MR1 deficiency enhances IL-17-mediated allergic contact dermatitis

Major histocompatibility complex (MHC) class Ib molecules present antigens to subsets of T cells primarily involved in host defense against pathogenic microbes and influence the development of immune-mediated diseases. The MHC class Ib molecule MHC-related protein 1 (MR1) functions as a platform to select MR1-restricted T cells, including mucosal-associated invariant T (MAIT) cells in the thymus, and presents ligands to them in the periphery. MAIT cells constitute an innate-like T-cell subset that recognizes microbial vitamin B2 metabolites and plays a defensive role against microbes. In this study, we investigated the function of MR1 in allergic contact dermatitis (ACD) by examining wild-type (WT) and MR1-deficient (MR1-/-) mice in which ACD was induced with 2,4-dinitrofluorobenzene (DNFB). MR1-/- mice exhibited exaggerated ACD lesions compared with WT mice. More neutrophils were recruited in the lesions in MR1-/- mice than in WT mice. WT mice contained fewer MAIT cells in their skin lesions following elicitation with DNFB, and MR1-/- mice lacking MAIT cells exhibited a significant increase in IL-17-producing αβ and γδ T cells in the skin. Collectively, MR1-/- mice displayed exacerbated ACD from an early phase with an enhanced type 3 immune response, although the precise mechanism of this enhancement remains elusive.


Introduction
The skin is repeatedly exposed to various antigenic substances of natural origin, cosmetics, metal accessories, and medical products of artificial origins, in the broad context of the environment (1), and the immunogenicity of these substances as sensitizers has been investigated (2,3). Antigen-presenting cells (APC) in skin exposed to sensitizers migrate to the draining lymph nodes (dLN) via lymphatic vessels and present them in the context of gene products of self-major histocompatibility complex (MHC) class Ia and II to antigen (Ag)-specific T cells (4). Thus, Ag-specific CD8 + and CD4 + T cells are primed, and these Ag-specific T cells within the memory fraction may be activated upon Ag re-exposure and migrate to the site of Ag entry to induce allergic contact dermatitis (ACD). ACD is transferable with T cells but not with antibodies and is thus classified as T cell-mediated type IV hypersensitivity according to the Gell and Coombs classification (4).
An experimental model of ACD is often employed in mice by painting chemicals such as dinitrohalobenzene onto the skin to study the sensitization and elicitation phases in detail (4). In addition to T cells, various other immune and non-immune cells in the skin are involved in the pathogenesis of ACD, and the crosstalk among them has been studied (5-7). Keratinocytes are the main type of non-immune cells in the skin, constituting a barrier layer since they not only form a physical barrier against the entry of foreign substances and pathogens but also secrete IL-1b when sensing insults against the skin to transmit signals downstream to immune cells (5). The cells of innate immunity include Langerhans cells, dermal dendritic cells, macrophages, neutrophils, and mast cells, which present Ag information and affect the intensity of ACD (6). Natural killer (NK) cells and innate lymphoid cells (ILCs), lymphocytes without rearranged Ag-specific receptors, potentiate (NK and ILC1 in particular) or regulate (ILC2 in particular) the immune and inflammatory responses at both the sensitization and elicitation phases of ACD depending on the context (6, 7).
Innate-like lymphocytes with rearranged TCRs are also important components in ACD. Murine skin is known to harbor a special gd T-cell population referred to as dendritic epidermal T cells (DETCs) expressing invariant Vg3Vd1 (in Garman Nomenclature [GN], Vg5Vd1 in Heilig-Tonegawa Nomenclature [H-TN]) TCR in the epidermis (8). However, humans do not have an equivalent epidermal T-cell population, although they harbor Vd1 + and Vd2 + T cells in the epidermis and dermis (9). Murine DETCs express NKG2D, which recognizes stress molecules such as RAE-1 induced in keratinocytes when sensitizing chemicals are applied to the epidermis (10). Moreover, keratinocyte-derived IL-1b induces IL-17 expression by DETCs (11) and Vg2 + (in GN, Vg4 in H-TN) or Vg4 + (in GN, Vg6 in H−TN) gd T cell subsets, including others (collectively referred to as Tgd17 cells) in the dermis (9), where the latter appear to play a more important role in ACD.
