Distinct sets of alphabeta TCRs confer similar recognition of tumor antigen NY-ESO-1157-165 by interacting with its central Met/Trp residues

Abstract

Naturally acquired immune responses against human cancers often include CD8(+) T cells specific for the cancer testis antigen NY-ESO-1. Here, we studied T cell receptor (TCR) primary structure and function of 605 HLA-A*0201/NY-ESO-1(157-165)-specific CD8 T cell clones derived from five melanoma patients. We show that an important proportion of tumor-reactive T cells preferentially use TCR AV3S1/BV8S2 chains, with remarkably conserved CDR3 amino acid motifs and lengths in both chains. All remaining T cell clones belong to two additional sets expressing BV1 or BV13 TCRs, associated with alpha-chains with highly diverse VJ usage, CDR3 amino acid sequence, and length. Yet, all T cell clonotypes recognize tumor antigen with similar functional avidity. Two residues, Met-160 and Trp-161, located in the middle region of the NY-ESO-1(157-165) peptide, are critical for recognition by most of the T cell clonotypes. Collectively, our data show that a large number of alphabeta TCRs, belonging to three distinct sets (AVx/BV1, AV3/BV8, AVx/BV13) bind pMHC with equal antigen sensitivity and recognize the same peptide motif. Finally, this in-depth study of recognition of a self-antigen suggests that in part similar biophysical mechanisms shape TCR repertoires toward foreign and self-antigens.

Fig. 1.

Sequence analysis of TCR junctional transcripts derived from BV1⁺, BV8⁺, and BV13⁺ NY-ESO-1_157–165 multimer⁺ T cell clones from five melanoma patients. Transcripts were reverse-transcribed and amplified from multimer⁺ T cell clones by using either pairs of BV1/BC, BV8/BC, or BV13/BC primers. PCR products were directly purified and sequenced. Conserved residues forming the CDR3 loop and preferential BJ gene element usages are highlighted in gray. The number of in vitro generated T cell clones displaying a given sequence is depicted. Of note, the NY-ESO-1-specific CD8⁺ T cell response seen in patients LAU 156 and LAU 198 was composed of, respectively, a single dominant BV8 and BV13 T cell clonotype, confirming the results obtained by using anti-BV antibodies and flow cytometry (Fig. S1).

Fig. 2.

Junctional amino acid sequence of TCR α-chains associated with each TCR β transcript derived from dominant NY-ESO-1_157–165-specific T cell clonotypes. Conserved residues forming the CDR3α/β loops and preferential AJ/BJ gene element usages are highlighted in gray. *, Clonotypes recently described in ref. .

Fig. 3.

Fine specificity of NY-ESO-1 antigen recognition of tumor-reactive T cell clones derived from patient LAU 50. The relative TCR functional avidity was compared by using T2 target cells (HLA-A2⁺/TAP⁻/⁻) pulsed with graded concentrations of either the native NY-ESO-1_157–165 peptide (SLLMWITQC) or the analog NY-ESO-1_157–165 peptide (SLLMWITQA). Complete collection of data (n = 82 clones) representing the peptide concentration (either native or analog) used to achieve 50% of maximum lysis is shown. Each data point represents the result of an individual clone. Mean ± SD values (nanomolar) are shown. *, P ≤ 0.03 with the Welch two-sample t test.

Fig. 4.

Contribution of individual NY-ESO-1_157–165 amino acid residues to functional antigen recognition by specific T cell clonotypes. (A) The relative antigenicity of alanine-substituted NY-ESO-1 peptide analogs was assessed on T2 target cells in the presence of graded peptide concentration. We used five and six distinct T cell clonotypes (BV1, BV8, and BV13) isolated from patients LAU 155 and LAU 50, respectively. The molar concentration required for 50% maximal lysis by each T cell clone and for each peptide was calculated from peptide titration experiments. The relative antigenic activity of each peptide analog was determined by using the native peptide NY-ESO-1 as a reference (SI Text). The recognition patterns of BV1-, BV8-, and BV13-derived clonotypes are depicted in separate panels of graphs. (B) Estimation of the binding free energy contributions of each residue of the NY-ESO-1_157–165 peptide for the TCR BV13 clonotype 1 isolated from patient LAU 155. Of note, the side chains of the central residues (P4–P8) showed important favorable contributions in the binding between peptide and TCR, whereas the contributions of the outer side chains (P1, P2, P3, and P9) were not significant. (C) Comparison of the sum of the binding free energy contributions made by the TCR Vβ residues and the TCR Vα residues toward the HLA-A*0201/NY-ESO-1_157–165 complex. Histogram count of 19 of the 20 TCR models was generated in silico, as a function of the ΔG_β,bind − ΔG_α,bind difference. The ΔG_β,bind − ΔG_α,bind difference is in kilocalories per mole. Of note, the 2BNR (15) ΔG_β,bind − ΔG_α,bind difference was −6.4 kcal/mol, conferring again a preponderant role to the β-chain in the binding process.