Broad Repertoire of the CD4⁺ Th Cell Response in Spontaneously Controlled Hepatitis C Virus Infection Includes Dominant and Highly Promiscuous Epitopes

Subjects

Twenty-two subjects with spontaneously controlled HCV infection and 23 subjects with chronic HCV infection gave written informed consent and were enrolled in this study which was approved by the Institutional Review Boards of the Massachusetts General Hospital and the Lemuel Shattuck Hospital. All subjects were anti-HCV positive as measured by third-generation enzyme immunoassay, and no subject had been treated with HCV antiviral therapy. HCV RNA was measured by the Roche Amplicor Monitor assay (detection limit of 600 HCV RNA IU/ml plasma). All individuals with spontaneously controlled HCV viremia had documented undetectable plasma viral loads (<600 IU/ml plasma) at multiple time points for >7 mo (range, >7 mo to >3 years). The HCV viral loads of the individuals with chronic HCV infection ranged from 2240 to >850000 IU/ml plasma (HCV RNA positive >1 year). Serum alanine transaminase levels (ALT) were normal (<28 U/L) in individuals with resolved infection, with the exception of subject R1, who had intermittent mild elevation in ALT levels (range, 23–72 U/L) attributed to chronic hepatitis B virus infection. Subjects R16, R17, R18 were HIV positive, with HIV viral loads <50 copies/ml for >3 mo and CD4⁺ counts >350/μl (patients R17 and R18 were being treated with highly active antiretroviral therapy; R16 had never been on HIV antiretroviral therapy). In subjects with spontaneously resolved infection, serotyping for the HCV genotype was performed using the Murex HCV Serotyping 1–6 assay (Abbott) according to the manufacturer’s instructions except in cases where the genotype had already been documented during acute infection. There was no serum or plasma available for subject R9. Data from subjects R1–3 have been described (22). Twenty-three subjects with persistent HCV viremia served as a control group.

Synthesis of HCV-derived peptides

Peptides corresponding to the amino acid sequences of the HCV-1a strain HCV-1 (GenBank accession number: M62321) were synthesized as free acids using the 9-fluorenylmethoxy carbonyl method. A total of 301 peptides spanning the entire polyprotein were used for the screening. These peptides were 20 aa in length, overlapping adjacent peptides by 10 aa (22).

Recombinant HCV proteins

The recombinant HCV proteins used in this study were expressed as C-terminal fusion proteins with human superoxide dismutase in Saccharomyces cerevisiae or Escherichia coli and were kindly provided by M. Houghton (Chiron, Emeryville, CA). These proteins were derived from the HCV-1 sequence and encoded core (C22-3 aa 2–120), NS3 (C33C aa 1192–1457), NS4 (C100.3 aa 1569–1931), NS3/NS4 (C200 aa 1192–1931), and NS5 (NS5 aa 2054–2995).

HLA typing

DNA for HLA typing was extracted using the Puregene DNA Isolation kit for blood (Gentra Systems) according to the manufacturer’s instructions. HLA class II typing was performed at the MGH Tissue Typing Laboratory and by Dynal Biotech HLA Diagnostics using sequence-specific primer PCR (24, 25).

Proliferation assays

Lymphocyte proliferation assays were performed as described previously (22, 26), using c22.3, c33c, c100.3, C200, and NS5 HCV proteins (Chiron) at concentrations of 10 μg/ml. PBMC were isolated from fresh heparinized blood by Ficoll-Hypaque-density gradient centrifugation and plated at 100,000 cells/well in 96-well U-bottom plates (Costar) in 200 μl of R10/HAB medium (RPMI 1640 (Sigma-Aldrich), 10% human AB serum (Sigma-Aldrich), and 10 mM HEPES buffer (Sigma-Aldrich) with 2 mM glutamine and antibiotics (penicillin-streptomycin, 50 U/ml)) and the designated proteins in quadruplicate wells. After a 6-day incubation at 37°C and 5% CO₂, wells were pulsed for 6 h with 1 μCi of [³H]thymidine (PerkinElmer). Cells were then collected on filters, and the amount of incorporated radiolabel was measured with a beta counter (22).

In vitro HLA-DR peptide binding assays

Frequently targeted peptides were tested for in vitro binding to up to 15 common HLA-DR molecules. HLA-DR molecules were purified, and binding assays were performed as previously described (27). Purified human HLA-DR molecules were incubated with unlabeled HCV peptides and 1–10 nM ¹²⁵I-radiolabeled probe peptides for 48 h. MHC binding of the radiolabeled peptide was determined by capturing MHC-peptide complexes on LB3.1 (anti-HLA-DRA) Ab-coated Lumitrac 600 plates (Greiner Bioone) and measuring bound cpm using the TopCount (Packard Instrument) microscintillation counter. The binding data were analyzed and IC₅₀ (nanomolar) determined as previously described (27, 28).

