HLA-E-restricted, Gag-specific CD8+ T cells can suppress HIV-1 infection, offering vaccine opportunities

Abstract

Human leukocyte antigen-E (HLA-E) normally presents an HLA class Ia signal peptide to the NKG2A/C-CD94 regulatory receptors on natural killer (NK) cells and T cell subsets. Rhesus macaques immunized with a cytomegalovirus-vectored simian immunodeficiency virus (SIV) vaccine generated Mamu-E (HLA-E homolog)-restricted T cell responses that mediated post-challenge SIV replication arrest in >50% of animals. However, HIV-1-specific, HLA-E-restricted T cells have not been observed in HIV-1-infected individuals. Here, HLA-E-restricted, HIV-1-specific CD8 ⁺ T cells were primed in vitro. These T cell clones and allogeneic CD8 ⁺ T cells transduced with their T cell receptors suppressed HIV-1 replication in CD4 ⁺ T cells in vitro. Vaccine induction of efficacious HLA-E-restricted HIV-1-specific T cells should therefore be possible.

Figure 1.. Priming and cloning of HLA-E restricted RL9HIV-specific CD8+ T cells from HIV naïve donors.

(A) PBMCs from 9 HIV-1 seronegative HLA-A2 negative donors (HD 1 to 9) were stimulated with autologous activated dendritic cells and the RL9HIV peptide for 9 days. HLA-E restricted RL9HIV specific CD8+ T cells were identified using HLA-E-RL9 disulfide trapped tetramer (RL9HIV-(D)) for donors HD1 to 6 or (B) a HLA-E RL9 tetramer generated by UV exchange RL9HIV-(UV)) for donors HD 7 to 9. (C) Disulfide trapped RL9HIV-(D) tetramer+ cells from donor HD1 were sorted for single cell cloning, and positive clones were identified using disulfide trapped RL9HIV-(D) tetramer. (D) RL9 positive clones were CD94 negative. Clones were dual stained with RL9 and canonical VL9 signal peptide disulfide trapped HLA-E tetramers.

Figure 2.. Functional analysis of HLA-E restricted RL9HIV-specific CD8+ T cell clones.

(A) IFN-γ, TNF-α, CD107a/b and CD137 expression by 15 RL9 clones upon stimulation with K562 cells transduced with single chain trimer (SCT) disulfide trapped HLA-D-RL9 was assessed using flow cytometry-based readouts; responsive clones could be detected using multiple functional readouts. (B) Six positive clones were further assessed by comparison of stimulation with K562 transduced with HLA-E and pulsed with RL9 peptide, K562 transduced with SCT linked HLA-E RL9 (K562-E-RL9), K562 transduced with SCT disulfide-linked HLA-E RL9 (K562-D-RL9), and K562 transduced with SCT disulfide linked HLA-E VL9 (K562-D-VL9) as a negative control. Horizontal lines indicate means. Groups were analyzed by 2-way ANOVA with Tukey’s multiple comparisons. Data shown are representative of six donors and two independent experiments. (C) Blockade of TNF-α secretion and CD107 expression of clone p13c7 by addition of the canonical VL9 signal prior to stimulation with HLA-E transduced K562 cells pulsed with RL9 peptide. Data shown were from two independent experiments, analyzed by paired t test.

Figure 3.. Recognition of naturally presented RL9 epitope on HIV-1 virus infected cells and inhibition of HIV-1-infected targets by RL9 clones.

(A) 771.221 cell lines transfected with CD4 were infected with HIV-1NL4.3 virus and cultured alone or with RL9-specific CD8+ T cell clones at an E: T ratio of 1:1. Frequencies of HIV-infected cells (Gag p24+) and the percentage reduction in the frequency of Gag p24+ cells in the presence of the RL9 clones after 5 days of culture (calculated as described in Materials and Methods) are shown. (B) The CD137 activation marker was assessed on CD8+ T cell clones when co-cultured with HIV-1NL4.3 infected CD4.221 cells compared to un-infected CD4.221 cells. Furthermore, the CD137 expression of CD8+ T cell clones was assessed with the canonical HLA-E binding VL9 signal peptide in the assay, when cocultured with HIV-1NL4.3 infected CD4.221. A control EBV-specific CD8+ T cell clone generated from the same donor was included in the assay as a negative control. Horizontal lines indicate means. Data were analyzed using a non-parametric Wilcoxon signed rank test. (C) Purified autologous CD4+ T cells were stimulated with anti-CD3 for 3 days prior to HIV-1NL4.3 infection, then either cultured alone or with clones at E: T ratios of 1:1 and 5:1 for 5 days. Gag p24+ cells were gated on CD3+/CD8-/CD4+ and CD4- T cells. Data shown were from 3 experiments, were analyzed by unpaired t test with Welch’s correction.

Figure 4.. Specificity and function of RL9-specific TCRs transduced into J8 Jurkat and primary CD8+ T cells.

(A) Five TCR transduced J8 Jurkat cell lines were stained with disulfide trapped RL9HIV-(D) tetramer and the tetramer+ population was enriched by sorting. (B) CD69 and /or GFP expression were assessed on transduced J8 Jurkat cells (pre-labeled with CellTrace Violet to facilitate gating for subsequent analysis) on exposure to HLA-E transduced K562 cells pulsed with RL9 peptide, and by K562 cells transduced with SCT HLA-E-RL9 or SCT disulfide-linked HLA-D-RL9. Horizontal lines indicate means. Statistical analysis of the data was performed using a non-parametric Wilcoxon signed rank test. (C) Two RL9 TCRs, p9c1 and p13c7, were transduced into primary CD8+ T cells. CD8+ transductants were stained with disulfide trapped RL9HIV-(D) tetramer initially, washed with PBS, and then stained with anti-mouse TCR Vβ antibody, anti-CD8 and Live/Dead Fixable Aqua. (D) CD8+ transductants were co-cultured with HIV-1NL4.3-infected primary CD4+ T cells at E:T ratios of 1:1 and 5:1. Gag p24+ cells were gated on CD3+/CD8−/CD4+ and CD4− T cells. Bars indicate mean reduction in % of p24+ cells. CD8+ T cells transduced with an irrelevant TCR (IG4) were included as a negative control and CD8+ T cells transduced with a TCR recognizing the B*2705 restricted Gag KK10 epitope were included as a positive control when CD4+ T cells from B*2705+ donor were infected with HIV-1NL4.3 virus and used as targets. Statistical analysis of the data was performed by One-way ANOVA Kruskal-Wallis test. Data shown are representative of four independent experiments.