Human CD8+ T-cells recognizing peptides from Mycobacterium tuberculosis (Mtb) presented by HLA-E have an unorthodox Th2-like, multifunctional, Mtb inhibitory phenotype and represent a novel human T-cell subset

Abstract

Mycobacterial antigens are not exclusively presented to T-cells by classical HLA-class Ia and HLA-class II molecules, but also through alternative antigen presentation molecules such as CD1a/b/c, MR1 and HLA-E. We recently described mycobacterial peptides that are presented in HLA-E and recognized by CD8+ T-cells. Using T-cell cloning, phenotyping, microbiological, functional and RNA-expression analyses, we report here that these T-cells can exert cytolytic or suppressive functions, inhibit mycobacterial growth, yet express GATA3, produce Th2 cytokines (IL-4,-5,-10,-13) and activate B-cells via IL-4. In TB patients, Mtb specific cells were detectable by peptide-HLA-E tetramers, and IL-4 and IL-13 were produced following peptide stimulation. These results identify a novel human T-cell subset with an unorthodox, multifunctional Th2 like phenotype and cytolytic or regulatory capacities, which is involved in the human immune response to mycobacteria and demonstrable in active TB patients' blood. The results challenge the current dogma that only Th1 cells are able to inhibit Mtb growth and clearly show that Th2 like cells can strongly inhibit outgrowth of Mtb from human macrophages. These insights significantly expand our understanding of the immune response in infectious disease.

Fig 1. Peptide specific HLA-E restricted CD8⁺ T-cell clones are activated through TCR ligation by specific peptide/HLA-E.

A. T-cell clones were stimulated for 16 hours with K562 cells either expressing or lacking HLA-E, which had been pre-loaded with the specific or control peptide. Induction of CD137 expression was determined by flow cytometry. T-cell clones upregulate CD137 only in response to specific but not control Mtb peptide presented in the context of HLA-E. Left panel, T-cell clone specific for peptide 62; Right panel, T-cell clone specific for peptide 68. B. Summary of CD137 upregulation for all T-cell clones tested: T-cell clones from donor 2 recognize peptide 62, whereas T-cell clones from donor 6 recognized peptide 68. Data are expressed as delta MFI, which was calculated by substraction of the MFI obtained with non-peptide pulsed K562/HLA-E from the specific peptide pulsed K562/HLA-E stimulated samples. * indicates no CD137 expression by these T-cell clones despite repetitive measurements. C. T-cell clones were stimulated for 5 minutes with peptide pulsed HLA-E expressing Meljuso cells and stained for 2 different phosphorylation sites of ZAP70 (Y292 in black, Y319 in grey) as indicator of TCR activation. Left panel, T-cell clone specific for peptide 62; Right panel, T-cell clone specific for peptide 68. D. Summary of ZAP70 phosphorylation for all T-cell clones tested for Y292 (left) and Y319 (right). Data are expressed as delta MFI, which was calculated by substraction of the MFI obtained with control-peptide pulsed Meljuso from the specific peptide pulsed Meljuso samples.

Fig 2. HLA-E restricted T-cell clones possess either suppressive or cytolytic activity.

A. T-cell clones were expanded, after which they were added in different ratios to an unrelated reporter Th1 cell clone (Rp15 1-1; Mtb hsp65 p3–13 specific, HLA-DR3 restricted) in the presence of irradiated HLA-DR3 expressing PBMCs as antigen presenting cells together with the cognate peptide recognized by Rp15-1-1 (closed bars). After 3 days of co-culture the proliferative response of the Th1 clone was determined by 3H-TdR incorporation. There was no proliferation in the absence of the cognate p3–13 peptide stimulating the Th1 clone (open bars). Data are expressed in counts per minute (CPM), averaged for triplicate wells (+/- standard deviation). B. T-cell clones were titrated onto ⁵¹Cr labelled adherent HLA-A2 negative monocytes that were infected with live BCG, and the release of ⁵¹Cr was determined after 5 hours. Data are expressed as percentage specific lysis. Black bars represent BCG infected monocytes, open bars represent uninfected control monocytes. A ratio of 10:1 (T-cells: monocytes) is shown here. T-cell clones from donor 2 (peptide 62 specific) and from donor 6 (peptide 68 specific) specifically lysed BCG infected target cells. Data represent the average +/- standard deviation of triplicate wells. C. Combined analyses of suppressive and cytolytic activity for all clones tested. The percentage of suppression was calculated by dividing deltaCPM (CPM in presence of p3–13 to activate Rp15 1-1 proliferation minus CPM in absence of p3–13) of 5x10e4 T-cell clones by the deltaCPM of the Th1 clone in the absence of HLA-E restricted T-cell clones (left panel) as described in [58]. Similarly, the percentage specific lysis was plotted in the middle panel. The percentage of specific Mtb killing for each individual clone is plotted in the right panel. The percentage of Mtb killing was calculated after subtraction of the average experimental variation within each experiment and was tested in 3–4 different macrophage donors for each T-cell clone. Nt = not tested. D. T-cell clones were added to Mtb (H37Rv) infected HLA-A2 negative macrophages for 24 hours in a ratio of 5:1 (T-cells: monocytes), subsequently macrophages were lysed and plated for assessment of colony forming units (CFU). CFU were counted and are expressed as CFU/ml lysate. Data represent the average +/- standard deviation of duplicate wells. E. The percentage of intracellular Mtb growth inhibition was calculated by dividing CFU outgrowth from infected macrophages with and without the addition of T-cells for each individual clone. All clones were tested in duplicate in at least 3 independent experiments, using independent macrophage donors. The percentage Mtb growth inhibition was expressed as average of these experiments. The percentage of Mtb growth inhibition was plotted against the percentage of CD8⁺ T-cells expressing perforin (left), perforin and granulysin (middle) and perforin and granzyme B (right), as assessed by flow cytometry. Linear regression analysis was performed to obtain an R² value.