Bulk stimulation of PBMC

Twenty million fresh or thawed PBMC were depleted of CD8⁺ cells with magnetic beads according to the manufacturer’s instructions (M-450 CD8 beads; Dynal Biotech) and the CD8⁺-depleted cells were stimulated with 1 μg/ml recombinant HCV Ags (either C22–3, C33C, C100.3, C200, or NS5) in 2 ml of R10/50 medium (RPMI 1640 medium, 10 mM HEPES with 2 mM glutamine and antibiotics (penicillin-streptomycin, 50 U/ml)) and 10% FCS (Sigma Aldrich) supplemented with recombinant IL-2 (rIL-2) (50 U/ml)) (22). Alternatively, for the regions of the HCV genome for which recombinant Ags were not available (E1, E2, p7, and NS2) PBMC were stimulated with overlapping 20-mer peptides (1 μg/ml each) in pools of 10 peptides (22). After 11–13 days cells were assayed for IFN-γ production by ELISPOT and intracellular cytokine staining in response to stimulation with HCV peptides. In fine-mapping and restriction experiments, HCV-specific CD4⁺ T cell lines were directly used if the specificity of the cell line was ≥3%. If the specificity was <3%, peptide-specific cells were selected by an IFN-γ capture assay (Miltenyi Biotec), as previously described (29). Briefly, the cells were stimulated with 10 μg of peptide/ml and costimulatory CD28 and CD49d mAbs (1 μg/ml; BD Biosciences). After incubation for 5 h at 37°C, IFN-γ was processed according to the manufacturer’s protocol and isolated on a magnetic cell sorting column. Selected cells were expanded with IL-2 and irradiated allogenetic feeder cells for an additional 10–14 days and then used in HLA restriction experiments.

ELISPOT assay

ELISPOT assays were performed as previously described (30). Polyvinylidene plates (96-well; Millipore) were coated with 2.5 μg/ml recombinant anti-human IFN-γ Ab (Endogen) in PBS at 4°C overnight. Peptides were directly added to the wells at a final concentration of 10 μg/ml–0.01 ng/ml. Cells from polyclonal HCV-specific CD4⁺ T cell cultures were taken out of R10/50 by washing three times with media and incubating for 6 h at 37°C in R10 and were then added at 50,000 cells/well in 140 μl of R10 medium. The plates were incubated for 18 h at 37°C in 5% CO₂. Plates were then washed, labeled with 0.25 μg/ml biotin-labeled anti-IFN-γ (Endogen), and developed by incubation with streptavidin-alkaline phosphatase (Bio-Rad), followed by incubation with 5-bromo-4-chloro-indolylphosphate-NBT (Bio-Rad) in Tris buffer (pH 9.5). The reaction was stopped by washing with tap water, and the plates were dried overnight, before counting on an ELISPOT reader (AID). Responses were considered positive if the number of spots was three times greater than the background; PHA served as a positive control. All positive responses were reconfirmed by intracellular cytokine assay (ICS) following stimulation with the specific peptide.

Intracellular IFN-γ staining by flow cytometry

Intracellular cytokine staining was performed as previously described (22, 31). Briefly, 5 × 10⁵ bulk expanded PBMC were added to 10⁵ autologous B cells and incubated with 4 μM peptide and anti-CD28 and anti-CD49d mAb at 37°C and 5% CO₂ for 1 h before the addition of GolgiPlug (1 μl/ml; BD Biosciences). The cells were incubated for an additional 5 h at 37°C and 5% CO₂. PBMC were then washed and stained with surface Abs, allophycocyanin-conjugated anti-CD3, and peridin-chlorophyll protein-conjugated anti-CD4 (BD Biosciences) at room temperature for 20 min. Following washing, the PBMC were fixed and permeabilized (Caltag Laboratories), and the FITC-conjugated anti-IFN-γ mAb (BD Biosciences) was added. Cells were then washed and analyzed on a FACSCalibur flow cytometer using CellQuest software (BD Biosciences). Single peptides that elicited distinct responses of more than three times background IFN-γ production were considered positive (22).

