Mycobacterium tuberculosis peptides presented by HLA-E molecules are targets for human CD8 T-cells with cytotoxic as well as regulatory activity

Abstract

Tuberculosis (TB) is an escalating global health problem and improved vaccines against TB are urgently needed. HLA-E restricted responses may be of interest for vaccine development since HLA-E displays very limited polymorphism (only 2 coding variants exist), and is not down-regulated by HIV-infection. The peptides from Mycobacterium tuberculosis (Mtb) potentially presented by HLA-E molecules, however, are unknown. Here we describe human T-cell responses to Mtb-derived peptides containing predicted HLA-E binding motifs and binding-affinity for HLA-E. We observed CD8(+) T-cell proliferation to the majority of the 69 peptides tested in Mtb responsive adults as well as in BCG-vaccinated infants. CD8(+) T-cells were cytotoxic against target-cells transfected with HLA-E only in the presence of specific peptide. These T cells were also able to lyse M. bovis BCG infected, but not control monocytes, suggesting recognition of antigens during mycobacterial infection. In addition, peptide induced CD8(+) T-cells also displayed regulatory activity, since they inhibited T-cell proliferation. This regulatory activity was cell contact-dependent, and at least partly dependent on membrane-bound TGF-beta. Our results significantly increase our understanding of the human immune response to Mtb by identification of CD8(+) T-cell responses to novel HLA-E binding peptides of Mtb, which have cytotoxic as well as immunoregulatory activity.

Figure 1. Predicted HLA-E binding Mtb peptides induce T-cell proliferation.

CFSE-labeled PBMCs of 20 donors were stimulated with 10 µg/ml of peptide in the presence of IL-7, on day 7 low dose IL-2 was added and at day 10 proliferation was assessed by flowcytometry. Plots are gated on CD3⁺CD8⁺CD56⁻ cells in the lymphocyte gate, the percentage of proliferation indicated is within CD3⁺CD8⁺CD56⁻ cells. Peptide and donor numbers correspond to the numbers in Table 1.

Figure 2. PPD responders and BCG vaccinated infants recognize predicted HLA-E binding peptides derived from Mtb.

(A) Summary of all proliferation data from 10 PPD responders and 10 PPD non-responding donors. Proliferation was calculated using the following formula: Geometric mean of undivided population minus geometric mean of all cells. Subsequently the percentage was calculated by: (delta geo mean of sample- delta geo mean of negative control)/ delta geo mean of maximal proliferation (PHA). Samples with 10% specific proliferation or more were considered positive (black boxes). Blue highlights identify the peptides that were selected for detailed analysis. Umbilical cord blood (UCB) (n = 5) cell proliferation was assessed by the percentage of proliferating CD3⁺CD8⁺CD56⁻ cells and proliferation >10% was considered positive. (B) PBMCs from BCG vaccinated infants (n = 12) were tested in the same way as the cells from PPD responders for 10 selected peptides. Proliferation of 5–10% is indicated in grey and proliferation >10% in black. (C) Comparison of PPD responders, non-responders & cord blood samples for peptide recognition, i.e. the number of peptides recognized per donor (n = 10 for adults, n = 5 for cord blood). Analyzed using Students T-test with a p<0.05 considered significant. (D) CD8⁺ T cells were purified from PBMCs by magnetic bead separation of donors 2,4 and 6 and proliferation (CFSE) in response to peptide pre-pulsed K562 cells with (right plots) or without (left plots) HLA-E was assessed. Plots are gated on CD8⁺CD56⁻ cells and CFSE dilution is plotted.

Figure 3. Surface-, functional- and Treg marker- phenotypes of predicted HLA-E binding peptide specific T-cell lines.

(A) CD8⁺ T-cell lines were generated by peptide stimulation, and activated with αCD3/28 for 24 hours before staining. Brefeldin A was added for the last 16 hours only (3 µg/ml). Cells are gated on a lymphocyte gate combined with CD3 and CD8 gates. (B) Summary of phenotyping data from 3 donors and 4 peptides selected for further detailed analysis.

Figure 4. Predicted HLA-E binding Mtb peptide specific T-cell lines are cytotoxic.