The skin also harbors innate-type T cells with ab-type TCRs, including natural killer T (NKT) cells (12), and mucosal-associated invariant T (MAIT) cells (13), whose reactivities are restricted by MHC class Ib molecules, cluster differentiation 1d (CD1d), and MHC-like protein 1 (MR1), respectively. These T-cell subsets are also categorized as preset T cells and resemble each other in several ways (14): 1) They recognize non-peptide antigens of microbial origin in the context of the restricting class Ib (glycolipids/CD1d vs. vitamin B 2, 9 metabolite/MR1), 2) major subsets utilize invariant Va chain (mouse Va14Ja18/human Va24Ja18 vs. mouse Va19/Ja33/ human Va7.2Ja33) with limited yet diverse Vb chains, respectively, 3) the invariant subsets of T cells exhibit effector/memory phenotypes and may function as either effector or regulatory cells in health and diseases (15). Studies of these invariant T cells may provide insights as to controlling ACD with low-molecular-weight ligands without concerns about MHC barriers because the restriction molecule is homogenous in an allogeneic relationship and highly conserved even in xenogeneic combinations (14).
The role of NKT cells in ACD has already been investigated by employing gene knockout ( -/-) mice, CD1d -/-(whole NKT celldeficient), or Ja18 -/-(invariant NKT [iNKT] cell-deficient) mice compared with wild-type (WT) mice with a C57BL/6 or BALB/c background (16)(17)(18). Initial studies demonstrated that ear swelling was reduced in both CD1d -/and Ja18 -/mice, suggesting that iNKT cells appear to function in the initiation and enhancement of ACD through prompt induction of IL-4 after Ag exposure, with involvement of IgM + B-1 B cells and effector ab T cells (16,17). Subsequent studies revealed that the differential functions of iNKT cells were dependent on the contact sensitizers employed in each study, with iNKT cells playing either pathogenic or regulatory roles (18,19). Human studies have also demonstrated that iNKT cells are detected in ACD lesions, implying some critical roles (20, 21).
The involvement of another innate ab type T cell, MAIT cells, in ACD has been limited to date and has been reported for palladium allergy in the foot pad lesions of BALB/c mice, where MAIT TCR was detected with iNKT TCR and presumed to display Ag-specificity (22). The role of MAIT cell accumulation in the lesion remains elusive in the development of ACD as a player in either inflammatory or regulatory responses. Thus, in the present study, we examined the effect of MR1/MAIT deficiency on ACD by comparing DNFB-induced ACD in WT versus MR1 -/mice to probe for altered responses in MR1 -/mice. The involvement of other subsets of innate-like T cells was also revealed in MR1 -/mice, and their relevance in ACD is discussed.
2 Materials and methods 2.1 Mice C57BL/6 (B6) mice were purchased from CLEA Japan, Inc. (Tokyo, Japan) and B6.MR1 -/mice were kindly provided by Dr. Susan Gilfillan (Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, USA) (23) and housed and maintained in an animal facility at the Analysis Center for Integral Genomic Functions at Kitasato University School of Medicine. The mice were provided food and water ad libitum. All animals were humanely treated and housed under pathogen-free conditions. All experimental procedures involving mice conformed to the guidelines of the Animal Experimentation and Ethics Committee of Kitasato University School of Medicine (#2017-143, 2018-119, 2019-025, and 2022-079).

Cell preparation from pinnae and lymph nodes of treated mice
Right and left pinnae and inguinal lymph nodes (draining lymph nodes [dLN]) on the right and left sides of the mice were obtained after euthanasia using a confirmed procedure. The pinnae were used for histology, flow cytometry, functional analyses of infiltrated cells, and gene expression analyses. A single-cell suspension was prepared according to the protocol previously described with slight modifications (25). In brief, the removed pinnae were cut into pieces and incubated with 100 mg/mL Liberase ® and 400 ng/mL DNase I (both from Roche Diagnostics, K.K., Tokyo, Japan) at 37°C with gentle shaking for 1 h. The digestion was stopped by adding ice-cold phosphate-buffered saline without Ca 2+ and Mg 2+ [PBS (-)], and the solution was layered on Lympholyte ® -M medium (Cedarlane Laboratories Ltd., Ontario, Canada) followed by centrifugation at 1,800 × g for 20 min. Cells at the interface were collected, washed with medium, and used for flow cytometry and cell culture. The lymph nodes were gently dispersed using a frosted-glass homogenizer to obtain a single-cell suspension, which was used for flow cytometry and cell culture.

