A molecular basis underpinning the T cell receptor heterogeneity of mucosal-associated invariant T cells

Abstract

Mucosal-associated invariant T (MAIT) cells express an invariant T cell receptor (TCR) α-chain (TRAV1-2 joined to TRAJ33, TRAJ20, or TRAJ12 in humans), which pairs with an array of TCR β-chains. MAIT TCRs can bind folate- and riboflavin-based metabolites restricted by the major histocompatibility complex (MHC)-related class I-like molecule, MR1. However, the impact of MAIT TCR and MR1-ligand heterogeneity on MAIT cell biology is unclear. We show how a previously uncharacterized MR1 ligand, acetyl-6-formylpterin (Ac-6-FP), markedly stabilized MR1, potently up-regulated MR1 cell surface expression, and inhibited MAIT cell activation. These enhanced properties of Ac-6-FP were attributable to structural alterations in MR1 that subsequently affected MAIT TCR recognition via conformational changes within the complementarity-determining region (CDR) 3β loop. Analysis of seven TRBV6-1(+) MAIT TCRs demonstrated how CDR3β hypervariability impacted on MAIT TCR recognition by altering TCR flexibility and contacts with MR1 and the Ag itself. Ternary structures of TRBV6-1, TRBV6-4, and TRBV20(+) MAIT TCRs in complex with MR1 bound to a potent riboflavin-based antigen (Ag) showed how variations in TRBV gene usage exclusively impacted on MR1 contacts within a consensus MAIT TCR-MR1 footprint. Moreover, differential TRAJ gene usage was readily accommodated within a conserved MAIT TCR-MR1-Ag docking mode. Collectively, MAIT TCR heterogeneity can fine-tune MR1 recognition in an Ag-dependent manner, thereby modulating MAIT cell recognition.

Figure 1.

MR1 tetramer staining and MR1 up-regulation by 6-FP and Ac-6-FP. (a) Chemical structures of MR1-restricted pterin ligands (6-FP [6-FP, i] and acetyl-6-FP [Ac-6-FP, ii]) and pterin-based compounds that did not bind to MR1 (pterin [iii], 6,7-dimethylpterin [6,7-diMePterin, iv], xanthopterin [v], pterin-6-carboxylic acid [vi], 6-hydroxymethylpterin [6-OHMePterin, vii], l-monapterin [viii], tetrahydrobiopterin [ix]). (b) Human PBMCs were stained with CD3- and CD161-specific mAbs and human tetrameric MR1–5-OP-RU, MR1-6-FP, or MR1–Ac-6-FP, respectively. Live, CD3⁺ cells were considered for CD161 (x axis) and tetramer staining (y axis), displaying dot plots and percentages of cells within boxed regions. Displayed are the results of three representative of 6 healthy blood donors tested. These experiments were performed twice using two independent batches of tetramer and yielding similar results. (c) MR1 expression levels on C1R.MR1 cells upon incubation with 6-FP, Ac-6-FP, 6,7-dimethylpterin, 6-hydroxymethylpterin, and 5-OP-RU added in indicated doses and for indicated time spans. Cells were stained with anti-MR1 mAb 26.5. Data shows MFI fold of background of MR1 expression levels, mean ± SEM of three independent experiments done in triplicates. These experiments were performed three times, yielding similar results and a representative experiment is shown. (d) Comparison of MR1 expression levels on C1R and C1R.MR1 cells upon incubation with 10 µM 6-FP or Ac-6-FP, or without ligand (media), for different time periods. Cells were stained with anti-MR1 mAb 26.5. Data shows untransformed geometric MFI (gMFI) of MR1 expression levels for each time point. Mean ± SEM of triplicates. These experiments were performed twice, yielding similar results and a representative experiment is shown.

Figure 2.

MAIT cell inhibitory ligands and MR1 stability. (a) Inhibition of Jurkat.MAIT (original WT, clone A-F7) cell activation by 6-FP or Ac-6-FP. 6-FP, Ac-6-FP (or controls – 6,7-diMePterin or no ligand) were added to C1R.MR1 cells at the indicated concentrations for 1 h before addition of Jurkat.MAIT cells and 0.02 µM synthetic rRL-6-CH₂OH. Data shows MFI of CD69 expression levels for gated Jurkat.MAIT cells. Mean ± SEM of triplicates. These experiments were performed twice, yielding similar results and a representative experiment is shown. (b) Inhibition of Jurkat76.MAIT (original WT TCR, clone A-F7) cell activation by 6-FP or Ac-6-FP. 6-FP, Ac-6-FP (or controls, PBS or no ligand) were added to C1R.MR1 cells at the indicated concentrations and co-incubated with Jurkat76.MAIT cells and 0.02 µM synthetic rRL-6-CH₂OH. Data shows emission at 492 nm correlating with IL-2 production. These experiments were performed three times, showing mean ± SD. (c) Thermostability of soluble MR1-ligand by fluorescence-based thermal shift assay. Shown is baseline-corrected, normalized emission at 610 nm plotted against temperature. Mean ± SEM of triplicate samples, and nonlinear curve-fits. The half maximum melt point (Tm50) is indicated as a dashed line. Displayed is a representative of three independent experiments yielding similar results. The table summarizes the data of three independent experiments, each in triplicate.

Figure 3.

