Signals that control MAIT cell function in healthy and inflamed human tissues

1.1 |. Discovery of MAIT cells, MR1 restriction, and antigen

MAIT cells are characterized by the usage of a semi-invariant T-cell receptor. Most frequently their TCRα chain is composed of TRAV1–2 (Vα7.2) partnered with TRAJ12, TRAJ20, or TRAJ33.^1–3 The use of this TCRα chain allowed for the initial discovery and identification of MAIT cells.^1,2 The TCRβ chain usage has more variability with bias toward particular TRBV genes.⁴ MAIT cell frequencies vary, and MAIT cells make up between ~0.1% and 10% of the total T-cell compartment in human peripheral blood.⁵ Within the liver, MAIT cells are typically enriched compared to blood where they on average comprise 15% of the total lymphocyte population.^6–8 Importantly, MAIT cells are not enriched in all mucosal tissues and typically found in frequencies (as a % of T cells) comparable to circulating levels in the colon,^6,9 lung,¹⁰ and oral mucosa.^11,12

MAIT cells are restricted to a nonclassical MHC, the MHC class I-related (MR1) protein.^13,14 As indicated by the name, the MR1 gene is located outside the MHC locus (the MR1 gene is on Chromosome 1, and the MHC locus is on Chromosome 6), but is in fact very similar to classical MHC class I, including the use of β2M. As MR1 has limited genetic variation in humans,¹⁵ to the point of being functionally monomorphic, and high sequence homology across mammalian species,¹⁶ it has been suggested that MAIT cells have an important evolutionary conserved role in immune responses.¹⁶ The antigens that have been identified so far as MR1 ligands are metabolites.¹⁷ The most studied of these are pyrimidines and lumazines that are formed in the riboflavin biosynthesis pathway found in species of bacteria and yeast.^18,19 Folic acid derivatives,^18,20 and synthetic modified ligands have also been reported.²¹ Host-derived ligand for MAIT cells was reported after the initial submission of our review: Ito et al. identified cholic acid 7-sulfate (CA7S) as an MR1 ligand that promotes MAIT cell homeostasis rather than activation.²² We will discuss the significance of the T-cell receptor signal in regulating MAIT cell function in Chapter 2. Of note, nearly all nucleated cells express MR1 transcript, but surface MR1 protein expression is more restricted as discussed in Chapters 2 and 3.

1.2 |. Identification and definition of human MAIT cells

MAIT cells have been identified with flow cytometry by detection of TCR Vα7.2 expression and high expression of the c-type lectin CD161 (KLRB1).⁸ The cytokine receptor IL-18Rα, the ectopeptidase CD26, and chemokine receptor CCR6 have additionally been used with/or without TCR Vα7.2 to identify MAIT cells.^8,23,24 Discovery of MR1 ligands led to the creation of the MR1 tetramer [ligand: 5-OP-RU] (hereafter referred to as MR1 Tet) allowing for the identification of MAIT cells by TCR specificity.^3,19 Comparative studies have since determined that blood MAIT cells identified by TCR Vα7.2+ and CD161high expression are nearly, but not fully congruent with the cell population that binds to MR1 Tet.²⁵ Comparisons revealed that not all MR1 Tet-binding T cells express TCR Vα7.2+, and furthermore, that some TCR Vα7.2+, CD161high cells do not bind to MR1 Tet. Thus, two different definitions have been used in studies—the classic MAIT cell definition based on expression of Vα7.2+ in context of high CD161 expression, and the tetramer-based definition of MR1-restricted T cells. Of note, as the MR1 tetramer is loaded with one specific ligand, 5-OP-RU, and considering the variable TCR beta chain use, it may not be able to identify all MR1-restricted T cells. Given the large congruence of the cell populations identified with both definitions, we will refer to these cells as MAIT cells throughout the review, but specify the definition when particularly relevant in context of a study. Finally, it is noteworthy that T cells, including MAIT cells, internalize the TCR following engagement. In the absence of the ability to use the TCR to identify MAIT cells, a recently described core gene signatures (including KLRB1, ZBTB16, and SLC4A10)²⁶ may prove useful to identify MAIT cells in tissues that could otherwise potentially be missed due to low TCR expression.

