Absence of viral escape within a frequently recognized HLA-A26-restricted CD8+ T-cell epitope targeting the functionally constrained hepatitis C virus NS5A/5B cleavage site

Footnotes

A table with details of the study population and a figure showing that cleavage at the NS5A/5B cleavage site is highly sensitive to mutations at the P1 and P1' positions are available as supplementary material with the online version of this paper.

Recently, an association between expression of particular MHC class I alleles and the natural course of HCV infection has been described. For example, HLA-B27 is associated with spontaneous viral clearance, whilst HLA-B8 is associated with HCV persistence (McKiernan et al., 2004). We have recently studied the virus-specific CD8⁺ T-cell response and virus evolution in the context of HLA-B27 and HLA-B8, and found that, for both HLA alleles, viral escape within dominant virus-specific CD8⁺ T-cell epitopes is a common feature in chronically infected patients (Neumann-Haefelin et al., 2006; Timm et al., 2004).

HLA-A2 has not been found to be associated with a specific outcome of infection, even though most studies about the role of viral escape have been biased towards this allele and have focused on previously described epitopes (Chang et al., 1997; Spangenberg et al., 2005; Urbani et al., 2005). The frequency of viral escape seems to be different for individual HLA-A2 epitopes. For example, selection of escape mutations within the epitope NS3 1406 (KLVALGINAV) has been described, whereas they are rare in the epitopes NS5B 2594 (ALYDVVTKL) and NS3 1073 (CINGVCWTV) (Chang et al., 1997; Kantzanou et al., 2003; Spangenberg et al., 2005). For the latter, substantial constraints on the ability to accommodate sequence variation due to fitness costs have been suggested (Soderholm et al., 2006).

The aim of this study was to characterize further the determinants of viral escape and its relative contribution to CD8⁺ T-cell failure and viral persistence. In order to exclude a methodological bias by using previously described (but not necessarily immunodominant) epitopes, we used a comprehensive approach to determine the full breadth and hierarchy of the CD8 response and its impact on virus evolution in acute and chronic HCV infection in HLA-A26⁺ patients. Importantly, the relatively rare HLA allele A26 has not been shown previously to restrict HCV-specific CD8⁺ T-cell responses.

First, we performed a comprehensive analysis of the CD8⁺ T-cell response in three HLA-A26⁺ patients with acute, self-limited HCV infection (patients A1A3; see Supplementary Table S1, available in JGV Online, for patients' characteristics), as well as six HLA-A26⁺ patients with chronic HCV infection (patients C1C4, C7 and C13; see Supplementary Table S1). A screening ELIspot assay was performed in a matrix set-up, using 441 overlapping peptides spanning the complete HCV polyprotein (18-mers overlapping by 11 aa, derived from HCV strain H77, kindly provided by the NIH Reference and Reagent Program). Positive peptides were confirmed by intracellular gamma interferon (IFN-γ) staining as described previously (Neumann-Haefelin et al., 2006). Positive overlapping peptides were screened by an epitope-prediction program () and given a score, leading to the identification of four novel HLA-A26-restricted HCV-specific CD8⁺ T-cell epitopes (Table 1). The predicted optimal epitopes were confirmed by serial dilution in comparison with the corresponding overlapping peptide; HLA-A26 restriction was confirmed by functional assays using autologous and partially HLA-matched EpsteinBarr virus (EBV)-infected cell lines [representative data for epitope NS5 2416 are shown in Fig. 1(a)]. Binding assays (Ruppert et al., 1993) showed that all four epitopes had a good HLA-A26-binding capacity (IC₅₀, <500 nM), with the NS5 2416 epitope displaying the best binding (IC₅₀, 25 nM; Table 1). Next, we tested the four novel HLA-A26-restricted epitopes in a larger cohort of chronically HCV genotype 1-infected HLA-A26⁺ patients (in patients C1C3, infected with genotype 1a, peptides with the genotype 1a consensus sequence were used, whilst in patients C4C8, infected with genotype 1b, peptides with the genotype 1b consensus sequence were used; see Table 1 for details). Importantly, one of the four epitopes, NS5 2416, was targeted in all (three of three) patients with acute, resolving HCV infection (genotype 1) and in two of eight patients with chronic HCV infection (genotype 1). The sequence-matched peptide with a D2416S substitution was also tested in six patients infected with genotype 3 (patients C9C14); two of the six patients targeted the epitope. Of note, a similar frequency has been described recently for immunodominant HLA-A2- and HLA-B27-restricted epitopes in chronic HCV infection (Lauer et al., 2004; Neumann-Haefelin et al., 2006; Spangenberg et al., 2005), indicating that the NS5 2416 epitope is recognized frequently in the background of HLA-A26⁺ patients. Interestingly, this response was detectable directly without previous peptide-specific stimulation in patients with acute, resolving infection, but only after peptide-specific stimulation in patients with chronic HCV infection; in addition, peptide-specific cell lines remained reproducibly negative in patient A1, despite the presence of a detectable response directly ex vivo. These results support previous findings that some HCV-specific T-cell responses display an impaired proliferative capacity (Urbani et al., 2002; Wedemeyer et al., 2002).

Table 1. HLA-A26-restricted CD8+ T-cell responses in patients with acute, resolving or chronic HCV infection Values are number of IFN-γ+ CD8+ cells in CD8+ PBMCs/number of IFN-γ+ CD8+ cells in a peptide-specific cell line (%). indicates no response detectable (<0.02 %). ND, Not determined.

Taken together, our findings suggest that strong functional constraints in the NS3/4A and especially in the NS5A/5B cleavage sites may prevent mutational escape in HLA-A26-restricted epitopes, and illustrate the limits of virus evolution in the presence of selection pressure by a strong and functionally intact immune response. This supports the hypothesis that selection towards a consensus sequence providing optimal viral fitness is a major player acting as a driving force against evolution of a highly mutable RNA virus (Altman & Feinberg, 2004). It is important to note that, in addition to the hypothesis suggested above, host factors, such as the genetic restriction of the immune response (Neumann-Haefelin et al., 2006), T-cell receptor diversity (Meyer-Olson et al., 2004) or strong CD4⁺ T-cell help (Grakoui et al., 2003), may also contribute to the observed lack of viral escape. However, in the presence of viral factors constraining evolution of viral mutations CD8 escape will not occur, despite a strong and functional virus-specific CD8⁺ T-cell response. These findings have important implications for vaccine design, as the identification of CD8⁺ T-cell epitopes in a virus is an important prerequisite for the development of an effective therapeutic vaccine. Targeting of conserved viral regions with strong functional constraints should result in stable control of virus replication.