Quantification of cytokines
The concentration of Th1/Th2/Th17 cytokines in the culture supernatant was quantified by flow cytometry using a BD CBA Mouse Th1/Th2/Th17 Cytokine Kit (BD Biosciences, CA, USA) according to the manufacturer's protocol.

Histology and quantitative analyses of microscopic images
The pinna tissue was fixed with buffered formaldehyde solution (10%) (Fujifilm-Wako Pure Chemical), followed by the standard protocol for paraffin-embedded sections and hematoxylin-eosin (HE) staining. Images of the HE-stained tissue were captured using a BIOREVO microscope (BZ-X800, KEYENCE Corp., Osaka, Japan), and the thickness of the pinna was quantified using image analysis software (BZ-X) for the microscope, in addition to manual measurement with a digital micrometer, as described in Section 2.2.

Statistics
The results are presented as means ± standard deviation (s.d.). Statistical analysis between two groups was performed using the Mann-Whitney U test, and comparison among three groups was performed using ANOVA followed by Tukey-Kramer tests. Values with p < 0.05 were considered statistically significant.

MR1 -/mice develop augmented ACD
To examine the role of MR1/MAIT cells in ACD, WT and MR1 -/mice were sensitized with DNFB in an acetone/olive oil solvent on the abdominal skin and challenged five days later on the right pinna, and the increment in thickness of the pinna in each mouse was calculated. MR1 -/mice exhibited a significantly greater increase in ear swelling than WT mice on days 1 and 2 after DNFB challenge ( Figure 1A). MR1 -/mice exhibited thicker pinnae, with severe intercellular edema and augmented infiltration of inflammatory cells compared to WT mice, as shown by histopathology ( Figure 1B). The inflammatory cells in the DNFBpainted pinnae appeared to consist mainly of polymorphic neutrophils in both WT and MR1 -/mice. The mean ear thickness of pinna painted with DNFB was also quantified using histological images, and that of MR1 -/mice was greater than that in WT mice ( Figure 1C). Although the representative histology of the vehicle control in MR1 -/mice was slightly thicker than that in WT mice ( Figure 1B), the mean ear thickness in the control group was similar between the WT and MR1 -/mice.

More neutrophils are recruited into the ACD-induced pinna in MR1 -/mice
To analyze inflammatory cells infiltrating the pinna challenged with control vehicle or DNFB, cells infiltrated into the pinna were obtained by disintegration of the tissue and analyzed by flow cytometry, as described in the Materials and Methods. The acquired mononuclear cells were gated as described (Supplementary Figure 1A). The neutrophils in the pinnae were identified as Ly6G + CD11b + cells (Figure 2A). More neutrophils were recruited to the DNFB-challenged pinnae in both WT and MR1 -/mice ( Figure 2A, right panels) than to the control pinnae ( Figure 2A, left panels). Furthermore, the number and frequency of neutrophils in the challenged pinnae of MR1 -/mice were significantly higher than those in the pinnae of WT mice ( Figure 2B). The expression of genes related to neutrophil migration and survival, such as Cxcl1, Cxcl2, and Csf3, was significantly increased or tended to be increased in MR1 -/mice compared to WT mice ( Figure 2C). Il17, which stimulates the expression of these genes, also tended to be increased in MR1 -/mice after two days of DNFB challenge ( Figure 2C).