MAIT TCR recognition of MR1–Ac-6-FP. (a) Electron density of Ac-6-FP in omit and final 2Fo-Fc maps of the ternary complex MR1–Ac-6-FP-MAIT original WT (clone A-F7) TCR. Electron density in mesh format, the ligand in ball and stick, MR1 in ribbon representation in white, a Fo-Fc omit map contoured at 3σ (green). (b–d) Structural basis for Ac-6-FP recognition by MR1 and original WT (clone A-F7) MAIT TCR compared with the previously published MR1-6-FP original WT (clone A-F7) MAIT TCR structure (PDB accession no. 4L4T). (b) Superposition of Ac-6-FP and 6-FP and MR1 induced conformational changes. (c) CDR3β conformational changes induced by Ac-6-FP. (d) Contacts between Ac-6-FP and Tyr95α from the CDR3α loop. MR1-6-FP residues are shown in white, pale yellow, and pale orange for the CDR3α and CDR3β loops, respectively, and the 6-FP ligand is shown in cyan. Equivalent residues from the MR1–Ac-6-FP shown in slate, yellow, and orange for the CDR3α and CDR3β loops, respectively and Ac-6-FP in green with contacts between TCR and Ac-6-FP indicated by dashed lines.

Figure 4.

MAIT TCR binding affinity and activation. (a) Relative binding of MAIT TCRs original WT (clone A-F7), B-B10, B-C10, C-A11, C-C10, B-G8, and B-F3-C1 against MR1–Ac-6-FP determined by SPR at 100 µM TCR. (b–k) Equilibrium binding curves for MAIT TCRs. (b) original WT (TRBV6-1, clone A-F7), (c) B-B10, (d) B-C10, (e) C-A11, (f) C-C10, (g) B-G8, (h) B-F3-C1, (i) #6 (TRBV6-4/TRAJ33), (j) C-F7 (TRBV20), and (k) #4 (TRAJ20) against MR1–5-OP-RU determined by SPR. Data are representative of the mean response for each TCR concentration and SEM in duplicate (n = 2). Dose response to synthetic rRL-6-CH₂OH (l) or 5-A-RU (m) preincubated with C1R.MR1 cells by TRAV1-2-TRAJ33-TRBV6-1 MAIT TCR⁺ SKW3.B-F3-C1, SKW3.B-G8, and SKW3.original WT (A-F7) or control TCR⁺ SKW3.LC13 cells. Data shows mean ± SEM fold of background MFI of CD69 expression for gated SKW3.TCR cells from 1 experiment (triplicate samples). These experiments were performed twice, yielding similar results.

Figure 5.

Differential CDR3β usage and MAIT TCR-MR1-Ag recognition. Structural representation of the MR1 surface in white with Ac-6-FP bound, shown in green. MAIT TCR CDR3β loops for the 7 structures are indicated for (a) original WT (TRBV6-1, clone A-F7) in orange, (b) B-B10 in green, (c) C-A11 in purple, (d) B-C10 in cyan, (e) B-G8 in salmon, (f) B-F3-C1 in magenta, and (g) C-C10 in slate. Contacting residues between MR1 and the CDR3β loops are shown in stick representation with the contact surface on MR1 shown according to the MR1 element type: carbon, orange; nitrogen, blue; oxygen, red; and sulfur, yellow. CDR boundaries are listed in Table 1.

Figure 6.

Impact of different Ags on MAIT TCR-MR1-Ag recognition. Structural differences between the ternary complexes of MR1–Ac-6-FP-MAIT TCR (MR1: white, Ac-6-FP: green, MAIT TCR: cyan) and MR1–5-OP-RU-MAIT TCR (MR1: slate, 5-OP-RU: yellow, MAIT TCR: orange). Remodeling of the (a) C-A11 and (b) C-C10 CDR3β loop between Ac-6-FP and 5-OP-RU complexes, with hydrogen bonds indicated in black dashed lines and vdw contacts in red dashed lines. CDR boundaries are listed in Table 1.

Figure 7.

TRBV usage and MAIT TCR-MR1-Ag recognition. Comparison of the MR1–5-OP-RU- MAIT TCR C-F7 (TRBV20), original WT (TRBV6-1, clone A-F7) and #6 (TRBV6-4) complex variable domains and CDRβ loops in magenta, green and cyan, respectively. (a) Overlay of the CDR loops from the three complexes on the MR1–5-OP-RU groove. (b) Comparison of relative angles for the Vβ-domains of original WT (TRBV6-1, clone A-F7) and C-F7 (TRBV20) MAIT TCR structures. Footprints of the MAIT TCRs on the MR-5-OP-RU surface for (c) original WT (TRBV6-1, clone A-F7), (d) #6 (TRBV6-4), and (e) C-F7 (TRBV20). Contacting residues between MR1 and the (f) original WT (TRBV6-1, clone A-F7), (g) #6 (TRBV6-4), and (h) C-F7 (TRBV20) CDR2β loops shown in stick representation with the contact surface on MR1 colored according to the MR1 element type: carbon, orange; nitrogen, blue; oxygen, red. CDR boundaries are listed in Table 1.

Figure 8.

TRAJ usage and MAIT TCR recognition. (a) Dose response to synthetic rRL-6-CH₂OH preincubated with C1R.MR1 cells by WT (a) or mutant Y95F (b) TRAV1-2-TRBV6-4 TRAJ12⁺, TRAJ20⁺, and TRA33⁺ SKW3.MAIT. Data shows mean ± SEM fold of background MFI of CD69 expression for gated SKW3.TCR cells from one experiment (triplicate samples). These experiments were performed twice yielding similar results. Comparison of the MAIT TCR TRBV6-4 (c) TRAJ20 (#4) and (d) TRAJ33 (#6) MR1–5-OP-RU structures with MR1, 5-OP-RU, CDR3α, and CDR3β loops shown in white, yellow, pale yellow, and orange, respectively. (e) Comparison of the positioning of CDR3α loops and Tyr95α for TRAJ20 (#4) and TRAJ33 (#6) in slate and pale yellow, respectively. Contacts between CDR3α, CDR3β, and 5-OP-RU are shown with dashed lines in black for hydrogen bonds and red for vdw contacts. CDR boundaries are listed in Table 1.