CD1b-restricted GEM T cell responses are modulated by Mycobacterium tuberculosis mycolic acid meromycolate chains

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains a major human pandemic. Germline-encoded mycolyl lipid-reactive (GEM) T cells are donor-unrestricted and recognize CD1b-presented mycobacterial mycolates. However, the molecular requirements governing mycolate antigenicity for the GEM T cell receptor (TCR) remain poorly understood. Here, we demonstrate CD1b expression in TB granulomas and reveal a central role for meromycolate chains in influencing GEM-TCR activity. Meromycolate fine structure influences T cell responses in TB-exposed individuals, and meromycolate alterations modulate functional responses by GEM-TCRs. Computational simulations suggest that meromycolate chain dynamics regulate mycolate head group movement, thereby modulating GEM-TCR activity. Our findings have significant implications for the design of future vaccines that target GEM T cells.

Keywords: CD1b; GEM T cells; Mycobacterium tuberculosis; molecular dynamics; mycolate lipids.

Fig. 1.

CD1b expression within human TB granulomas. Human lung biopsies from patients with confirmed TB were stained for the macrophage marker CD68 (A) and CD1b (B). (C) Negative control with secondary antibody (Ab) and avidin biotin–peroxidase complex (ABC) detection only. (A–C, Insets) Large box is a 2.1× magnified version of the small box. (Scale bars: A–C, 50 μm.)

Fig. 2.

Cross-reactivity of GEM18-TCR. (A) Representative flow cytometry plots showing GEM18-TCR activation through up-regulation of CD69 (y axis) on Jurkat T cells. GEM18-TCR–transduced Jurkat T cells, but not mock-transduced Jurkat T cells, up-regulate CD69 when cultured with CD1b⁺ T2 lymphoblasts in the presence of the MA JR1080 (Table S1). Phorbol ester PMA and ionomycin (PMA/Iono) was used as a positive control. (B) Structure of free MA JR1080 and its Gro-MM OTA-23, GMM SMP74, and TMM MH176 studied in C–E. Activation of GEM1 (C), LDN5 (D), and GEM18 (E) Jurkat T cell lines cultured with CD1b⁺ T2 lymphoblasts in the presence of MA (JR1080), Gro-MM (OTA-23), GMM (SMP74), TMM (MH176), or no antigen. Data are representative of at least three independent experiments (C–E) Mean and SEM of duplicate measurements are shown. Ag, antigen.

Fig. 3.

Meromycolate chain structure determines GEM18-TCR activity. (A) Representative flow cytometry plots of CD69 expression on GEM18 Jurkat T cell lines cultured with CD1b⁺ T2 lymphoblasts in the presence of no antigen (No Ag), weakly stimulatory MA (JRRR121), and strongly stimulatory MA (JR1080). (B) Activation of GEM18 Jurkat T cells cultured with CD1b⁺ T2 lymphoblasts in the presence of various MAs at 10 μg/mL that contain different meromycolate chain structures. (C) Activation of GEM18 Jurkat T cells when cultured with CD1b⁺ T2 lymphoblasts in the presence of MAs that represent the cyclopropane stereoisomers of the Mtb α-MA (JR1080). Stimulations were performed in the form of a dose–response, with TMM (MH176) as the negative control. (D) Activation of GEM18 Jurkat T cells cultured with CD1b⁺ T2 lymphoblasts in the presence of Gro-MM that contains different meromycolate chain structures. Stimulations were performed in the form of a dose–response, with TMM (MH176) as the negative control. Structures of lipids are shown next to the graphs. All graphs are representative of at least three independent experiments performed in duplicate. Mycolate concentrations were 0.1 μg/mL (clear), 1 μg/mL (light gray), 10 μg/mL (dark gray), and 20 μg/mL (black). (B–D) Mean and SEM of duplicate measurements are shown.

Fig. 4.

MA meromycolate variants induce diverse functional responses. (A) Ex vivo T cells from human patients with TB were stimulated with autologous moDC in the presence of one strongly stimulatory MA (JR1080) and GMM (SMP74) or weakly stimulatory MAs (JRRR121 and MMS130). Intracellular IL-2, IFN-γ, and TNF-α were measured by flow cytometry. Cells were pregated on CD3⁺ CD161⁻ live lymphocytes. Cytokine-positive cells are plotted relative to a negative control. (B) Cell viability of GEM18-transduced ex vivo T cells cultured with CD1b⁺ T2 lymphoblasts in the presence of indicated MAs. Targeted cell killing was assessed using a Cytotox-Glo assay. (C) Absolute values of IFN-γ cytokine secretion from GEM18-transduced ex vivo T cells cultured with CD1b⁺ T2 lymphoblasts in the presence of indicated MAs. Cytokine secretion was measured by a Luminex array. (D) Heat map summarizing Luminex array data showing relative concentrations of cytokines in response to lipid antigen. Values were normalized to the mean cytokine concentrations measured in supernatants following stimulation with weakly stimulatory MAs. Red indicates high concentrations, and blue indicates low concentrations. Data are representative of three independent experiments performed in triplicate. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 (A, Mann–Whitney U test; B and C, one-way ANOVA). (B and C) Mean and SD of duplicate measurements are shown.

Fig. 5.

Differential binding of GEM18-TCR to soluble CD1b monomers treated with MAs. (A) SDS/PAGE analysis of recombinant GEM18-TCR under reducing (R) and nonreducing (NR) conditions. The predicted molecular weights (MW; molecular masses) of the TCR-α and TCR-β monomeric proteins and of TCR-αβ heterodimeric proteins are indicated. (B) GEM18-TCR dextramer binding to MACSibeads conjugated to CD1b treated with the methoxy MAs JRRR121, JRRR124, and HA56. Untreated CD1b MACSibeads were used as a control. (C) Specific staining of Jurkat T cells expressing GEM18-TCR with CD1b dextramers treated with weakly stimulatory (JRRR124) and strongly stimulatory (HA56) methoxy MAs. HA56-loaded CD1b-dextramer binding of Jurkat T cells expressing CD1d-restricted [invariant natural killer T (iNKT)] and CD1c-restricted (NM4) TCRs is shown as background controls. MFI, mean fluorescence intensity.

Fig. 6.

Meromycolate chain immobilization affects ligand head group dynamics. (A) Value of rmsd for MA head group movement relative to the head group of GMM in the previously determined CD1b–GMM complex (PDB ID code 1UQS). Higher rmsd values indicate conformations less similar to that observed in 1UQS, while a greater spread of values indicates increased mobility of the ligand head group. Vertical bars mark mean values for histograms of the corresponding color. Highly stimulatory antigens have a lower mean rmsd, while the less stimulatory antigens have a higher mean rmsd. (B and C) Geometric functional group positions are indicated by centroids (colored balls). Data were generated from 200-ns molecular dynamic simulations. Representative weakly stimulatory ligands (B; JRRR121 and JR1046) and strongly stimulatory ligands (C; JR1080 and MH157) are shown. Positions of proximal (blue) and distal (red) functional groups are shown.