Current T cell-based cancer therapies are personalized to each patient (to ensure HLA matching) and are only suitable against certain cancers (with known target antigens). So the idea of a T cell-based therapy that is not limited by polymorphic MHC restriction or known tumour antigens is very attractive. A study published in Nature Immunology describes a new T cell population that specifically kills many human cancer cell types via recognition of the non-polymorphic MHC class I-related protein MR1. MR1 is widely expressed and varies little between individuals, raising the prospect of a universal cancer immunotherapy.

Credit: Simon Bradbrook/Springer Nature Limited

A T cell population that proliferated in response to A549 cancer cells was isolated from peripheral blood mononuclear cells from an HLA-mismatched healthy donor. Remarkably, the resultant T cell clone — named MC.7.G5 — could kill multiple cancer cell lines regardless of their HLA allomorph expression and did not kill healthy cells. The authors used a genome-wide CRISPR–Cas9 screening approach to identify the genes essential for recognition of target cells by MC.7.G5 T cells. In this screen, only transduced target cells in which the key gene involved in recognition was ablated survived selection with MC.7.G5 T cells. This led to the identification of six genes in the resistant target cells: B2M, MR1, STAT6 and three genes associated with transactivation of β2M and MHC class I and II promoters (RFX, RFXANK and RFXAP), suggesting that MC.7.G5 T cells recognize cancer targets via MR1, which forms heterodimers with β2M. Accordingly, anti-MR1 antibody, but not MHC class I or II antibodies, blocked target cell recognition by MC.7.G5 T cells. In addition, reintroduction of MR1 into CRISPR–Cas9-mediated MR1-knockout A549 cells restored recognition by MC.7.G5 T cells.

MR1 molecules presenting intermediates of riboflavin synthesis are known to be sensed by mucosal-associated invariant T (MAIT) cells. But MC.7.G5 T cells did not recognize MR1 bound with MAIT cell-activating ligands nor empty MR1 molecules or MR1 expressed by healthy monocyte-derived dendritic cells. Moreover, target cell recognition by MC.7.G5 T cells was reduced in the presence of riboflavin-derived microbial MR1 ligands. This suggested that MC.7.G5 T cells recognize MR1 molecules presenting cargo that is specific to cancer cells. Importantly with respect to their potential use as cancer immunotherapy, MC.7.G5 T cells remained unresponsive to healthy cells from various tissues even after MR1 surface expression was upregulated by cell activation or after exposure to cell stressors.

To test their capacity to target cancer in vivo, MC.7.G5 T cells were adoptively transferred into mice engrafted with leukaemia cells. Mice receiving MC.7.G5 T cells showed a marked reduction in cancer burden 18 days later and increased survival compared with mice that did not receive MC.7.G5 T cells. This in vivo targeting was dependent on expression of MR1 by the leukaemia cells, as shown by co-transfer of wild-type and MR1–/– leukaemia cells. Finally, therapeutic potential was highlighted by the finding that T cells purified from patients with stage IV melanoma and transduced with the MC.7.G5 T cell receptor could recognize and kill autologous and non-autologous melanomas but not healthy cells or MR1–/– melanomas.

MC.7.G5 T cells recognize MR1 molecules presenting cargo that is specific to cancer cells

This paper identifies an intriguing new population of cancer-specific T cells that has potential as a pan-cancer therapy. Future work will be needed to establish the exact nature of the ligand recognized by the MC.7.G5 T cells.