Fig 3. Peptide specific HLA-E restricted CD8⁺ T-cell clones have an effector memory phenotype.

T cell clones were cultured in the absence of peptide specific stimulation and RNA was isolated from T-cell clones, RNA expression profiles were determined using dcRT-MLPA and data were normalized for GAPDH expression within each sample (A-C). A. RNA expression levels of classical cellular subset and memory markers of T-cell clones; B. RNA expression levels of cytotoxic effector function associated molecules; C. RNA expression levels of lineage associated transcription factors. D. Flow cytometric analysis of T-cell phenotype, T cell clones were directly stained from culture, a representative T-cell clone is shown (D6-2B4). Gating strategy in S1 Fig. E. Flow cytometric analysis of effector molecules, T-cell clones activated with αCD3/28 beads for 24 hours followed by intracellular staining, a representative T-cell clone is shown (D6-2B4). F. Flow cytometric analysis of lineage determining transcription factors, T-cell clones were directly stained from culture using intracellular staining protocols. Dashed lines represent transcription factor staining in PBMCs, grey (D2–1B9) and black are examples of different T-cell clones (D2-4A1, D6-1F11, D2-2A9).

Fig 4. Mtb specific HLA-E restricted T-cell clones produce Th2 cytokines.

A. T-cell clones were cultured and RNA was isolated for gene-expression measurement using dcRT-MLPA, data are normalized to GAPDH as housekeeping gene (left panel); clones were stimulated for 24 hours with αCD3/28 beads before supernatants were collected to determine their maximum cytokine secretion profiles (in pg/ml) using multiplex bead arrays (middle panel); similarly, cells were stimulated for 16 hours with αCD3/28 beads in the presence of brefeldin A followed by intracellular cytokine staining to determine intracellular cytokine levels (right panel). Data are expressed as % of the CD3⁺CD8⁺ T-cell population. Data are coloured according to the amount of the molecules detected, according to the legend in the figure. Nt = not tested. B. HLA-E restricted CD8⁺ T-cell clones were stimulated for 24 hours with αCD3/28 T-cell activator beads and stained intracellular for IL-4 or IL-5 or IL-13 as well as IFN-γ. C. T-cell clones were cultured with peptide pulsed macrophages to assess their specific cytokine production in response to peptide presented by professional antigen presenting cells (top panel), or with BCG infected macrophages to assess their specific cytokine production induced by naturally presented antigen during in vitro mycobacterial infection (bottom panel). Supernatants were collected and cytokine/ chemokine levels were determined using multiplex bead arrays.

Fig 5. HLA-E restricted Mtb specific T-cell clones utilize IL-4 to provide B-cell help.

T cell clones were co-cultured with primary CD19⁺ B-cells in a 1:1 ratio for 48 hours, subsequently B-cell activation was determined by flow cytometry and by measurement of IL-6 in supernatants. A. Flow cytometric analysis of B-cells only (top row), or B-cells co-cultured with 2 independent T-cell clones and stained for CD80, CD86, and CD25. Cells are gated on CD3⁻CD19⁺ cells. B. B-cell activation induced by HLA-E restricted Mtb specific T-cell clones as indicated by expression of CD80, CD86, and CD25. C. B-cell activation induced by panel of unrelated, (CD4⁺) control T-cell clones and by recombinant cytokines. B-cell activation is assessed by flow cytometry. D. IL-6 production in supernatants of co-cultures of B-cells with HLA-E restricted Mtb specific CD8⁺ T-cells, B-cell activators (CpG, αIgG/M), recombinant cytokines and unrelated control T-cell clones. Data are expressed as pg/ml in supernatant. E. Co-culture of B-cells with HLA-E restricted Mtb specific T-cell clones in the presence of blocking antibodies against IL-4, IL-5 or IL-13, supernatants were collected and IL-6 measured by ELISA. Data are expressed as percentage inhibition of IL-6 production in supernatants of specific antibody blocking compared to the isotype control.

Fig 6. CD8⁺ T-cells from TB patients bind HLA-E/ peptide tetramers and produce Th2 cytokines following peptide stimulation.

PBMCs from patients with pulmonary TB were stained directly ex vivo with HLA-E/ peptide tetramers and analysed by flow cytometry. Data are expressed as the percentage tetramer positive cells within the CD8⁺ population. PBMCs were stimulation with either peptide 62 or peptide 68 for 16 hours in the presence of monensin. Cytokines were stained by intracellular staining followed by flow cytometric analysis, data are expressed as percentage of CD8⁺ T-cells. A. Example flow cytometry results for a representative single TB patient following staining with HLA-E tetramers containing peptide 62 or peptide 68, cells are gated on CD8⁺ T-cells. B. Results of combined TM staining on PBMCs from TB patients for both tetramers containing P62 or P68, data are expressed as percentage of CD8⁺ T-cells. C. Example of intracellular cytokine staining following peptide stimulation for a single representative TB patient, cells are gated on CD8⁺ T-cells. D. Cytokine production by CD8⁺ T-cells following stimulation with peptide 62 (left) and peptide 68 (right). Open circles represent patients with tetramer staining <0.1%, close circles represent patients with tetramer staining >0.1%. Groups were compared using a Mann-Whitney U test and p<0.05 was considered significant.