Synthetic 5-amino-6-D-ribitylaminouracil paired with inflammatory stimuli facilitates MAIT cell expansion in vivo

Introduction Mucosal-associated invariant T (MAIT) cells are a population of innate-like T cells, which mediate host immunity to microbial infection by recognizing metabolite antigens derived from microbial riboflavin synthesis presented by the MHC-I-related protein 1 (MR1). Namely, the potent MAIT cell antigens, 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU) and 5-(2-oxoethylideneamino)-6-D-ribitylaminouracil (5-OE-RU), form via the condensation of the riboflavin precursor 5-amino-6-D-ribitylaminouracil (5-A-RU) with the reactive carbonyl species (RCS) methylglyoxal (MG) and glyoxal (G), respectively. Although MAIT cells are abundant in humans, they are rare in mice, and increasing their abundance using expansion protocols with antigen and adjuvant has been shown to facilitate their study in mouse models of infection and disease. Methods Here, we outline three methods to increase the abundance of MAIT cells in C57BL/6 mice using a combination of inflammatory stimuli, 5-A-RU and MG. Results Our data demonstrate that the administration of synthetic 5-A-RU in combination with one of three different inflammatory stimuli is sufficient to increase the frequency and absolute numbers of MAIT cells in C57BL/6 mice. The resultant boosted MAIT cells are functional and can provide protection against a lethal infection of Legionella longbeachae. Conclusion These results provide alternative methods for expanding MAIT cells with high doses of commercially available 5-A-RU (± MG) in the presence of various danger signals.

Introduction: Mucosal-associated invariant T (MAIT) cells are a population of innate-like T cells, which mediate host immunity to microbial infection by recognizing metabolite antigens derived from microbial riboflavin synthesis presented by the MHC-I-related protein 1 (MR1).Namely, the potent MAIT cell antigens, 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU) and 5-(2oxoethylideneamino)-6-D-ribitylaminouracil (5-OE-RU), form via the condensation of the riboflavin precursor 5-amino-6-D-ribitylaminouracil (5-A-RU) with the reactive carbonyl species (RCS) methylglyoxal (MG) and glyoxal (G), respectively.Although MAIT cells are abundant in humans, they are rare in mice, and increasing their abundance using expansion protocols with antigen and adjuvant has been shown to facilitate their study in mouse models of infection and disease.
Due to the relatively low abundance of MAIT cells in unaltered laboratory mouse strains (0.1% of circulating ab T cells) the study of MAIT cells in the context of diseases in vivo in mice can be difficult.Therefore, for the effective study of MAIT cells in the context of disease in vivo, methods of increasing the level of MAIT cells have been developed (13)(14)(15)(17)(18)(19)(20)(21).Significant MAIT cell expansion in mice has been demonstrated previously by administration of synthetic 5-OP-RU in combination with various danger signals (such as TLR agonists or defined cytokines), or bacterial infection with riboflavin autotrophs (14,15).These "boosting" methods provide an inflammatory environment that provides sufficient co-stimulation to activate and increase the MAIT cell frequency and abundance in the tissues from ~0.1-1% of total ab T cells in naive mice to upwards of 10-50% at day 7 post boosting and ∼10% during convalescence (15,17).This increase in MAIT cell number has enabled the characterization of MAIT cells in mouse models.Nonetheless, there are limitations to the current published MAIT cell boosting methods.For instance, while infection with Salmonella Typhimurium, Legionella longbeachae or Francisella tularensis has been shown to markedly increase the absolute numbers of MAIT cells within mice (13,15,17), these infections also skew the expanded MAIT cell populations to distinct functional phenotypes and are also associated with significant non-MAIT ab T cell accumulation, thus complicating downstream data interpretation (19).Furthermore, it may take several weeks for the bacteria to be cleared from the host (15,17) creating a risk of residual bacterial contamination when performing MAIT cell isolation.This limits the suitability of infection-induced MAIT cell boosting in some models.Sterile MAIT cell-boosting strategies have also been developed and may be preferred for targeted augmentation of MAIT cells while minimising off-target effects on other immune cells (13).These boosting methods rely on co-administration of danger signals to simulate an infection, producing a co-stimulatory environment, which, in combination with synthetic 5-OP-RU, robustly increases MAIT cell abundance and frequencies amongst ab T cells (13,19).
Synthetic 5-OP-RU (22) is not commercially available but can be produced from the conversion of 5-A-RU in a non-enzymatic manner (23).Therefore, we assessed whether an alternative boosting strategy involving the administration of commercially available synthetic 5-A-RU ± MG to mice together with costimuli, provided in the form of riboflavin pathway deficient bacteria, TLR9 agonist (CpG combo), or IL-23-Ig encoding plasmid DNA, was sufficient to induce MAIT cell expansion in vivo.Our results demonstrate that 5-A-RU ± MG, in combination with each one of the co-stimuli represent robust MAIT cell boosting strategies in mice.