1.3 |. General phenotypic and functional properties of human MAIT cells in peripheral blood

MAIT cells from human peripheral blood (usually processed as peripheral blood mononuclear cells, PBMC) have been well-characterized phenotypically. Most MAIT cells in human blood are CD8+, but a fraction of MAIT cells express CD4, and some express neither CD4 nor CD8.⁸ MAIT cells are often referred to as having an “effector memory” phenotype analogous to conventional (MHC class I-restricted) effector memory CD8 T cells which lack expression of CD62L and CCR7,²⁷ and thus cannot enter lymph nodes via the high endothelial venules (HEVs).^28,29 In line with this expression pattern, MAIT cells are found in human lymph but in lower frequency compared to blood.³⁰ Presumably, these MAIT cells entered the lymph by exiting tissue via lymphatic capillaries. Of note, MAIT cells express CD45RO akin to conventional central memory CD8 T cells (which are CD62L+ and CCR7+), but express limited CD45RA (³¹ and our own unpublished observation). MAIT cells in the blood appear CD27hi, CD28+, and have limited PD-1 expression,^8,32 thus resembling a typical central memory phenotype rather than an effector memory phenotype.^33,34 Thus, the commonly used “effector memory” designation for MAIT cells is somewhat unfortunate as it pertains to the trafficking properties of conventional effector memory CD8 T cells, but not necessarily other phenotypic characteristics.

In regards to their functional properties, most MAIT cells in peripheral blood have a typical CD8 effector T-cell response and express granzyme B and IFNg upon activation. MAIT cells in the blood express intermediate levels of T-bet, but high levels of Helios, and Eomes.³² Circulating MAIT cells express chemokine receptors CCR2 (ligands: CCL2, CCL7, CCL8, CCL13, and CCL16), CXCR4 (ligands: CXCL12, MIF), CCR5 (ligands: CCL3, CCL4, CCL3L1, and CCL5), CCR6 (ligand: CCL20), and CXCR6 (ligand: CXCL16) with intermediate levels of CCR9 (ligand: CCL25) (as previously reviewed in³⁵). Overall, these phenotypic traits—the cytotoxic properties and ability to respond to inflammation-elicited chemotactic cues—certainly fit with the notion that MAIT cells are able to perform antibacterial surveillance in tissues.¹⁷

1.4 |. General phenotypic and functional properties of MAIT cells in human tissues

Studies examining MAIT cells in human tissues revealed that some MAIT cells have the ability to secrete IL-17, while the majority of MAIT cells expressed granzyme B and IFNg upon activation.¹⁷ In line with the notion that MAIT cells in tissues express IL-17, a single-cell analysis approach showed that MAIT cells isolated from rectal punch biopsies express IL23R,²³ and MAIT cells also express the transcription factor RORγt,³⁶ the master regulator of Th17s, used by other innate/innate-like lymphocytes such as γδ T cells, NKT cells, NK cells, and innate lymphoid cells.^37,38 A small subset of MAIT cells in human mucosal tissues appears to be able to secrete the cytokine IL-22 upon stimulation,^7,32 which has been associated with restoring epithelial cell proliferation.³⁹

These observations could indicate that there are distinct MAIT cell lineages similar to Th1, Th2, and Th17 CD4 T cells and Types 1, 2 and 17 NKT cells. MAIT cells have also been reported to be able to provide B cell help akin to Tfh cells.^40,41 Of note, Foxp3 expression in MAIT cells has also been reported, but appears to be reflective of its activation state rather than indicative of a regulatory T-cell-like phenotype.⁴² It is also noteworthy in this context, that inflammatory signals (IL-12, -15, and -18) are sufficient to elicit surface expression of the inhibitory receptor CTLA-4 on MAIT cells.¹² This could allow MAIT cells to interfere with antigen-presenting cells (APCs) providing B7–1 and B7–2 co-stimulatory signals to CD28-expressing conventional T cells.⁴³ Alternatively, it may predominantly be a mechanism to curtail MAIT cell effector function.