The Ly6G lo-(-) CD11b + population that appeared straight below the neutrophil gate in Figure

Increased dermal gd T cells in the pinnae of MR1 -/mice
To examine another major population of cells residing in control pinnae or infiltrating inflamed pinnae, we analyzed aband gd-type T cells of vehicle-and DNFB-treated pinnae by flow cytometry based on gating, as shown in Supplementary Figure 1B. Both ab and gd T cells were detected in vehicle control and DNFBpainted pinnae, and gd T cells were clearly separated according to the fluorescence intensity as epidermal (Epi: TCR hi ) and dermal (Der: TCR lo ) gd T cells ( Figure 3A) (25). Notably, in the pinnae of the vehicle control group, the contour of ab T cells was more evident in WT mice, whereas that of Der gd T cells was more evident in MR1 -/mice ( Figure 3A, upper and lower left panels). Elicitation by DNFB caused a reduction (dense contour to scarce one or dots) of Epi and Der gd T cells, whereas a clear population of ab T cells was observed in both WT and MR1 -/mice ( Figure 3A, upper and lower right panels). When the number of T cells was analyzed further, the ab T cells in the pinnae significantly increased after challenge with DNFB in both WT and MR1 -/mice at similar levels, suggesting that the sensitized population of ab T cells was vigorously recruited into the pinnae after painting in both strains of mice ( Figure 3B, lower right panels), although the percentage of ab T cells of MR1 -/mice was significantly lower than that of WT mice due to the increased Der gd T cells as described below ( Figure 3B, lower left panels). Accordingly, the percentage of Epi gd T cells was markedly decreased in the DNFB-challenged pinnae compared to the vehicle controls ( Figure 3B, upper left panels). The number of Epi gd T cells also appeared to be decreased in the DNFB-challenged group, whereas the extent was not as marked as that of the frequency, and MR1 -/mice exhibited higher numbers than WT mice ( Figure 3B, upper right panels). Another subset, Der gd T cells, appeared to be decreased in frequency in WT and MR1 -/mice in the DNFB-treated group under the influence of the dominant recruitment of ab T cells, whereas the frequency was significantly higher in MR1 -/mice than in WT mice in both the control and DNFB groups ( Figure 3B middle left panel). Moreover, the number of Der gd T cells was not reduced, even in WT mice, and was significantly increased in DNFB-challenged pinnae in MR1 -/mice compared with WT mice ( Figure 3B middle right panel). In contrast, the cells of interest in the present study, MAIT cells detected by 5-OP-RU/MR1 tetramer, were reduced in DNFBchallenged pinnae in comparison with those in the vehicle control, ( Figure 3C) both in terms of frequency and cell number ( Figure 3D). Since MR1 -/mice lack MAIT cells due to the Mr1disruption, there was no difference between the control and DNFB groups.
The increment and reduction of each T cell subset, as compared with other immune cells among the different panels, are not evident, because the cell numbers during the pre-and post-elicitation stages of each cell number differ over several log scales. To better visualize the relationship of each subset of cells in the pinnae of vehicle-and DNFB-painted WT or MR1 -/mice, the cumulative graph of cells for lymphocytic and phagocytic lineages is shown in Supplementary  Figure 3. The majority of cells infiltrating the pinnae consisted of neutrophils, Mo/Mf, and ab T cells in both WT and MR1 -/mice, although there was a difference in the composition between WT and MR1 -/mice in the vehicle control and DNFB-challenged pinnae.

Enhanced Th17 immune responses in MR1 -/mice
To examine the effect of MR1/MAIT cell deficiency on T-cell cytokine production in ACD, the whole draining LN (dLNs; inguinal) cells of abdominal skin from WT and MR1 -/mice were stimulated with 2,4-dinitrobenzene sulfonic acid (DNBS) in vitro. The level of IL-17A in the culture supernatant when stimulated with DNBS was significantly higher for the LN cells of MR1 -/mice than for those of WT mice (Figure 4). The production of other cytokines such as IL-10, TNF-a, IFN-g, and IL-6 was comparable between WT and MR1 -/mice (Figure 4), and IL-4 production was almost undetectable (data not shown).