Co-administration of synthetic 5-A-RU alone or with MG facilitates the expansion of MAIT cells in mice following infection with riboflavin-deficient bacteria
First, we tested the ability of 5-A-RU to facilitate MAIT cell accumulation using bacterial infection as a means of co-stimulation.S. Typhimurium strain HW501 has a deletion of the ribD and ribH genes (hereafter referred to as S. Typhimurium DRibD/H) preventing the production of 5-A-RU and subsequently riboflavin (15,24,25) and is, therefore, unable to produce known MAIT cell antigens (15).We reasoned that a sufficiently high concentration of synthetic 5-A-RU with or without MG in conjunction with S. Typhimurium DRibD/H infection would facilitate the production of MAIT cell antigens in vivo and restore MAIT cell accumulation to the levels observed with a riboflavin producing strain of S. Typhimurium, BRD509 (15).
Mice were infected intratracheally (IT) with S. Typhimurium DRibD/H, resuspended in PBS containing either 5-OP-RU, 5-A-RU, 5-A-RU that had been pre-incubated with MG (5-A-RU+MG), or MG alone.As a positive control, mice were infected with S. Typhimurium BRD509.Mice received three follow-up doses of either 5-OP-RU, 5-A-RU+MG, 5-A-RU, MG, or PBS on days 1, 2 and 4 (Figure 1A).Consistent with previous findings (15), both 5-OP-RU and 5-A-RU+MG fully restored MAIT cell frequency of ab T cells and absolute numbers in the lungs of mice infected with S. Typhimurium DRibD/H to levels comparable to those in mice infected with S. Typhimurium BRD509 (Figures 1B, D, E) (15).Compared to mice that had received S. Typhimurium DRibD/H alone, mice that also received 5-A-RU displayed an ~36-fold increase in MAIT cell absolute numbers (Figure 1D).No significant change in total ab T cell numbers or non-MAIT ab T cell numbers was observed between treated groups, indicating that the 5-OP-RU, 5-A-RU+MG and 5-A-RU were acting specifically on MAIT cells (Supplementary Figure 1A).Similarly, no significant differences in other immune cell subsets (NKT cells, gd T cells, NK cells, neutrophils, inflammatory monocytes, macrophages, conventional dendritic cells (cDCs, cDC1s and cDC2s) were detected among the groups of mice treated with the various compounds (Supplementary Figure 2A).The resultant expanded MAIT cells were predominantly CD4 and CD8 double negative across all treatments (Supplementary Figure 3A), in line with previous observations (15).Consistent with our previous work (14,17,19), infection with S. Typhimurium BRD509 induced a MAIT17 population, as defined by the expression pattern of transcription factors (TF) T-bet and RORgT (Figures 1C, F) (15,19).Similarly, mice infected with S. Typhimurium DRibD/H with and without 5-OP-RU, 5-A-RU+MG and 5-A-RU administration followed the same MAIT17 phenotype suggesting a dominant IL-17 cytokine secreting capacity upon boosting (Figures 1C, F).The accumulation of MAIT cells in the 5-A-RU alone group was ~2-fold lower than in the 5-A-RU+MG group (Figure 1D), suggesting that 5-A-RU alone in vivo is less efficient in inducing MAIT cell accumulation than 5-A-RU that had been pre-incubated with MG in vitro, which fully restored MAIT cell accumulation, relative to the S. Typhimurium DRibD/H with 5-OP-RU and S. Typhimurium BRD509 infections.Nevertheless, these data demonstrate that synthetic 5-A-RU is capable of stimulating MAIT cell accumulation in vivo when paired with a bacterial infection.