Importantly, a recent single-cell RNA-Seq study by Garner and colleagues provided compelling evidence that MAIT cell functional properties reflect plasticity rather than distinct lineages.²⁶ Typically, the mouse model is ideally suited to track cell fate decisions and differentiation over time. However, as highlighted in other reviews and studies, murine MAIT cells appear to have an inherent Th17 lineage bias and thus appear distinct from the human MAIT cell population.^44–49 The notion of MAIT cell functional plasticity is intriguing, because it would allow MAIT cell to continuously integrate signals from the environment to adjust functional properties. Finally, the Garner et al. study also provided evidence for a link between TCR clonotype identity and transcriptional characteristics in MAIT cells.²⁶ This link strongly indicates a key role for the TCR signal in controlling MAIT cell function.

In this next section, we will highlight some important considerations when dissecting the consequences of TCR and cytokine signals on MAIT cells. Specifically, we will highlight important lessons learned from conventional T cells about the ability of the TCR to sense antigen (quantity and quality) and relay very distinct downstream signals, which further integrate and even synergize with cytokine signals. We then discuss the relevance of these findings for MAIT cells.

2.1 |. The challenges of properly discerning the consequences of T-cell receptor signals from pro-inflammatory signals

Conventional naïve CD8 T cells need to receive three signals to become fully activated, divide, and acquire effector function (expression of granzyme B, IFNg): a TCR signal (Signal 1), co-stimulation (Signal 2), and a pro-inflammatory signal (Signal 3, which can be provided by Type I interferon, IL-12, IL-21, and possibly other cytokines⁵⁰). A Signal 1 without a Signal 2 leads to tolerance.⁵¹ If Signals 1 and 2 are provided without a Signal 3, then the T cells become activated, but do not acquire effector function.⁵¹ Once conventional CD8 T cells differentiated into memory T cells, then effector function can be elicited either via the TCR or pro-inflammatory signals (as long as the TCR can still interact with self-peptide/MHC).^52,53 As outlined earlier, peripheral blood MAIT cells possess the trafficking properties of effector memory CD8 T cells, but the functional properties of central memory CD8 T cells (low/no PD-1 expression, granzyme B expression upon reactivation, but not directly ex vivo).⁵⁴ Early studies concluded that stimulating MAIT cells via their TCR was sufficient to elicit effector function (IFNg secretion, granzyme B expression), which seemed plausible given their memory-like phenotype. However, these initial experiments were not designed in a manner to directly assess TCR-driven activation of MAIT cells. It is important to consider that stimulation of PBMCs in a manner that activates conventional memory CD8 T cells leads to the secretion of cytokines that can elicit and amplify activation.⁵⁵ Importantly, even the activation of just a small subset of antigen-specific CD8 T cells is sufficient to then subsequently drive the activation of other memory phenotype T cells, which leads to acquisition of effector function and expression of biomarkers previously used to extrapolate TCR-mediated activation.⁵⁵ Similarly, “feeding” human monocytes with gently fixed bacteria or yeast does not only provide MAIT cell antigen in the form of metabolic intermediates of the riboflavin synthesis pathway, but also various TLR agonists that trigger secretion of pro-inflammatory cytokines. We previously demonstrated that TCR stimulation alone (in the absence of pro-inflammatory cytokines) is insufficient for MAIT cells to have sustained IFNg secretion. This observation could in part explain why MAIT cells do not acquire effector function in healthy human barrier tissues that are microbially colonized. Importantly, we also demonstrated that TCR- and cytokine-mediated signals together have a synergistic effect on the acquisition of effector function.²³ This is a critically important observation to consider for data interpretation to ensure that the role of an activating signal is neither over-interpreted nor underestimated: MAIT cells cocultured with bacteria/yeast-fed monocytes lose much of their effector function when anti-MR1 antibody is added to prevent MR1 interactions with the TCR.^23,54 Given the synergistic action between TCR and cytokine signals, the magnitude of an effector response should not be interpreted in an additive manner in that the TCR signal is being primarily responsible for acquiring effector function. Overall, these studies highlight that dissecting the effect of TCR from cytokine signals on MAIT cell effector function is not trivial due to these synergistic effects. Importantly, TCR-mediated activation of MAIT cells in the absence of pro-inflammatory signals is not inconsequential, but elicits a tissue repair program.⁵⁶