Next, we examined the frequency and number of T helper (Th) subsets in the dLNs of WT and MR1 -/mice after 5 days of sensitization. There were no differences in the frequencies or the numbers of CD4 + CD3 + (T) cells in dLNs between WT and MR1 -/- mice ( Figure 5A). When T-bet + cells (Th1) were analyzed among CD4 + CD3 + cells, the frequency of T-bet + cells in the dLNs of MR -/mice was lower than that in WT mice, and the number of T-bet + CD4 + T cells also exhibited a decreasing trend ( Figure 5B). The CD4 + T cells were analyzed for RORgt and Foxp3 expression ( Figure 5C). The frequencies and numbers of RORgt + Foxp3 -(Th17) cells were significantly higher in MR1 -/than in WT mice ( Figure 5D, upper left and lower left panels). There were no differences in the frequency and number of RORgt -Foxp3 + (Treg) cells between WT and MR1 -/mice ( Figure 5D, upper and lower middle panels). Of note, RORgt + Foxp3 + cells, which may represent stable Treg effector cells (26), although a small population in comparison with RORgt -Foxp3 + cells, appeared more frequently (2×) in MR1 -/mice than in WT mice, as shown in Figure 5C. However, there were no statistical differences in the mean frequencies and cell numbers of the population between WT and MR1 -/mice ( Figure 5D, upper and lower right panels). Additionally, there were no differences in each fraction of Th cells in unsensitized mice (Supplementary Figures 4A, B). Consistent with the above findings, staining for intracellular cytokines in CD4 + T cells treated in vitro with PMA and ionomycin for 4 h ( Figure 5E) demonstrated that the CD4 + T cells of MR1 -/mice exhibited a higher frequency and number of IL-17A + cells than those of WT mice ( Figure 5F, upper and lower left panels), whereas frequency and number of IFN-g + cells differed between WT and MR1 -/mice ( Figure 5F, upper and lower right panels). These cytokine profiles are consistent with the data obtained from the culture experiments in Figure 4. Comparable production of IFN-g protein was found in the culture supernatant detected by CBA (Figure 4), as intracellular protein detected by flow cytometry (Figure 5E), whereas T-bet + T cells were reduced in frequency ( Figure 5B).

Increased IL-17A-producing dermal gd T cells in MR1 -/mice
We then examined the population of T cells in the pinnae of MR1 -/mice, because the source of IL-17A production was assumed to be Th17 cells as well as Tgd17 cells (27). Notably, the pinnae of the vehicle control mice contained dermal T cells at a higher frequency in MR1 -/mice ( Figures 3A, B). Der gd T cells, especially Vg2 + gd T cells, contain the Tgd17 cell population in the skin (26). Thus, we examined Der gd T cells for IL-17 expression in pinnae of unsensitized WT or MR1 -/mice after stimulation with PMA and ionomycin in vitro. Not only total T cells but also Der Vg2 + T cells exhibited a high frequency of IL-17A + -T cells in the pinnae of WT mice and an even higher frequency in MR1 -/mice than in WT mice under unsensitized conditions (Figures 6A, B). When Vg2 -Epi gd and Der gd T cells were analyzed for IL-17A in the same settings as in Figure 6A (gated as Vg2for Supplementary Figure 5A), Vg2 -Der gd T cells in MR1 -/mice were also significantly increased, but to a lesser extent than Vg2 + Der gd T cells, whereas the Vg2 -Epi gd T cells exhibited a decreasing trend (p = 0.05) in MR1 -/mice compared to WT mice in terms of the frequency of IL-17A + -cells (Supplementary Figures 5B, C). We next examined Der gd T cells for the expression of Vg2 in DNFB-challenged pinnae two days after elicitation ( Figure 6C). The number of Der gd T cells that expressed the Vg2 chain increased, even with a decreasing trend for the infiltration of ab T cells ( Figure 6D).
To examine CD4 + T cells in DNFB-challenged pinnae, we analyzed the cells obtained on day 2 of elicitation, and a similar frequency was observed for WT and MR1 -/mice, although an increasing trend in the number of Th cells was observed in MR1 -/mice compared to WT mice ( Figure 6E). The CD4 + T cells were also analyzed for the expression of Foxp3 and RORgt ( Figure 6F). Both the frequency and cell number of all subsets, RORgt + Foxp3 -(Th17; Figure 6G, left panels), RORgt -Foxp3 + (Treg; Figure 6G, middle panels), and RORgt + Foxp3 + (stable Treg Cytokine production by antigen-specific T cells in draining lymph nodes. Lymph node T cells harvested from inguinal lymph nodes at day 5 in vehicle-and DNFB-painted WT and MR1 -/mice were cultured for three days in the presence and absence of DNBS (100 mg/mL). Cytokines (IL-10, IL-17A, TNF-a, IFN-g, IL-6) in the supernatant were quantified as described in the Materials and Methods. Representative data of two experiments of three to four mice/experiment. Mann-Whitney U test. **p < 0.01. effector; Figure 6G, right panels), were increased in MR1 -/mice compared with those in WT mice. To explain the upstream events that led to the above differences, the relevant cytokine mRNAs were analyzed 6 h after elicitation. Both Il17 and Il1b expression were significantly increased in the pinnae of MR1 -/mice compared with WT mice as early as 6 h ( Figure 6H), suggesting that the expression of IL-1b might enhance the responses of both Th17 and Tgd17 cells. The ear swelling induced by DNFB challenge in MR1 -/mice was already augmented at 6 h (Supplementary Figure 6A), and the expression of genes relevant to neutrophils, such as Csf3, Cxcl1, and Cxcl2, was also increased (Supplementary Figure 6B), although neutrophil recruitment was comparable at this time point between WT and MR1 -/mice (Supplementary Figure 6C).

