I will firstly summarize how Albert Bendelac and I discovered the NKT and MAIT cells in the early 90. I will outline the commonalities and differences between these two subsets regarding their development and peripheral seeding in different species. In particular, the dependency on the SLAM/SAP pathway is slightly different for NKT and MAIT cells in mice and humans. I will discuss how the evolutionary features of the MAIT TCR and MR1 molecule indicates a specific selection pressure on MR1 to educate and present antigen to MAIT cells.
In mice, both NKT and MAIT cells develop in the thymus into highly similar subsets which are either type-1 (T-bet) or type-17 (RORgt+). However, rodents are the exception as in 4 other mammalian species a common mixed 1/17 phenotype is observed. In mice, this blended 1/17 phenotype is only found in the gut or after infection. scRNAseq+TCR analysis of developing MR1:5-OP-RU and CD1d:aGC tetramer+ murine thymocytes indicates that TCR characteristics do not determine the type-1 vs type-17 subset choice. These results suggest a stochastic process in the 1/17 choice as confirmed by a predominant type 1 for both NKT cells and the few MAIT cells that develop in germ free mice. The higher proportion of MAIT17 cells found in SPF mice is related to an intra-thymic proliferation of this subset driven by the microbial 5-OP-RU originating from the periphery.
Contrary to NKT cells that are exclusively selected by CD4+CD8+ thymocytes, some MR1:5-OP-RU tetramer+ cells are selected by thymic epithelial cells and following interaction with MR1-presenting DP thymocytes, mature into MAIT cells. MAIT cells exit the thymus in a S1PR-dependent manner at several stages of maturation. The different MAIT subsets seed specific organs as, for instance, semi-mature MAIT cells go to the intestine in a CCR9-dependent manner where they become fully mature.