HLA restriction experiments

HLA restriction experiments were performed as described before (22, 29). Autologous and partially HLA-matched EBV-transformed B-lymphoblastoid cell lines (BCL) and L cell lines (LCL) stably transfected with HLA class II molecules were pulsed with 0.4 μg of peptide/ml for 90 min at 37°C and 5% CO₂. HLA-DR-transfected cell lines were kindly donated by Epimmune. The peptide-pulsed BCL or LCL were then washed four times, incubated together with peptide-specific cells at a ratio of 1:10, and stained as described above. The peptide-specific CD4⁺ T cell response was analyzed by intracellular cytokine staining for IFN-γ production as described above.

Statistical analysis

Nonparametric Mann-Whitney tests were performed using GraphPad Prism 4.0 software (GraphPad Software). To avoid overestimation of the total breadth of antiviral responses, responses to adjacent overlapping peptides were counted as responses to 1 epitopic region, with only the stronger response shown.

Immunodominance of selected peptides located within NS3, NS4, and NS5 in spontaneously resolved HCV viremia

Spontaneous resolution of HCV infection has been shown to be associated with a vigorous lymphocyte proliferative response (8, 10, 32). Therefore, we stimulated PBMC with recombinant HCV core (c22.3), NS3 (c33c), NS4 (c100.3), and NS5 proteins in standard lymphoproliferative assays (LPA) in all 22 subjects with resolved HCV infection and 23 subjects with chronic HCV infection (for clinical and virological data for the cohort see Table I⇓). There was a significant difference in the magnitude of the lymphoproliferative responses to all three nonstructural proteins when compared with persons with unresolved infection (Fig. 1⇓). The lowest magnitude of responses in persons with resolved infection was to the relatively small core protein, and these responses were not significantly different from persons with chronic infection, although there was a trend toward significance (p = 0.08). Proliferative responses were infrequently detected in persons with chronic infection (Fig. 1⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Comparison of lymphoproliferative responses to recombinant HCV Ags in subjects with spontaneously resolved and chronic infection. LPAs with recombinant HCV Ags expressing the HCV proteins: core (c22.3), NS3 (c33c), NS4 (c100.3), and NS5 were performed on PBMC of 22 spontaneously resolved and 23 chronically HCV infected subjects as described in Materials and Methods. Eight uninfected seronegative subjects served as controls. In contrast to the weak proliferative responses displayed by subjects with chronic HCV infection, multiple regions of HCV were targeted by subjects with spontaneously resolved HCV, with recombinant NS3, NS4, and NS5 proteins each inducing similar strong proliferative responses (NS3 mean stimulation index (SI) = 18.3, range 0.5–101; NS4 mean SI = 18.9, range 1.2–124; NS5 mean SI = 17, range 1–100.3). The recombinant core protein induced the weakest average proliferative response in subjects with resolved HCV (mean SI = 6.6, range 0.7–46).

We next performed a more detailed analysis of each subject to define the fine specificities of the HCV CD4⁺ T cell epitopes targeted within each of the HCV proteins by testing HCV-specific polyclonal CD4⁺ T cell cultures that were generated with slight modifications of the protocol described previously (Ref.22 , see also Materials and Methods). Cells were initially tested with separate pools of 10 overlapping peptides covering the whole HCV polyprotein after 11–13 days. Positive pools were deconvoluted the next day by single peptide ELISPOT. Single-positive responses in ELISPOT were then confirmed in a single peptide intracellular cytokine staining assay for specific IFN-γ production.

The 22 subjects with resolved HCV infection demonstrated broadly distributed responses (mean 10; range 3–28) against different 20-mer peptides (Fig. 2⇓A) and most subjects (21 of 22) each showed one or more responses in at least three of four of the nonstructural proteins (NS2, NS3, NS4a/b, or NS5a/b). In all cases (22 of 22) of resolved viremia one or more epitopes in NS3 were targeted (Fig. 2⇓B) whereas responses were less frequently detected in the highly conserved, but small, core protein (consistent with the data generated in LPA assays), and in the variable regions of the HCV envelope proteins (Figs. 2⇓B and 3⇓A) (33, 34, 35, 36).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Spontaneous resolution of HCV viremia is associated with recognition of multiple CD4⁺ T cell epitopes. PBMC from 22 subjects with resolved HCV infection and 23 subjects with chronic HCV infection were CD8⁺-depleted and stimulated with recombinant Ags (or overlapping peptides) covering the entire polyprotein in the presence of recombinant IL-2. Expanded cell lines were tested for IFN-γ production upon stimulation with pools of 10–20 overlapping peptides by ELISPOT. All peptide pools eliciting a positive response by IFN-γ production were further deconvoluted to identify specific 20-mer peptides targeted by the CD4⁺ T cell response. Each single response detected in the ELISPOT assay was confirmed by ICS for IFN-γ production as well. Horizontal bars indicate the median numbers of targeted HCV CD4⁺ T cell responses against single peptides for each subject as tested by ICS. The number of responses is significantly higher (p < 0.0001) in persons with resolved HCV than in chronically infected persons when comparing the number of CD4⁺ T cell responses detected in short-term lines across the entire polyprotein (resolvers: mean 10, range 3–29; chronic infection: mean 1, range 0–8) (A), and when comparing the number of responses in HCV resolvers (R) or chronically infected (C) analyzed and plotted for each of the proteins E1/E2, NS2, NS3, NS4, and NS5 individually (B). Although recombinant core protein induced more responses in persons with resolved HCV than in chronically infected persons this trend did not reach significance.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Frequency of detection of HCV peptides in HCV-specific CD4⁺ T cell cultures and localization of these specificities within the HCV polyprotein of 22 subjects with spontaneously resolved HCV infection (A) and 23 chronically HCV-infected subjects (B).