Discussion
In the present study, we demonstrated that ACD was augmented in MR1 -/mice compared to WT mice because of the increased numbers of Th17 and Tgd17 cells in MR1 -/mice. MAIT cells were markedly reduced upon elicitation with DNFB in WT mice. MAIT cells (5-OP-RU/MR1 tetramer + cells) in the dLN on day 3 of DNFB challenge expressed Nur77 in Nur77 gfp mice (data not shown), suggesting that MAIT cells were activated during the elicitation phase. Furthermore, the deficiency of MAIT cells appears to be related to an altered distribution and/or number of T cells and a bias towards the type 3 immune response in a direct or indirect manner, although the mechanism remains elusive.

MR1 deficiency may cause wider defects in MR1-restricted T cells (MR1T) (28) besides MAIT cells, as the diversity of MR1T
(including MAIT and MR1-reactive T) cells extends to six different groups with unique modes of recognition, binding, and reactivity (29), most of which are ab type but include a gd type, such as Vd3Vg8 T cells that bind and recognize MR1 at its membrane proximal region, similar to an a3 domain-recognizing antibody (30,31). Notably, the significant role of MR1T cells have been shown to play an important role in antitumor immunity (32) and have already been implicated in infectious and autoimmune diseases (33). It should be noted that T cells obtained from Va19Ja33 Tg Ca -/mice that overexpress MAIT cells were previously tested for suppression of delayed-type hypersensitivity (34). Transfer of invariant Va19 + T cells but not control non-transgenic T cells suppressed foot pad swelling induced by sheep red blood cells in B6 hosts prior to sensitization, accompanied by a reduction of serum IL-17 and IFN-g. Nevertheless, further studies will be needed to explore the mechanisms by which MR1T/MAIT cells interact with other immune cells to suppress ACD response.
The significant increase and bias towards Tgd17 cells we observed in the skin of MR1 -/mice is likely associated with the enhanced ACD response, although the timing and site of the developmental characteristics (35) in the biased distribution of Vg2 + Tgd17 in MR1 -/mice remain to be determined. Interestingly, MAIT cells and gd T cells have a reciprocal relationship, similar to the expansion of MAIT cells in NKT celldeficient mice (36). For instance, a patient with a homologous MR1 mutation at position 31 Arg to His substitution (position 9 in mature MR1 protein: MR1 R9H /MR1 R9H ) was discovered to display primary immunodeficiency due to functional MR1 deficiency, with no circulating MAIT cells (37). Notably, the patient had increased circulating T cells expressing Vg9Vd2 with the CD27 -CD28phenotype. Conversely, MAIT cells have been reported to be increased in gd -/mice (38). These results indicate that there is an equipoise among the three types (NKT, MAIT, and gd) of innatelike T cells by competing with a homeostatic factor or niche (39). Hence, it is tempting to speculate that NKT cells, MAIT cells and gd T cells all contribute in a reciprocal manner, as a meńage à trois, to various inflammatory diseases such as ACD. The present study demonstrated that in MR1 -/mice, the dominant Vg2 + Tgd17 cells in the skin ( Figure 6) and increased Th17 cells ( Figure 5) upon sensitization in the dLN migrated to the skin and enhanced ACD. It is intriguing that skin MAIT cells are biased towards IL-17 production (MAIT17) and promote tissue repair (38), and that their deficiency appears to be compensated by the dominance of Tgd17 cells in the skin. It is not known whether there are any direct interactions between MAIT cells and gd T cells that limit each other's effector functions. However, one may speculate that MAIT cells and gd T cells compete with each other for homing niches within the dermis, where MAIT cells localize near the dermalepidermal interface (38) and gd T cells localize to also in superficial regions (40). The most critical factor for MAIT cell tissue homing and homeostasis is likely their early life exposure to and sustained interaction with the microbiota that synthesize riboflavin (38). Tgd17 cells are similarly influenced by microbes for their expansion and functional activity (41), suggesting that skin commensals affects the balance between MAIT cells and gd T cells. Furthermore, cytokines such as IL-1b and IL23 (38) in the environment are also thought to be important factors that affect the balance between these T cell subsets.
