Unconventional T cells in anti-cancer immunity

Abstract

Unlike conventional T cells that detect peptide antigens loaded to major histocompatibility complex (MHC) molecules, unconventional T cells respond to non-peptidic metabolite antigens presented by MHC class I-like proteins, such as CD1 and MHC-related protein 1 (MR1). Semi-invariant mucosal-associated invariant T (MAIT) cells, γδ T cells, and invariant natural killer T (iNKT) cells, together with other CD1- or MR1-restricted T cell subsets expressing diverse T cell receptors (TCR), elicit an innate-like response independent of diverse MHC genetics. In contrast to an overall enhanced response to bacterial-derived riboflavin precursor metabolites in infections, MAIT cells often exhibit an immunosuppressive or exhausted phenotype in glioblastoma, lung cancer, colorectal cancer, and various hematological malignancies. Whereas some tumor cells can activate MAIT cells, the structures and functions of tumor-derived MR1 ligands remain largely unknown. Novel discoveries of mammalian-derived agonists and antagonists binding to MR1 protein are our knowledge of MR1 ligand structures and functions from MAIT cell activation in healthy conditions to anti-cancer immunity. Recent findings reveal that nucleoside and nucleobase analogs, as self-metabolites to activate MR1-restricted T cells, are regulated in the tumor microenvironment. Likewise, iNKT cells exhibit a dynamic role in cancer, capable of both protumor and antitumor immunity. Similarly, γδ T cells have also demonstrated both protective and tumor-promoting roles, via recognizing stress-induced protein and metabolite ligands. This review further depicts the distinct kinetics of responses, highlighting a rapid activation of unconventional T cells in solid versus hematological cancers. Emerging therapeutic strategies, including antigen-loaded MR1 and CD1, adoptive T cell transfer, chimeric antigen receptor-T (CAR-T) cells, T cell receptor-T (TCR-T) cells, and combination treatments with immune checkpoint inhibitors, yet remain challenging, hold promise in overcoming tumor-induced immunosuppression and genetic restriction of conventional T cell therapies. By addressing critical gaps, such as novel structures and functions of cancer metabolite antigens, unconventional T cells offer unique advantages in anti-cancer immunotherapy.

Keywords: CD1; MHC class I-related protein 1 (MR1); cancer; immunotherapy; lipids; polar metabolites; unconventional T cells.

Figure 1

Re-analyses of tertiary crystal structures of MR1 or CD1, mammalian metabolite antigens, and invariant T cell receptors, in parallel with an HLA-A2-auto peptide-TCR complex. The reported crystal structures for human MR1, 5-formyl-deoxyuridine, and A-F7 MAIT cell TCR (9EK7 from the PDB database), human CD1d, β-galactosylceramide, and autoreactive Vα24 TCR (3SDX), human CD1d, sulfatide, and DP10.7 γδ TCR (4MNG), and HLA-A2, glioblastoma peptide HuD, and A6 conventional CD8+ T cell TCR (3PWP) were compared via Pymol for metabolite antigen interaction with MR1, CD1, and TCR chains. (A) Metabolite antigen binding to MR1, CD1, and TCR chains is shown through polar interactions (yellow dots), particularly hydrogen bonding, but hydrophobic interaction is not shown. Interacting residues were annotated. α1, α2, and α3 label α1, α2, and α3 domains with α2 domains partially removed to show protein-ligand interactions. Kd, the dissociation constant, represents the concentration of ligand binding to 50% of receptor molecules. (B) Metabolite antigen binding to MHC class I-like proteins differentially shapes the orientation of TCRα and TCRβ chains.

Figure 2

Polar metabolite antigen presentation to MR1-restricted T cells. (A) Vitamin B1, B6, and B9 precursors as MR1 ligands. (B) tumor-associated mammalian-derived metabolites as MR1 ligands. (C) synthetic analogues as MR1 ligands. (D) reported antitumor responses upon stimulation with polar metabolites. Black lettering denotes agonists and red lettering depicts antagonists. Red X indicates inhibited MR1T cell response. Abbreviations are as follows: 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU), 5-(2-oxoethylideneamino)-6-D-ribitylaminouracil (5-OE-RU), ribityl lumazine (RL), 7-methyl-ribityllumazine (RL-7-Me), 6,7-dimethyl-8-ribityllumazine (RL-6,7-diMe), 6-methyl-7-hydroxy-ribityllumazine (RL-6-Me-7-OH), 3-(2-deoxy-β-D-erythro-pentofuranosyl)-6-(hydroxymethyl)-8-oxo-9H-purine-2-carbaldehyde (M1Ado), 6-(hydroxymethyl)-8-oxo-9H-purine-2-carbaldehyde (M3Ade), and acetyl-6-formylpterin (Ac-6-FP).

Figure 3

Polar metabolites as MR1 ligands. (A) Vitamin B1, B6, and B9 precursors or related metabolites shown in relation to their metabolic pathways (GTP for riboflavin biosynthesis), with reactants labeled above arrows indicating product formation. RibH (lumazine synthase) and RibC (riboflavin synthase) enzymatically convert early vitamin B2 precursors into riboflavin derivatives. (B) Nucleoside and nucleoside adducts but with unknown MR1-binding capacity for guanosine and 2-deoxy-5-formyluridine. (C) Synthetic analogues associated with vitamin precursor metabolites. Ac-6-FP is derived from vitamin B9 precursors, while 2-oxopropyl pteridine, JYM72 (chemical name unknown), hydroxyethylamine, and propylamine are linked to vitamin B2 derivatives. MR1 agonists or MAIT stimulators are underlined. Abbreviations are as follows: guanosine triphosphate (GTP), 5-amino-6-(D-ribitylamino)uracil (5-A-RU), 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU), 5-(2-oxoethylideneamino)-6-D-ribitylaminouracil (5-OE-RU), ribityl lumazine (RL), 6,7-dimethyl-8-ribityllumazine (RL-6,7-diMe), 6-methyl-7-hydroxy-ribityllumazine (RL-6-Me-7-OH), pyridoxal (PL), 2-deoxy-5-formyluridine (fdU), 3-(2-deoxy-β-D-erythro-pentofuranosyl)-6-(hydroxymethyl)-8-oxo-9H-purine-2-carbaldehyde (M1Ado), 6-(hydroxymethyl)-8-oxo-9H-purine-2-carbaldehyde (M3Ade), and acetyl-6-formylpterin (Ac-6-FP).

Figure 4

MR1T cell responses in Hematological Malignancies (A) and Solid Tumors (B). Each panel is divided into quadrants to illustrate the relationship between MR1T cell phenotype (top, labeled “Expression”) and functional outcome (bottom, labeled “Response”). Top left quadrant: pro-tumor immunosuppressive or exhausted phenotypes. Bottom left quadrant: tumor-promoting cytokine profiles and dysfunctional responses. Top right quadrant: antitumor effector and activation phenotypes. Bottom right quadrant: tumor-controlling protective responses. Cancer cells depicted include (A) malignant B cell-chronic lymphocytic leukemia (B-CLL), multiple myeloma (MM); (B) Breast cancer (BC), colorectal cancer (CRC), glioblastoma (GBM).

Figure 5

CD1d-bound lipids detected in different cancers. Agonists are underlined. Glycophospholipid is identified from HLA class I deficient human lymphoblastoid, MS detection with m/z 861.8 of [M-H]-, but with an unknown hexosyl group.

Figure 6

Cancer-associated phosphoantigens (pAgs). HMBPP, a microbial metabolite, and IPP, a host-derived mevalonate intermediate, are presented by butyrophilin proteins to activate Vγ9Vδ2 γδ T cells.