It may still be a premature assumption that antigen with agonist properties for MAIT cells is not present in context of viral infections as also thoughtfully noted in a recent review.⁵⁷ How could a viral infection potentially provide antigen for MAIT cells? The possibility for such a scenario comes from an elegant mouse model study by Brenner and colleagues that examined the activation of NKT cells following a fungal infection.^58,59 In this scenario, the NKT cells were not activated by fungal-derived lipid antigen, but self-derived lipid that was only present following TLR-mediated activation of host cells. Akin to this scenario, virally infected cells could alter host cells in a manner that triggers the production of metabolites or other MR1-binding antigen not found in an uninfected state. The Ito et al. study that was published after the initial submission of our review further highlights that it is critically important to stringently define antigenic origin: although CA7S is a host-derived metabolite, it is greatly reduced (>98%) in germ-free mice.²² Most studies examined MAIT cells in context of an abundant, strong agonist TCR signal, while the consequences of other types of TCR signals are still poorly understood. We will next summarize studies with conventional T cells that are important to consider for further defining the impact of TCR and cytokine signals for MAIT cell function, particularly in tissues.

2.2 |. A brief overview of signaling lessons learned from conventional T cells that could also be relevant for MAIT cells

Much of our understanding of how T cells function stems from the mouse model system and has shown to also hold up with human T cells despite some species-specific differences.⁶⁰ Given the functional similarities between conventional memory CD8 T cells and MAIT cells, we review some key insights from CD8 T-cell studies that would be challenging to mimic with human MAIT cells, but are highly relevant given the emerging role of the MAIT TCR signal in controlling functional properties.

Studies in the past 25 years have highlighted the complexity of T-cell receptor signaling for T-cell fate decisions beyond eliciting effector function. T-cell receptor signals are critical for the survival of naïve and memory CD8 T cells, but these TCR signals are “weak” signals that stem from interactions with low affinity self-peptide presented by MHC class I.^61–65 A study that deleted the TCRα chain on T cells in an inducible manner, estimated that the half-life of naïve CD8 T cells decreased from 162 to 16 days in the absence of a TCR, and the half-life of memory phenotype CD8 T cells without a TCR was estimated at 52 days (and was stable for TCR-sufficient CD8 T cells over the experimental time-course of 100 days).⁶⁶ Of note, dissecting the role of the TCR signal versus other survival signals such as IL-7 is experimentally challenging even in the mouse model system. This was revealed by a number of studies starting in the early 2000s highlighting that increased cytokine availability can turn a pro-survival signal into a proliferation-inducing signal.^61,67–69 Specifically, the depletion of T cells leads to an increase in cytokines such as IL-7 which can drive homeostatic expansion of the remaining T cells in a TCR-dependent manner.^61,70 Of note, interplay between IL-7 signaling and the TCR has also been reported for MAIT cells.⁷ Importantly, it is self-peptides presented by MHC class I that drive homeostatic expansion of CD8 T cells in the absence of a Signal 2 (co-stimulation).⁷¹ Whether MAIT cells similarly depend on receiving a (low affinity) TCR signal for long-term survival is unknown. Nearly every nucleated cell expresses MR1 transcript, but only very few cells express MR1 protein on the surface in the periphery.⁷² The antigen processing machinery that is required for loading metabolites on MR1 is starting to be better understood (reviewed in⁷²), but does need further investigation. Furthermore, the nature of the antigens that are bound on these MR1 cell surface expressing cells at steady state is unclear. Thus, it is unknown if there is a MAIT cell interaction with MR1 that would be equivalent to the self-peptide/MHC signals for conventional memory CD8 T cells (although the recent identification of CA7S suggests that such an equivalent signal exists²²).