The macroscopic and microscopic appearances of skin pathology were markedly enhanced with edema and cellular infiltration in MR1 -/mice compared with WT mice (Figure 1). In severe cases in MR1 -/mice, the elicited pinnae were covered with crustae by frequent scratching, and a large area of the inflammatory lesion was sometimes lost, presumably due to necrosis or injury, which was not observed in WT mice. Thus, the skin thickness data for severe cases were inevitably unincorporated in the analyses. The severity of dermatitis may permit use of ACD in MR1 -/mice as an intractable model system to study disease pathogenesis and testing immune therapies. Notably, MAIT cells have been reported to display tissue repair functions, as wound healing by punch biopsy was significantly delayed in the absence of MAIT cells (38). If the keratinocytes injured during ACD by cytotoxic lymphocytes fail to be replaced with newly proliferated cells, the epithelial defect may cause infections and further damage the skin. A recent study also revealed that amphiregulin, a member of the epidermal growth factor family produced by MAIT cells, accelerated wound closure, but in an MR1-independent manner (42). In experimental autoimmune uveoretinitis, MAIT cells ameliorated disease, which was associated with anti-inflammatory/neuroprotective activities of IL-22 as well as IL-22-independent repair functions upon stimulation with 5-OP-RU (43). Accordingly, the severity of ACD response in MR1 -/mice observed in our study may result in part from defective repair due to MR1T/MAIT cell deficiency.
In the absence of exogenous stimulation, MR1 -/mice exhibited a similar pinna thickness compared to WT mice ( Figure 1C), suggesting that MR1 -/mice do not develop spontaneous dermatitis. However, increased production of IL-1b in mutant mice than WT mice was detected at pinnae after 6 h of elicitation with DNFB ( Figure 6H), since the barrier function of the skin was presumably weakened in MR1 -/mice due to MAIT cell deficiency (38,44). The ear swelling in MR1 -/mice was more enhanced than WT mice at 6 h of elicitation (Supplementary Figure 6A), whereas the level of neutrophil migration was similar between the two strains (Supplementary Figure 6C), suggesting that edematous changes at the very early phase appeared to be different between MR1 -/and WT mice.
The cellular infiltrates consisted mainly of Mo/Mf, neutrophils, and ab T cells in both MR1 -/and WT mice after DNFB elicitation (Figure 2 and Supplementary Figures 2, 3). Notably, there were significantly more neutrophils in terms of percentage and actual cell numbers in MR1 -/mice than WT mice. The recruitment of infiltrates was concordant with the enhanced expression of cytokines and chemokines by Th17 and Tgd17 cells in the pinnae stained with DNFB, which supported neutrophil generation, recruitment, and activation ( Figures 2C, 4). Resident Mf were reduced in percentage due to, in part, dilution by the recruitment of Mo/Mf and a reduction in the actual cell number in both WT and MR1 -/mice (Supplementary Figures 2, 3). As for eosinophils in pinnae, the cell number per 10 mg tissue was not significantly increased in DNFB-challenged pinnae in MR1 -/mice compared with WT mice (data not shown). Although these changes result from the MR1T/MAIT cell deficiency, the underlying mechanisms remain to be further investigated.