Altogether, 89 of the 301 overlapping peptides tested were targeted by at least one person (Fig. 4⇓). Whereas some peptides were only recognized by 1–2 subjects, each of the 24 peptides shown in Table II⇓ was recognized by at least 3 of the 22 subjects with resolved infection, even though in some cases these subjects did not share class II molecules (see below).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Summary of all the specificities recognized by each individual with resolved HCV infection. Shaded boxes indicate a positive CD4⁺ response. Black boxes represent the strongest CD4⁺ response in a single subject. Peptides: p2 = aa 21–40; p3 = aa 31–50; p294 = aa 2941–2960.

The two most frequently detected specificities in persons with resolved infection, p124 (aa 1241–1260) and p177 (aa 1771–1790) (situated in NS3 and NS4, respectively) were each recognized by >50% of the subjects with resolved HCV infection. These two peptides contain CD4⁺ T cell epitopes that have previously been described in smaller and less comprehensive studies as frequently recognized epitopes (37, 38). An additional four peptides were targeted in >30% of persons with resolved infection (p153 (aa 1531–1550), p158 (aa 1581–1600), p257 (aa 2571–2590) and p294 (aa 2941–2960)), all of which are located within one of the different HCV nonstructural proteins. Each of the 22 subjects with spontaneously resolved infection targeted at least one (mean 2.5; range 1–4) of these six most frequently targeted peptides.

In addition to defining the breadth of the targeted epitopes, we also examined the dominant response in each subject with resolved viremia, defined as the peptide that induced the strongest IFN-γ response as measured by single peptide ICS in expanded cells (Fig. 4⇑). Even though peptide p124 (aa 1241–1260) is one of the most frequently recognized peptides in persons with resolved HCV, this peptide was found to be the dominant response in only 1 of 22 subjects. In contrast, p257 (aa 2571–2590) was the dominant response in HCV-specific short-term lines of 6 of the 10 subjects that recognize this peptide.

In the comprehensive high-resolution screening of HCV-specific polyclonal cell lines, only 7 of the 23 subjects with chronic infection showed one or more positive HCV-specific CD4⁺ T cell response, and these responses were narrowly targeted against few peptides (average 1, range 0–8) (Fig. 3⇑B). Of those epitopes targeted in chronic infection, all were also detected in resolved infection.

Frequently detected HCV specificities show degenerate binding in in vitro assays

The finding that certain peptides were recognized by a large proportion of subjects with resolved HCV infection, despite the fact that they expressed different HLA class II molecules, raised the possibility of promiscuous HCV CD4⁺ T cell epitopes that are recognized in the context of multiple HLA class II molecules. We first tested this hypothesis by analyzing the capacity of frequently recognized HCV peptides to bind to a panel of up to 15 of the most prevalent human HLA-DR molecules in an in vitro assay (29, 39, 40). Peptides with an IC₅₀ of 1000 nM or lower were considered to bind to the prospective DR molecule since this has been shown in previous studies to represent an affinity threshold associated with immunogenicity in the context of HLA-DR molecules (28, 29). These most frequently recognized peptides were highly promiscuous in binding to multiple HLA-DR molecules with each of these six peptides binding between 7 and 12 different molecules (Table III⇓). Furthermore, some of these peptides were able to bind to several HLA-DR molecules with extremely high affinities (IC₅₀ <10 nM). These results indicate that HCV peptides capable of binding multiple HLA-DR molecules with high affinity are likely to be frequently recognized by subjects with resolved infection.

