CD8+ T lymphocyte responses target functionally important regions of Protease and Integrase in HIV-1 infected subjects

Abstract

BACKGROUND: CD8+ T cell responses are known to be important to the control of HIV-1 infection. While responses to reverse transcriptase and most structural and accessory proteins have been extensively studied, CD8 T cell responses specifically directed to the HIV-1 enzymes Protease and Integrase have not been well characterized, and few epitopes have been described in detail. METHODS: We assessed comprehensively the CD8 T cell responses to synthetic peptides spanning Protease and Integrase in 56 HIV-1 infected subjects with acute, chronic, or controlled infection using IFN-gamma-Elispot assays and intracellular cytokine staining. Fine-characterization of novel CTL epitopes was performed on peptide-specific CTL lines in Elispot and 51Chromium-release assays. RESULTS: Thirteen (23%) and 38 (68%) of the 56 subjects had detectable responses to Protease and Integrase, respectively, and together these targeted most regions within both proteins. Sequence variability analysis confirmed that responses cluster largely around conserved regions of Integrase, but responses against a large, highly conserved region of the N-terminal DNA-binding domain of Integrase were not readily detected. CD8 T cell responses targeted regions of Protease that contain known Protease inhibitor mutation residues, but strong Protease-specific CD8 T cell responses were rare. Fine-mapping of targeted epitopes allowed the identification of three novel, HLA class I-restricted, frequently-targeted optimal epitopes. There were no significant correlations between CD8 T cell responses to Protease and Integrase and clinical disease category in the study subjects, nor was there a correlation with viral load. CONCLUSIONS: These findings confirm that CD8 T cell responses directed against HIV-1 include potentially important functional regions of Protease and Integrase, and that pharmacologic targeting of these enzymes will place them under both drug and immune selection pressure.

Figure 1

CD8 T cell responses to HIV-1 Protease. PBMC were stimulated with the indicated peptide in an overnight IFN-γ ELISPOT assay. Each row represents an individual peptide. (A) Amino acid sequence, using standard single-letter amino acid abbreviations. Numbers above peptide sequences refer to the amino acid position within Protease, with key Protease inhibitor mutation residues indicated in bold. (B) Bars represent the percentage of 56 study subjects who responded to the peptide in an ELISPOT assay. Peptides with the highest number of responses are shaded gray. (C) The magnitude of every CTL response detected in the study cohort. Each symbol represents a single CTL response against that peptide by one individual. Magnitudes of responses are shown after subtraction of background, which in all cases was <30 SFC/million PBMC. Closed circle (●): acute cohort. Open circle (○): chronic cohort. Shaded triangle (): controllers.

Figure 2

CD8 T cell responses to HIV-1 Integrase. PBMC were stimulated with the indicated peptide in an overnight IFN-γ ELISPOT assay. Each row represents an individual peptide. (A) Amino acid sequence, using standard single-letter amino acid abbreviations. Numbers above peptide sequence indicate amino acid position within Integrase. The conserved HHCC zinc finger-like domain and DDE element are indicated in bold. (B) Bars represent the percentage of 56 study subjects who responded to the peptide in an ELISPOT assay. Peptides with the highest number of responses are shaded gray. (C) The magnitude of every CTL response against Integrase peptides detected in the study cohort. Each symbol represents a single CTL response against that peptide by one individual. Magnitude of responses are after subtraction of background, which in all cases was <30 SFC/million PBMC. Symbols are the same as in Figure 1: Closed circle (●): acute cohort. Open circle (○): chronic cohort. Shaded triangle (): controllers.

Figure 3

Fine-mapping of one novel epitope within Protease and two within Integrase. Peptide-specific CD8 cell lines were generated for three peptides, Protease 6, Integrase 17, and Integrase 29/30. PBMC collected from subjects with strong responses by ELISPOT were expanded using a bispecific CD3/4 antibody. Following expansion, peptide-specific cells were collected using an IFN-γ catching assay after stimulation with the appropriate peptide. Peptide specificity was confirmed by flow cytometry. HLA-restriction was then determined using peptide-pulsed target cells matched at only one MHC class I allele in a ⁵¹Cr-releasse assay at an E:T ratio of 10:1; peptide-pulsed autologous cells were used as a positive control. The sequences of the optimal epitopes were also determined by testing peptide-specific cell lines against serial dilutions of truncations of the original peptide in an ELISPOT assay. Data are shown for three epitopes: (A, B) – Protease 6. (C, D) – Integrase 17. (E, F) – Integrase 29/30.

Figure 4

Correlation of amino acid sequence variability with frequency of CD8 T cell responses targeting Protease. For Protease, amino acid sequences were obtained from at the HIV-1 Molecular Immunology Database (27), and aligned relative to the HIV-1 clade B consensus sequence. Entropy scores for each amino acid residue were calculated based on this alignment, smoothed over nine amino acids, and plotted for all sequences (n=155, blue line, left axis) and clade B sequences only (n=34, red line, left axis). Entropy scores of 1 correspond to 100% conserved residues, while lower scores (plotted here on an inverse scale) correspond to increasing sequence variability. The number of responses in the 56 study subjects against peptides containing each amino acid was also plotted (purple line, right axis) to correlate regions with high sequence variability with regions targeted by CD8 T cells. Spearman’s rank-order correlation coefficient was calculated to correlate CD8 T cell responses against sequence variability for each protein.

Figure 5

Correlation of amino acid sequence variability with frequency of CD8 T cell responses targeting Integrase. For Integrase, amino acid sequences were obtained from at the HIV-1 Molecular Immunology Database (27), and aligned relative to the HIV-1 clade B consensus sequence. Entropy scores for each amino acid residue were calculated based on this alignment, smoothed over nine amino acids, and plotted for all sequences (n = 155, blue line, left axis) and clade B sequences only (n = 34, red line, left axis). Entropy scores of 1 correspond to 100% conserved residues, while lower scores (plotted here on an inverse scale) correspond to increasing sequence variability. The number of responses in the 56 study subjects against peptides containing each amino acid was also plotted (purple line, right axis) to correlate regions with high sequence variability with regions targeted by CD8 T cells. Spearman's rank-order correlation coefficient was calculated to correlate CD8 T cell responses against sequence variability for each protein.

Figure 6

Frequency of CD8 T cell responses to HIV-1 proteins relative to protein size. The frequency of responses directed against Protease and Integrase in the study cohort are plotted against the size of the proteins, in number of amino acids. Published data from cohorts where the frequency of CD8 T cell responses against at least one HIV-1 protein are plotted for comparison (see text for references).