When the T cells were compared in MR1 -/and WT mice, a large number of ab-type T cells specific for the sensitizer Ag in pinnae was equally recruited in both MR1 -/and WT mice after challenge with DNFB. Thus, the percentage of Epi and Der gd T-cell fractions decreased accordingly after challenge ( Figure 3B). The apparent reduction was simply due to dilution by the migrated ab T cells into the pinnae, whereas the number of gd T cells in each fraction increased after challenge to enhance the ACD response via production of cytokines and chemokines from Th17 and Tgd17 cells. Notably, MR1 -/mice harbored a significantly higher percentage of Der gd T cells, even in unsensitized states, and exhibited a higher percentage of IL-17A + cells in both the Vg2 + and Vg2fraction (Vg2 + > Vg2 -) upon in vitro stimulation with PMA and ionomycin ( Figures 6C, D; Supplementary Figures 5B, C). The abundance of Tgd17 cells in the skin of MR1 -/mice may result in a robust type 3 immune response at the site of ACD since more Tgd17 cells during the initiation phase in the dermis effectively boosted the response compared with WT mice.
The frequency of MAIT cells in mouse skin is strikingly different from that in human skin, with approximately 10% of ab T cells being MAIT cells in mice and 0.5%-2% of ab T cells being MAIT cells in humans, with the remainder being the conventional type and NKT cells (38,44). Therefore, the present results must be considered when assessing whether they are readily applicable to human cases of ACD. However, the involvement of innate T cells in ACD is not compromised in humans, as iNKT cells presumably participate as effectors (21) and the role of NKT cells in ACD may vary depending on different sensitizers (19). MAIT cells were detected in palladium allergy in a previous report (22), and the involvement of iNKT cells has already been demonstrated in allergies to metals such as nickel, cobalt, and chromium (21, [45][46][47] that are present in accessories, biomedical devices, and food constituents (1). It is intriguing to consider whether MAIT cells and iNKT cells adopt a common or distinct pathway that affects the ACD response. Notably, MAIT cells have been examined as promising targets for immunotherapy in the skin for phototherapy of atopic dermatitis (48) and as effectors of a major inflammatory disease, psoriasis (13). The utilization of MAIT ligands as therapeutic agents may be associated with low resistance by patients, since they are vitamin B-related compounds with either inhibitory (VB 9 -folate) or stimulatory (VB 2 -riboflavin) activities (14,49). To examine whether MAIT cells can be modulated to protect against ACD, further investigations that clarify their immunoregulatory role will be required.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Materials. Further inquiries can be directed to the corresponding author.

Ethics statement
The animal study was reviewed and approved by the Animal  Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material
The Gating strategy for the flow cytometric analyses performed in the present study. A. Gating strategy for detecting granulocytes. The acquisition was mostly ungated except for small particles (with very low FSC-A) and the acquired cells were widely gated with FSC-A/SSC-A as shown to include larger cells with intracellular granules. Cells were removed from doublets and dead cells followed by gating for the CD45 + population as shown in sequence. Finally, CD11b + Ly6G + cells in the CD45 + cell population were designated as neutrophils. B. Gating strategy for detecting T-cell subsets. The acquired cells were lymphocyte-gated in tighter FSC-A/SSC-A than that of A, as shown. The CD45 + population after removal of doublets and dead cells was further analyzed with TCRb/TCRgd or TCRgd/Vg2 to discriminate ab T, gd T, and Vg2 + T subsets in the gd T cell population. T-helper (Th) cell subsets in draining lymph nodes from unsensitized WT and MR1 -/mice. Cells in inguinal lymph nodes were obtained from each unsensitized strain of mice and stained for the analyses according to the Materials and Methods. A. Representative flow cytometric profiles of CD3 + CD4 + cells of the Foxp3 + and RORgt + population in WT and MR1 -/mice. B. Frequencies and cell numbers of RORgt + Foxp3 -(Th17; left panels), RORgt -Foxp3 + (Treg; middle panels), and RORgt + Foxp3 + (stable Treg effector; right panels) cells in WT and MR1 -/mice represented in panel A. Representative data of at least two experiments of four mice/experiment.  Figure 1A. B. The expression of Cxcl1, Cxcl2, and Csf3 related with neutrophil recruitment and activation was examined with mRNA obtained from the left pinnae (vehicle control) and the right pinnae (DNFB) of either WT (closed bar) or MR1 -/mice (open bar) at 6 h after challenge. C. Frequency of neutrophils in the CD45 + fraction in MR1 -/mice compared with those of WT mice as in 6 h after challenge (vehicle: closed bar, DNFB: open bar). Representative data of at least two experiments of four mice/experiment. Mann-Whitney U test. *p < 0.05