Frequently targeted peptides are restricted by multiple HLA class II molecules and share largely overlapping binding repertoires

The above binding data as well as the diverse HLA class II molecules expressed in persons targeting the same peptides (Fig. 4⇑) suggest that these peptides may be promiscuously presented by multiple class II molecules. To address this experimentally, cells were stimulated with peptide-pulsed partially HLA-matched BCL and/or the corresponding HLA-DR-transfected fibroblasts in an assay described previously (22, 29). Representative data demonstrating HLA class II restriction of a p153 (aa 1531–1550)-specific cell line in subject R8 are shown in Fig. 5⇓, demonstrating recognition of this epitope in the context of HLA DRB1*0701, an allele to which p153 (aa 1531–1550) showed strong binding in the in vitro assay (Table III⇑).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

HCV-specific CD4⁺ T cell response against p153 (aa 1531–1550) is restricted by HLA-DRB1*0701 in subject R8. Peptide-pulsed (0.4 μg/ml autologous and partially HLA-matched B cells, as well as l cells L650.2 (DRB 1*1302), L605.1 (DRB1*0701), and L576.5 (DRB3*0301)) were incubated with CD4⁺ T cell lines from subject 1 for 6 h and analyzed by flow cytometry for IFN-γ production as described in Materials and Methods. Non-peptide-pulsed B cells and l cells were used as a control (data not shown).

Of the six most frequently targeted peptides in persons with resolved HCV infection, restricting HLA class II molecules were demonstrated for each epitope, and for three of these multiple restricting class II molecules were readily defined. As an example, peptide p124 (aa 1241–1260) overlaps with a previously described and well-defined highly promiscuous CD4⁺ T cell epitope. This epitope has been previously described to be restricted by at least five different HLA-DR or DQ molecules (37, 38, 41), and additionally was found to be restricted by HLA-DR1*1001 in subject R2 and HLA-DRB3 0301 in subject R8. Thus together with previous data, this peptide can be presented in the context of at least seven different class II molecules, further emphasizing the promiscuity of this peptide (Table IV⇓).

Data from all restriction experiments that we performed on individual subjects are summarized in Table IV⇑. Of note, using this experimental approach, which incorporates functional FACS analysis with partially HLA-matched cell lines, all HCV HLA class II epitopes for which HLA restrictions could be defined were presented by HLA-DR molecules and not by DQ molecules.

The above data indicate that the 20-mer peptides can be presented in the context of multiple class II molecules, suggesting that either the epitopes within the longer peptides are identical, or that the peptides contain multiple overlapping epitopes. The generation of CD4⁺ T cell lines provided a means to define the minimal (and optimal) length of some frequently recognized HCV 20-mer peptides in individual subjects with different HLA class II molecules. For the fine-mapping analysis of the six most frequently recognized peptides, we defined the “minimal-epitope” as the smallest peptide triggering a response at least 50% as strong as that observed with the full-length peptide in the ICS assay. This definition is consistent with criteria that have been used in studies of HIV and CMV-specific CD4⁺ T cell responses (29, 42). We confirmed the results from the ICS experiments by conducting ELISPOT dilution assays (data not shown).

The effects of serial truncations on the magnitude of the IFN-γ responses are illustrated in Fig. 6⇓ which shows representative ICS results from subjects R1, R3, R12, and R14, all of whom targeted peptide p177 (aa 1771–1790) in the context of different HLA class II molecules. Serial N-terminal and C-terminal truncations showed an almost identical pattern in all four subjects, suggesting that the minimal epitope was the 11-mer ¹⁷⁷⁵LAGLSTLPGNP.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Defining the fine specificity of the response against peptide p177 (aa 1771–1790) in cell lines from multiple subjects with resolved HCV. Representative ICS data from CD4⁺ T cell lines of subjects R12, R3, R1, and R14 (the restricting HLA DR molecule is shown for each subject). Cells were stimulated with 0.4 μg/ml peptide p177 (aa 1771–1790) or its truncations 177A-J. The minimal region essential for recognition of this epitope is highlighted in gray.

Optimal responses for this and most of the other epitopes we analyzed were seen with peptide truncations flanking (43) the minimal epitope that induced IFN-γ production by two to four N- and C-terminal residues (Table IV⇑). For some peptides (p227 (aa 2271–2290), p242 (aa 2421–2440)) the minimal epitope was only narrowed down to 14 aa. For many applications like the construction of new class II tetramers (23), the knowledge of the peptide which triggers the optimal response rather than the minimal epitope may have the greatest relevance.

The data from the functional truncation experiments performed in this study are supported in part by commonly used prediction algorithms (44, 45, 46), but there are multiple peptides for which the functional data indicate the need for a longer peptide than would be predicted (Table IV⇑).