As already briefly noted, the critically important role of cytokines, particularly cytokines that signal via the common gamma chain (IL-2, -4, -7, -9, -15, and -21), for T cell survival was reported in numerous studies in the early 2000s (reviewed in⁷³). The concentration-dependent effects of these cytokines on conventional T-cell survival, proliferation, and even induction of effector function is also highly relevant for considering MAIT cell function. For example, IL-15 signals are necessary for the homeostatic maintenance of memory CD8 T cells^74,75 and for the survival of NK cells.⁷⁶ However, at higher concentrations IL-15 can also induce cell proliferation and display pro-inflammatory properties.⁷⁷ IL-15 can even have superagonist properties when complexed with recombinant IL-15Ra and elicit substantial cell proliferation and even acquisition of effector function in memory CD8 T cells and NK cells.^78,79 Of note, this effect again depends on the availability of self-peptide MHC signals for conventional CD8 T cells.⁸⁰ The strength of the TCR signal can further shape the response to IL-15 delivered as a complex with IL-15Ra.⁸⁰ There is evidence for this interplay between TCR and IL-15 in MAIT cells as well: Sattler and colleagues reported that addition of IL-15 in context of suboptimal TCR activation increased MAIT cell effector function.⁸¹

The importance of the TCR signal strength in shaping the response of the CD8 T cells is well documented: clonal expansion is more limited and acquisition of effector function is diminished with lower affinity ligands.⁸² The interplay of TCR signal strength and pro-inflammatory signals together regulate the survival of responding CD8 T cells in a bim-dependent manner.⁸³ This documented interplay of cytokine signal and TCR signal strength on conventional T-cell fate decisions needs to be considered when interpreting MAIT cell experiments, particularly the ex vivo experiments that are needed to study human MAIT cells.

There are several other concepts around TCR signaling that are relevant for discussing and considering the plasticity of MAIT cell responses in tissues. A clever mouse model system developed by Labrecque and colleagues allowed for the tetracycline-inducible expression of the T-cell receptor.⁸⁴ A CD8 T cell expresses 3×10⁴ TCRs on the cell surface, and the tet-inducible system allowed for a titration from OFF (not Tet, no TCR), to 5×10², 5×10³, and 2×10⁴ TCRs per cell.⁸⁴ When antigen was abundant in the form of high affinity antigen, even 500 TCRs were sufficient to induce proliferation and acquisition of effector function indistinguishable from 2×10⁴ TCRs. However, a higher number of TCRs (5×10³) was required for full activation and proliferation when antigen was less abundant (but still high affinity).⁸⁴ MAIT cells appear to express the same quantity of TCR compared to conventional T cells, thus indicating that the number of TCRs expressed on a T cell allows for sensitive detection. What happens when antigen is really rare or of rather low affinity? We want to highlight a set of studies that demonstrate the exquisite sensitivity of TCR sensing. CD8 T cells can “count” and add up the TCR signals that they receive in the form of phosphorylated c-Jun even in the absence of apparent calcium mobilization and MAPK phosphorylation.⁸⁵ Whether MAIT cells operate in a similar manner is unknown, but it is important to consider that the TCR is not a simple ON/OFF switch, but allows for a nuanced surveillance of the environment. Data from Mark Davis’ lab indicate that a conventional T cell can detect a single peptide–MHC on the surface of an antigen-presenting cell.⁸⁶ This may seem puzzling given the observation that monomeric ligands are not stimulatory, but CD4 T cells appear to achieve this sensitivity by recognizing heterodimer of agonist peptide- and endogenous peptide–MHC complexes (with the help of CD4).⁸⁷ This is noteworthy given that MHC class I and II, and most likely MR1, are not homogenously loaded with the same antigen outside of experimental conditions, and may contain antigen with weak to strong agonist properties, but also null ligands and antagonists.^88,89 Briefly, peptide antigens with actual antagonistic effects on T cell activation and proliferation have been identified for conventional T cells,^88,89 and may also exist for MAIT cells (Figure 1). As highlighted in the introduction, MAIT cell effector function must be carefully regulated. If MAIT cells are as capable and nuanced as conventional T cells in sensing and responding to antigen, then there are presumably negative regulators or additional control signals in place to prevent unwanted MAIT cell responses, such as in microbially colonized, healthy tissues.

