MAITs onstage in mice and men with three acts for development

Mucosal-associated invariant T (MAIT) cells are a conserved T-cell subset found in humans, mice and cattle. Although putative ligands and functions of those cells have become the focus of recent studies, their development was still poorly understood. In a recent issue of Nature Immunology, Koay et al. have shed light on the MAIT cell developmental process in mice and men.1 Already in the 1990s, MAIT cells were characterized as a conserved subset of lymphocytes that expresses an invariant TCR α chain. MAIT cells are positively selected by MR1, a non-polymorphic ortholog of the major histocompatibility complex (MHC)-class I protein.2, 3 More recently, MR1 was shown to bind riboflavin (vitamin B) metabolites originating from microbes.4 In vitro, MAIT cells reacted to MR1-expressing cells infected with a variety of bacteria like Escherichia coli and species of Mycobacterium, Staphylococcus and Streptococcus, as well as yeast species like Candida and Saccharomyces.5, 6, 7 A close relationship of MAIT cells with commensal microbes was manifested by the absence of MAIT cells in germ-free mice, and MAIT cells localized to sites where microbial antigens likely are presented, including the intestines, liver and mesenteric lymph nodes. A protective, antimicrobial role of MAIT cells has been demonstrated, as MR1-deficient mice showed an increased bacterial burden after infection with Klebsiella, Mycobacterium and Francisella.6, 8, 9 Furthermore, patients with pulmonary infections had a lower MAIT cell frequency in their blood, suggesting these cells localized to the infected tissue. Despite interspecies conservation of their specificity and other properties, although MAIT cells are highly abundant in diverse human tissues, their frequency tends to be lower in mice.10 Most peripheral MAIT cells lack expression of either TCR co-receptor, CD4 or CD8αβ, and therefore are predominantly CD4, CD8αβ double-negative (DN) or CD8α+ lymphocytes that display a memory phenotype. A stepwise developmental program for MAIT cells had been proposed, with an intra-thymic, MR1-dependent, positive selection, and a B-cell and microbe-dependent peripheral expansion.11 Despite this, the phenotypic progression that occurs during the differentiation of MAIT cells, and the point at which they acquire characteristic effector functions, had not been clarified. As often happens in science, the development of a new tool, in this case MR1 tetramers loaded with antigen, has allowed for substantial progress. Using MR1 tetramers loaded with the ribitylaminouracil compound 5-OP-RU, Pellicci et al. identified new MAIT cell populations and convincingly described a three-stage MAIT cell developmental process (Figure 1). The most immature MR1 tetramer-positive precursors in murine thymus, stage 1, were CD4, CD8αβ double-positive (DP) or CD4 single-positive (CD4SP), and CD24+CD44−. A CD24−CD44− stage 2 subset was characterized that was mainly CD4+CD8α−. In contrast, stage 3 thymus MAIT cells, which were the majority, were CD24−CD44+ and either DN or CD8α+. Thymic stage 3 cells appeared functionally mature, with expression of the transcription factor PLZF, either RORɣt or T-bet, and secretion of interleukin (IL)-17 and IFN-ɣ upon ex vivo stimulation. Stage 3 MAIT cells were by far the major population in peripheral sites, such as spleen, lung and lymph nodes. The progression of stages was supported by an analysis of MAIT cells through ontogeny, with stage 1 predominant at 2 weeks of age and stage 3 the majority weeks later. In addition, in a very elegant in vitro cell development system, the authors were able to track the differentiation of sorted stage 1, stage 2 or stage 3 cells when cocultured for 5 days with OP9 bone marrow stromal cells. Whereas stage 1 MAIT cells could differentiate to stages 2 and 3 in an MR1-dependent manner, stage 2 cells could only develop into stage 3, whereas stage 3 cells maintained their mature phenotype. By analyzing mutant mouse strains, the authors identified several requirements that control MAIT cell differentiation. MR1 was critical, as it was required for transitions from stage 1 to 3, and for differentiation and/or survival of stage 2 and 3 cells. Transition to stage 3 also requires microRNA (miRNA) formation, as most MAIT cells were arrested at stage 1 when analyzed in mice deficient for the miRNA biogenesis enzyme Drosha. Analysis of mutant strains showed that stage 3 differentiation required expression of PLZF, IL-18 and also, commensal microbes, as the frequency of stage 3 cells in germ-free mice was strongly reduced. The authors also examined if there is a comparable differentiation of MAIT cell precursors in humans. Thus, human thymus, cord blood as well as peripheral blood from young and adult donors were compared. Matched blood and thymus tissue from donors aged 5 days to 14 years allowed for a well-controlled analysis. Although the markers for mouse and human MAIT cells were different, the data were consistent with a similar three-stage differentiation process, with stage 1 cells identified as CD161−CD27− (Figure 1), mostly DP and exclusively present in the thymus. Stage 2 (CD161−CD27+, DP/CD4+/CD8SP) cells could be detected in blood, although their frequency declines with age, and stage 3 (CD161+CD27+/lo, DN or CD8+) cells were the main fraction detected in the periphery. Similar to mouse MAIT cells, human stage 3 cells had the most mature phenotype, as they expressed PLZF, T-bet and RORɣt together, and some secreted IFN-ɣ and TNF-α upon ex vivo stimulation. Overall, this study by Pellicci et al. identified a number of important similarities in intra-thymic MAIT cell development in mice and men. There are differences as well, for example, the lower frequency of human thymus MAIT cells capable of producing cytokines after activation compared with their mouse counterparts, especially for IL-17A. Interestingly, despite some differences, there are also a few similarities with the differentiation of type I or invariant natural killer T (iNKT) cells, including the marking of immature cells by CD24 expression, the late acquisition of expression of NK receptors such as CD161, and a requirement for the transcription factor PLZF for functional maturity, although the kinetics of PLZF induction differ in the two cell types. The increased number of MAIT cells in Cd1d−/− mice lacking iNKT cells suggested that these two cell types compete for an intra-thymic niche. Mapping reliable cell markers for different developmental stages will be of great value for further analyses of MAIT cell populations, and related populations that are MR1-restricted, but that do not express the canonical TCR α, or that have different specificities.5 Simultaneously, this study raises new questions about the further maturation of MAIT cells in the periphery. Which factors lead to the switch from CD8αβ to CD8αα in peripheral MAIT cells or to production of different cytokines in tissues? Why is IL-18 required but not IL-18Rα? These issues still have to be elucidated, however, Pellicci et al. took the first step to stage the three acts of MAIT cells in proper light. The author declares no conflict of interest.