Putative immunodominant human immunodeficiency virus-specific CD8(+) T-cell responses cannot be predicted by major histocompatibility complex class I haplotype

Abstract

Recent studies of human immunodeficiency virus (HIV)-specific CD8(+) T cells have focused on responses to single, usually HLA-A2-restricted epitopes as surrogate measures of the overall response to HIV. However, the assumption that a response to one epitope is representative of the total response is unconfirmed. Here we assess epitope immunodominance and HIV-specific CD8(+) T-cell response complexity using cytokine flow cytometry to examine CD8(+) T-cell responses in 11 HLA-A2(+) HIV(+) individuals. Initial studies demonstrated that only 4 of 11 patients recognized the putative immunodominant HLA-A2-restricted p17 epitope SLYNTVATL, suggesting that the remaining subjects might lack significant HIV-specific CD8(+) T-cell responses. However, five of six SLYNTVATL nonresponders recognized other HIV epitopes, and two of four SLYNTVATL responders had greater responses to HIV peptides restricted by other class I alleles. In several individuals, no HLA-A2-restricted epitopes were recognized, but CD8(+) T-cell responses were detected to epitopes restricted by other HLA class I alleles. These data indicate that an individual's overall CD8(+) T-cell response to HIV is not adequately represented by the response to a single epitope and that individual major histocompatibility complex class I alleles do not predict an immunodominant response restricted by that allele. Accurate quantification of total HIV-specific CD8(+) T-cell responses will require assessment of the response to all possible epitopes.

FIG. 1

Assessment of p17 SLYNTVATL-specific IFN-γ production in 11 HLA-A2⁺ HIV-positive patients. PBMC from each patient were stimulated with 2 μg of SLYNTVATL peptide per ml and costimulatory antibodies as described in Materials and Methods. The number shown in each plot represents the percentage of CD3⁺ CD8⁺ CD69⁺ IFN-γ⁺ events, with background IFN-γ production subtracted. The background IFN-γ production was ≤0.02% in all patients except patients 1 (0.03%), 4 (0.15%), and 11 (0.07%). Responses equal to or greater than 0.05% above background are considered positive.

FIG. 2

Intracellular IFN-γ production from CD8⁺ T cells in response to HIV-1 peptide mixes. PBMC from HIV-seronegative (A) or HIV-seropositive (B) individuals were incubated with HIV-1 peptide mixes for 6 h in the presence of the costimulatory anti-CD28 and anti-CD49d antibodies and brefeldin A (final 5 h only). After incubation, the cells were surface stained as described in Materials and Methods. The number shown in the upper right corner of each plot represents the percentage of CD69⁺ IFN-γ⁺ CD3⁺ CD8⁺ cells responding to the indicated peptide mixture.

FIG. 3

Identification of individual peptide responses using a peptide pool matrix. PBMC from patient 10 were incubated in the presence of costimulatory antibodies and brefeldin A as described in Materials and Methods with 20 epitope pools (in bold), each containing up to 10 HIV-1 peptides (A). The percentage of IFN-γ production to each pool from CD3⁺ CD8⁺ CD69⁺ small lymphocytes is shown at the right and along the bottom. Those peptides that could potentially be recognized, as determined by the pool responses, are in gray boxes. The actual peptides recognized (B) are in red-bordered gray boxes. The plots in panel B show the CD8⁺ T-cell response to the peptide mixes and individual peptides (as identified through the matrix analysis) in patient 10, where the number shown in each box represents the percentage of CD3⁺ CD8⁺ CD69⁺ IFN-γ⁺ events. The response to each peptide is shown in the top of each plot, as is the peptide source (i.e., HIV protein) and sequence of the peptide. Background IFN-γ production in patient 10 was less than 0.01%.

FIG. 4

Additional HIV epitopes recognized in a p17 SLYNTVATL nonresponder and responder. Potentially recognized epitopes were identified by peptide matrix analysis in all 11 subjects. Single-peptide analysis was performed to determine the peptides recognized in each subject. Shown are those responses identified in a representative SLYNTVATL nonresponder (A) and SLYNTVATL responder (B). As in the previous figures, all events shown are CD3⁺ and CD8⁺ gated events. The percentages represent those events that are CD69⁺ and IFN-γ⁺ within the CD8⁺ T-cell subset. Peptide numbers correspond to those shown in Fig. 3A. Each plot also shows the HLA restriction, source of peptide, and sequence, consecutively, of the peptide tested. The HLA-A2-restricted CMV peptide pp65 NLVPMVATV was included in all experiments as a positive control to demonstrate that all HLA-A2⁺ patients recognized an HLA-A2-restricted peptide.