One of the most studied cytokines with an inhibitory function for T cells is TGF-β. Based on an enormous body of work with the mouse model system and work with naïve human T cells, TGF-β was established as a critically important negative regulator of CD8 T cell effector function.⁹⁰ Overall, TGF-β is associated with suppressing effector functions of CD8 T cells such as granzyme B and IFN-g expression.⁹⁰ Importantly, these studies typically relied on using genetic abrogation of TGF-β signaling, which precludes studying the dose-dependent effects of TGF-β.⁹¹ We recently examined the effect of TGF-β on the reactivation of (mouse) memory CD8 T cells with the following considerations that are also relevant for MAIT cells in tissues: (1) Availability of active TGF-β likely varies across tissue microenvironments and should thus be assessed across a titration curve. (2) Memory CD8 T cells can be reactivated by TCR signals as well as cytokine signals. (3) The strength of such reactivating stimuli can vary. The hypothesis that we tested was that the nature of the reactivating signal would affect the consequences of TGF-β signals. Contrary to the prevailing notion that TGF-β’s main role is as a master suppressor of CD8 T-cell effector functions, we found that TGF-β modulates several aspects of memory CD8 T cell function in a dose-dependent manner and based on the strength of the reactivation signal.⁹¹ We found that TGF-β suppressed granzyme B expression in a manner that was inversely proportional to the strength of the activating TCR or pro-inflammatory signals. However, even high doses of TGF-β had a very modest effect on IFNγ expression in the context of weak or strong reactivation signals. Unexpectedly, reactivation of memory CD8 T cells in the presence of TGF-β led to altered chemokine (CCL20) and chemokine receptor (CCR8, CXCR3, and others) expression.⁹¹ Again, these changes occurred in concert with TCR signals: TGF-β-induced expression of CCR8 was inversely proportional to the strength of the reactivating TCR signal. Finally, we found that TGF-β exerted these modulatory effects regardless if given before or after the reactivation signal.⁹¹ We could identify epigenetic changes as the mechanism for this temporal independence, which could potentially offer a mechanism to eliminate stochastically encountered order of signals as a variable to exerting function.⁹¹

We highlighted these studies to emphasize the complexity of TCR and cytokine signals that MAIT cells are likely to encounter in the tissues (Figure 1), and the still poorly understood cell fate decision processes that occur when T cells, including MAIT cells, integrate multiple different, possibly even opposing signals that eventually interact downstream at the transcription factor and epigenetic remodeling level. With this lens of signaling complexity we will next review MAIT cell function in two human mucosal tissues that are presumably quite distinct in regards to inflammatory processes and microbial colonization.

3.1 |. MAIT cell function/phenotype in oral mucosa

Healthy oral cavity is colonized with a diverse species of microbiota that is unique in composition to other anatomical site samples.^94–96 Periodontitis, an inflammatory disease of the gingiva, is a prevalent disease. Bacterial species containing riboflavin synthesis genes were found in both healthy and inflamed gingival tissues,⁹⁷ highlighting that bacterial-derived MAIT cell antigens are present in healthy tissues.

