genotype 1a. Here, a distinct substitution in the N-terminal flanking region located 5 residues upstream of the epitope (S1368P; P = 0.00002) was selected in HLA-B*51-positive individuals. Functional assays revealed that the S1368P substitution impaired recognition of target cells presenting the endogenously processed epitope. The results highlight that, despite an epitope being highly conserved between two genotypes, there are major differences in the selected viral escape pathways and the corresponding T cell responses.

Importance: HCV is able to evolutionary adapt to CD8(+) T cell immune pressure in multiple ways. Beyond selection of mutations inside targeted epitopes, this study demonstrates that HCV inhibits epitope processing by modification of the epitope flanking region under T cell immune pressure. Selection of a substitution five amino acids upstream of the epitope underlines that efficient antigen presentation strongly depends on its larger sequence context and that blocking of the multistep process of antigen processing by mutation is exploited also by HCV. The pathways to mutational escape of HCV are to some extent predictable but are distinct in different genotypes. Importantly, the selected escape pathway of HCV may have consequences for the destiny of antigen-specific CD8(+) T cells.

FIG 1

CD8⁺ T cell response against the epitope IPFYGKAI_1373–1380 in PWID exposed to HCV. (A) T cells were expanded for 10 days from PBMCs in the presence of the peptide IPFYGKAI. After in vitro expansion, the cells were restimulated with the peptide before intracellular IFN-γ staining. The autologous viral epitope sequences from GT3a-infected PWID with detectable responses are indicated. (B) HLA-B*51_1373–1380-specific T cells were detected ex vivo via IPFYGKAI-specific HLA-B*5101 dextramer staining. (C) The HLA-B*51_1373–1380 specific T cell response was determined by ELISpot assay at the indicated time points in a patient with acute HCV infection. (D and E) Serial peptide dilutions of the prototype (green), I1373V (blue), or I1380L (red) sequence of the B51-1373 peptide were tested in RNA-negative (D) or GT3a-infected (E) patients. Statistical comparisons between groups in panels A and B were done with a Kruskal-Wallis test, and significant P values are indicated.

FIG 2

Frequency of HLA-B*51-associated viral polymorphisms in the epitope region. The frequency of variations from the reported prototype sequence of the epitope region in GT1a (A), GT1b (B), the East-German Anti-D cohort (C), and GT3a (D) are shown for patients carrying the HLA-B*51 allele (red) and patients not carrying the HLA-B*51 allele (green). Positions with significant differences in polymorphism frequencies in the absence or presence of HLA-B*51 are marked, and the P values (Fisher exact test) and the most frequent variant amino acid are indicated.

FIG 3

CD8⁺ T cell response against the endogenously processed epitope IPFYGKAI_1373–1380. Effector T cells were expanded for 10 days from PBMCs in the presence of the peptide IPFYGKAI. HLA-B*51-positive target cells were generated by electroporation with a NS3(aa1330-2420)-GFP fusion protein. GFP-positive cells were sorted and used as targets for restimulation of IPFYGKAI-specific effector CD8⁺ T cells in an effector/target ratio of 1:1 for 4 h, followed by an intracellular cytokine staining (ICS). Mock transfected targets and peptide-pulsed targets served as negative or positive controls, respectively. (A) Representative fluorescence-activated cell sorting results of one ICS. (B) The IFN-γ response against targets transfected with GT1a prototype NS3 was normalized to 100% and compared to the response against other targets as indicated. The data represent results from at least three independent experiments. The P values were calculated using a one-sample t test with a hypothetical value of 100 (S1368P versus pt1a) or an unpaired t test (pt GT3a versus S1369P).

FIG 4

Impact of the S1368P substitution on proteasomal degradation. (A and B) The relative abundance of epitope-containing cleavage products after digestion of 25mer peptides with the prototype sequence (pt) or carrying the S1368P substitution (S1368P) with constitutive proteasome (A) or immune proteasome (B) is shown by normalized spectral indices. (C and D) The relative abundance of cleavage products with the correct C-terminal epitope end is shown by normalized spectral indices of individual peptides after digestion with constitutive proteasome (C) or immune proteasome (D).

FIG 5

Infectivity and replication of wild-type and mutant TNcc strains. RNA transcripts of the parental HCV TN genome and TN mutants were transfected into Huh-7.5 cells. (A) Seventy-two hours later, cell-free supernatant was used for inoculation of naive Huh-7.5 cells. The TCID₅₀ of the variants was determined by a limiting-dilution assay and staining with core and NS5A-specific antibodies. (B) RNA levels at 72 h posttransfection were measured by quantitative real-time reverse transcription-PCR. The RNA levels in the presence of 2′methyladenosin (2′CMA) was used for normalization. The ratio of measured HCV RNA to measured RNA levels in the presence of 2′CMA was calculated for each construct. The data from four independent experiments are shown.