(A) HLA-E expression on K562 cell lines after peptide loading. Upper graph, untransfected K562 cells do not stain with anti-HLA-E (3D12) antibodies, whereas HLA-E transfected K562 cells expressing the natural HLA-E ligand (HLA-B7) or cells loaded with a Mtb derived HLA-E binding peptide (#55, see Table 1) express similar levels of HLA-E. Middle graph, similar staining is observed when an antibody against HLA class I (W6/32) is used, with equal expression levels of HLA-E in the presence of a natural ligand or an Mtb derived peptide. Bottom graph, the isotype control staining for 3D12 and W6/32 was negative on all cell lines tested. HLA-E expression was always confirmed before functional experiments. (B) Peptide specific T-cell lines were generated by stimulation with peptide in the presence of IL-7 and further expansion using IL-2. K562 cells expressing HLA-E were loaded with specific and control peptides, labeled with 1 µCi ⁵¹Cr and co-cultured with T-cell lines for 5 hours in different ratios before ⁵¹Cr release was measured. Data are expressed as % specific lysis. Left graph: T-cell line derived from donor 4 directed against peptide 55 (using the high affinity HLA-E binding peptide 68 or the natural B7 ligand as irrelevant control), right graph: T-cell line derived from donor 2 directed against peptide 62 (using peptide 68 on transfected K562, or peptide 62 on untransfected K562 as irrelevant controls). Representative of 4 lines derived from 3 donors. (C) Fully HLA-A,B,C mismatched adherent monocytes were infected with live BCG (MOI 5, overnight) before labeling with 1 µCi ⁵¹Cr. Cells were co-cultured with T-cell lines for 5 hours in 6-replicate cultures. Grey bars represent BCG infected monocytes whereas white bars represent uninfected monocytes. All HLA-E peptide restricted T-cell lines recognized BCG infected target cells. In contrast a CD8⁺ T-cell clone restricted to the male HY antigen presented in HLA-A2 did not lyse infected monocytes, whereas the natural ligand (B-cells expressing HY in the context of HLA-A2) was specifically lysed, resulting in up to 100% target cell lysis. Data are depicted as mean+standard error of the mean and represent multiple experiments.

Figure 5. HLA-E binding peptide specific T-cell lines have immuno-regulatory activity.

(A) Peptide specific T-cell lines were generated by stimulation with peptide in the presence of rIL-7 and further expanded with rIL-2. To investigate their potential capacity to inhibit CD4⁺ T-cell activity, they were co-cultured with a well characterized CD4 Th1 clone (Rp15 1-1 (1×10e4 cells per well)) proliferating to its cognate peptide (0.5 µg/ml) presented by HLA-DR3⁺ cells. Dose-dependent addition of peptide stimulated cells (ranging from 0.6–5×10e4 added T-cells) reduced proliferation of the Th1 clone as measured by [³H] TdR incorporation at day 3. T-cell lines generated from 3 donors directed against 3 different peptides are shown (donor 2 against peptide 62, donor 4 against peptide 55 and donor 6 against peptide 68), the T cell line from donor 2 against peptide 54 had a similar suppressive capacity (not shown) CPM = counts per minute. (B) To exclude that suppression was merely the consequence of lysis of the responder T-cells, we analysed cell survival in a CFSE labeling experiment. The responder T-cell clone Rp15 1-1 was labeled with a low dose of CFSE (0.005 µM), whereas a second, isogenic T-cell clone with a different peptide specificity and HLA-DR2 restriction (R2F10), was labeled with a high concentration of CFSE (5 µM). Both responder and irrelevant T-cell clones were HLA-E negative. They were then co-cultured with the HLA-E binding Mtb peptide specific T-cell lines (“Treg”) in the presence of the peptide (0.5 µg/ml) recognized by the responder clone Rp15 1-1 and HLA-DR3⁺ APCs. After 16 hours CFSE intensity was measured by flowcytometry. In the absence of added Tregs, similar numbers of responder and irrelevant T-cell clones were retrieved (ratio of 1). The addition of “Tregs” to peptide activated or control cultures also resulted in similar numbers of both responder and irrelevant T-cells, thus indicating that the responder clone is not lysed by the Tregs. Simultaneously a 3 day co-culture was performed and analyzed by [³H] TdR uptake. This experimental set-up revealed inhibition of proliferation of the responder clone. Addition of different numbers of Tregs did inhibit proliferation in a suppression assay but not the ratio of responder over irrelevant T-cell numbers in a CFSE intensity assay. (C) Clonal populations were obtained by limiting dilution of the T-cell line of donor 2 against peptide 62, all derived from 0.1 cells/well cultures. Clones were co-cultured in different ratios to Rp15 1-1 as described in (a) and 3H TdR incorporation measured. Three out of the 5 tested clones inhibited proliferation of the indicator clone in a dose dependent manner. (D) K562 target cells selectively expressing HLA-E were loaded with peptide 62 (10 µg/ml) before ⁵¹Cr labeling, followed by 5 hour co-incubation with T-cell clones and determination of ⁵¹Cr release. Clone 4G10 strongly lysed peptide loaded target cells, whereas 3E11 and 3C1 had moderate lysing capacity.

Figure 6. Suppression is mediated by membrane-bound TGFβ.

(A) Co-culture assay of Treg line (donor 4, peptide 55) with responder T-cell clone (Rp15 1-1) in the presence of 0.5 µg/ml cognate peptide and HLA-DR3⁺ APCs results in inhibition of proliferation ([³H] TdR uptake; left graph). Right graph: Addition of LAP in this setting in the presence of a fixed number (2.5×10e4) of Tregs resulted in a dose-dependent abrogation of suppression. Experiment representative of 4 lines in 3 different experiments. *: p<0.05, **:p<0.01, ***: p<0.001 (Students T-test). (B) Surface staining of membrane bound TGFβ. T-cell lines were stained for the presence of membrane bound TGFβ using the TB21 monoclonal antibody, gated on CD25⁺ cells. mTGFβ staining was observed on all 4 lines, although to a variable extent, the isotype control was used to place markers and was negative for all lines (bottom).