MAIT cells in healthy buccal and gingival oral mucosa were found in frequencies (% of T cells) equivalent to those found in blood.^11,12 Within buccal mucosal tissue, MAIT cells are reported to localize in the epithelial layer, but were also found in the underlying connective tissue.¹¹ Additionally, MR1 and HLA-DR-expressing cells were present in the buccal mucosa suggesting the presence of antigen-presenting cells capable of presenting antigen to MAIT cells.¹¹ In gingival and buccal tissues, MAIT cells had elevated expression of CD69 and CD103 indicative of tissue residence.^11,12 In terms of effector function, both resident (CD103+) and non-resident (CD103-) buccal mucosa-derived MAIT cells produced less TNFa and IFNg (upon ex vivo stimulation) compared to circulating MAIT cells. Buccal mucosa-derived MAIT cells produced more IL-17 than blood MAIT cells, and tissue resident CD103+ buccal mucosal-derived MAIT cells produced significantly more than CD103 buccal mucosal-derived MAIT cells.¹¹ By transcript, our lab also observed the increase in IL-17A and IL-17F in gingival mucosal derive MAIT cells compared to circulating counterparts.¹² More CCL20 and CXCL16 was detected in periodontitis gingival tissue compared to healthy tissue which may act to recruit MAIT cells via CCR6 or CXCR6, respectively.⁹⁸ It is important to note that a comparison between blood and tissue derived MAIT cells needs to be done carefully and take any collagenase-mediated effects on protein and transcript expression into account.^99,100 Further insight about potential changes is needed from studies that examine MAIT cells at the single-cell level in the same tissue, but across a gradient of mild to severe inflammation (which is feasible with gingival tissues¹⁰¹), which would allow for the assessment of inflammation-associated MAIT cell activation states.

3.2 |. MAIT cell function/phenotype in placenta

Following initial reports of a placental microbiome, more recent studies indicate that the placenta does not contain a microbiome.^102–105 We thus use the placenta here as a tissue with distinct environmental exposures compared to the oral mucosa, while examining MAIT cell functional properties. Within the placenta, MAIT cells appear in higher abundance (% of T cells) in the intervillous space of the placenta compared to peripheral blood.³¹ MAIT cells are also found within the decidual tissue (the maternal-derived base of the placental bed): MAIT cell abundance (% of T cells) and phenotype depend on the location (decidua parietalis versus decidua basalis) as well as the gestational age.^31,106–108 Overall, MAIT cells isolated from the decidua showed increased expression of CD25, HLA-DR, and CD69 with a decrease in CD127 expression (compared to peripheral blood) indicative of either migration of activated MAIT cells to the decidua, or potentially in situ activation in these healthy and possibly sterile human tissues.^31,107

In regards to their ex vivo functional properties, decidual MAIT cells displayed enhanced effector function: a higher % of decidual compared to peripheral blood MAIT cells produced IFNg and granzyme B when stimulated with paraformaldehyde fixed E. coli.³¹ Whether the enhanced effector function by decidual MAIT cells is intrinsic or due to indirect effects (differences in the composition of the mononuclear immune cell population between decidua and blood, etc.) will need to be determined. Similarly, when activated with PMA/Iono, most decidual MAIT cells produced TNFa and IFNg, while IL-17 expression was very limited,¹⁰⁸ particularly when compared to MAIT cells from the oral mucosa.¹¹ Of note, decidual MAIT cells did not appear to produce IL-22,¹⁰⁸ in contrast to MAIT cells in the female genital mucosa.³²

Studies also reported MR1 expression on the cell surface of immune populations isolated from the placenta, with monocytes/macrophages expressing the most MR1.^31,108 The identity of the antigen presented by MR1 in this context is intriguing given the lack of bacterial colonization. It could suggest that MR1 expressing immune cells migrate to the site, or that bacterially derived antigen is transported to the site. It could also be indicative of the presence of endogenous self-antigen.²²

Overall, MAIT cells in environmentally distinct tissues (such as oral mucosa and placenta) have functional commonalities as well as differences. Ideally, a thorough understanding of which signals are received by MAIT cells in tissues would then lead to linking specific signals (or signal combinations) to cell activation states.