Immunogenicity and cross-reactivity of a representative ancestral sequence in hepatitis C virus infection

Abstract

Vaccines designed to prevent or to treat hepatitis C viral infection must achieve maximum cross-reactivity against widely divergent circulating strains. Rational approaches for sequence selection to maximize immunogenicity and minimize genetic distance across circulating strains may enhance vaccine induction of optimal cytotoxic T cell responses. We assessed T cell recognition of potential hepatitis C virus (HCV) vaccine sequences generated using three rational approaches: combining epitopes with predicted tight binding to the MHC, consensus sequence (most common amino acid at each position), and representative ancestral sequence that had been derived using bayesian phylogenetic tools. No correlation was seen between peptide-MHC binding affinity and frequency of recognition, as measured by an IFN-γ T cell response in HLA-matched HCV-infected individuals. Peptides encoding representative, consensus, and natural variant sequences were then tested for the capacity to expand CD8 T cell populations and to elicit cross-reactive CD8 T cell responses. CD8(+) T cells expanded with representative sequence HCV generally more broadly and robustly recognized highly diverse circulating HCV strains than did T cells expanded with either consensus sequence or naturally occurring sequence variants. These data support the use of representative sequence in HCV vaccine design.

Figure 1. Lack of correlation between HLA binding affinity and frequency of recognition

The frequency of epitope recognition among 60 HCV infected subjects with matching HLA allele and subtype was assessed. HLA binding affinity was determined in an in vitro assay using purified HLA molecules with known allele and subtype. The percent of subjects recognizing HLA-matched epitopes as derived in Table I was graphed relative to the HLA binding affinity for (A) all subjects and (B) subjects that clear infection.

Figure 2. Relative to the bole1a peptides, consensus peptides fail to or poorly expand T cells specific for peptides from either sequence

PBMC were stimulated for twenty days with peptides bearing either the bole1a or consensus cons1a sequences. Stimulating peptide sequence is indicated on each graph in the upper right corner. The resulting lines were tested against titrated concentrations of peptides encoding the (--•--) bole1a or cons1a sequence (--○--) in duplicate wells. (A) HLA A*01 HCV NS3 1436-1444, subject 65 (B) HLA C*07 HCV E2 610-618, subject 109 (C) Summation of the total magnitude of response (Total SFC/1e6 cells).

Figure 3. Relative to peptides encoding bole1a sequence, peptides encoding variant sequences fail to or poorly expand T cells specific for either bole1a or variant sequence

PBMC were stimulated for twenty days with a peptide of either the bole1a or an identified circulating sequence of that same epitope. In the upper right corner of each graph, the sequence used to expand is shown with amino acid differences from bole1a underlined. The resulting lines were tested against titrated concentrations of peptides in duplicate wells encoding the bole1a sequence (--•--) and all identified circulating variants of that epitope (open shapes) in an IFN-gamma ELISpot assay. (A) HLA B*07 HCV Core 41-49, subject 109 (B) HLA C*07 HCV E2 610-618, subject 109 (C) HLA A*02 HCV E2 614-622, subject 18 (D) HLA A*11 HCV E2 621-628, subject 109 (E) HLA A*29 HCV 97 790-799, subject 109 (F) HLA A*02 HCV NS3 1073-1081, subject 160 (G) HLA A*26 HCV NS3 1111-1120, subject 109 (H) Sum of the magnitude of ELISpot response at peptide dilutions of 10, 1, and 0.1uM (y-axis) when peptides encoding bole1a or a naturally circulating variant was used to expand the T cells (x-axis). The lines connect the bole1a epitope to its corresponding variants.

Figure 4. Bole1a reliably expands T cells with diverse cross-reactivity

PBMC were stimulated for twenty days with peptides bearing either the bole1a identified variant sequences. The resulting lines were tested against titrated concentrations of peptides encoding an HCV consensus sequence and identified variants in duplicate wells in an IFN-gamma ELISpot. (A) HLA A*01 HCV NS3 1436-1444, subjects 113 and 65 (B) HLA A*11 HCV E2 621-628, subjects 18 and 109 (C) HLA B*08 HCV NS3 1395-1403, subjects 109 and 181 (D) HLA A*26 HCV NS3 1111-1120, subjects 109 and 18 (E) HLA C*07 HCV E2 610-618, subjects 109 and 148 (F) HLA A*02 HCV NS3 1406-1415, subjects 26 and 18 (G) HLA A*02 HCV NS3 1073-1081, subjects 160 and 18.

Figure 5. Bole1a sequence contains the greatest number of optimal epitopes

The optimal epitope sequence was defined as the sequence that induced the largest response or within 2-fold response of the largest response on IFN-gamma ELISpot at a peptide concentration of 1uM. Lines generated that produced only responses less than 100 SFC/1E6 cells were excluded from consideration. Full-length genotype 1a sequences culled from GenBank (n=390), cons1a, and bole1a sequences were then screened for the presence of the optimal epitopes and the total number of optimal epitopes per sequence was tallied.