TLR9 agonist and synthetic 5-A-RU ± MG stimulation induces robust MAIT cell accumulation in mice in an MR1 dependent manner
Next, we examined MAIT cell accumulation in response to 5-A-RU paired with the synthetic TLR9-agonists of B-class and P-class CpG (CpG combo) which elicits an inflammatory environment that activates a suite of immune cells and provides the necessary costimulatory signals for MAIT cell activation in mice together with 5-OP-RU (13,(26)(27)(28)(29).We hypothesized that the combination of 5-A-RU, with or without MG, and CpG combo in the absence of pathogenic burden could stimulate MAIT cell accumulation.
Mice were administered intravenously (IV) with CpG combo in combination with either 5-OP-RU, 5-A-RU pre-incubated with MG (5-A-RU+MG), 5-A-RU alone, MG alone, or PBS alone as a vehicle control (Nil), followed by additional doses of respective ligands on days 1, 2 and 5 (Figure 2A).
As expected, the 5-OP-RU and 5-A-RU+MG treated mice showed a robust increase in MAIT cell accumulation (30-60-fold in the lungs and 20-fold in the liver) when compared to control groups of mice treated with CpG and MG or CpG alone (Figures 2B,  C, E, F and Supplementary Figure 4).Strikingly, in the livers of the mice treated with 5-A-RU, there was a similarly high increase in the accumulation of MAIT cells compared to the control groups, comparable to the mice treated with 5-OP-RU and 5-A-RU+MG (Figures 2B, C).This did not repeat in the lungs, where treatment of mice with 5-A-RU produced a 20-fold increase in MAIT cells from the control groups, compared to a 30-60-fold increase with 5-OP-RU or 5-A-RU+MG (Figures 2E, F).Consistent with this result, the percentage of MAIT cells of total ab T cells closely reflected trends seen in the absolute number of MAIT cells generated by 5-OP-RU, 5-A-RU+MG and 5-A-RU.To confirm the robust expansion of MAIT cells exhibited by the boosting methods was MR1dependent, we included control groups that received multiple doses of intraperitoneal anti-MR1 blocking antibodies (30) in addition to 5-A-RU+MG, 5-A-RU or CpG alone stimulation (Figures 2B, C, E, F).Functional blocking of MR1 significantly ameliorated MAIT cell accumulation in both lungs and liver in comparison to the non-blocked groups and are comparable in number to the Nil (CpG only) group and naive controls (Figures 2B, C, E, F).This confirmed the antigen-MR1 specific nature of the MAIT cell-boosting strategy.
Consistent with data using infection with S. Typhimurium DRibD/H to provide co-stimuli, we saw no significant differences in the accumulation of non-MAIT ab T cells, or other immune cells between groups, demonstrating minimal off-target effects from these ligands (Supplementary Figures 1B, 2B).Notably, we observed some tissue pathology (noticeable necrosis and/or thrombosis, Supplementary Figure 5) and leukocyte infiltration upon boosting with CpG combo in both the liver and the kidneys.However, tissue pathology was consistent between all treatment groups and the Nil group (CpG only).This suggests that the inflammation and leukocyte infiltration were likely due to the administration of the CpG combo and not the MAIT cell activating compounds 5-OP-RU, 5-A-RU+MG or 5-A-RU (Supplementary Figure 5).In contrast to infection with S. Typhimurium DRibD/H, CpG combo treatment together with 5-OP-RU or 5-A-RU+MG, yielded distinct coreceptor expression patterns by MAIT cells in both organs, involving a higher proportion of CD8 + MAIT cells compared to the 5-A-RU treatment groups or naïve mice (Supplementary Figure 3B).Despite these differences, expanded MAIT cells in both lungs and livers of mice stimulated with 5-A-RU+MG or 5-A-RU displayed a similar TF expression pattern as mice stimulated with 5-OP-RU (13,19), indicating that in all three boosting regimens caused a skewing towards MAIT1 phenotype (Figures 2D, G and Supplementary Figure 6A).This phenotype was most prevalent in the liver (Figure 2D and Supplementary Figure 6A) but was also present within the lungs (Figure 2G, Supplementary Figure 6A).Together, these data demonstrate that 5-A-RU administered with a CpG combo can facilitate MAIT cell accumulation in vivo to a number and phenotype comparable to MAIT cells boosted with CpG combo and 5-OP-RU.

Induction of IL-23 with administration of synthetic 5-A-RU induces MAIT cell accumulation in mice in an MR1 dependent manner
We have previously shown that MAIT cell accumulation can be achieved in vivo by inducing IL-23 expression in the presence of 5-OP-RU using hydrodynamic injection (HDI) of a DNA plasmid encoding IL-23-Ig (14,19).This treatment provides a single defined co-stimulatory signal (IL-23) to the mice.Therefore, we tested if the combination of 5-A-RU and HDI of IL-23-Ig, in the absence of infection or other co-stimuli, could induce MAIT cell accumulation in vivo.
Mice were initially administered with IL-23-Ig plasmid DNA (HDI), followed by 5-OP-RU, 5-A-RU+MG, 5-A-RU, MG, or PBS (Nil) after eight hours and again on day 2 (Figure 3A).Consistent with previously published data ( 14), both 5-OP-RU and 5-A-RU +MG facilitated a robust, nearly 60 and 80-fold increase in liver MAIT cells, respectively, and an ~45 and ~39-fold increase in lung MAIT cells, respectively, when compared to the control IL-23 alone group (Figures 3B, E and Supplementary Figure 7).Surprisingly, MAIT cell accumulation in the 5-A-RU treated mice was limited to a 6-fold increase from the IL-23 group in the lungs and an 11-fold increase in the liver (Figures 3B, E).Similar to boosting with CpG combo, MAIT cell accumulation in both the lungs and liver in the presence of 5-A-RU+MG or 5-A-RU was inhibited by the addition of anti-MR1 blocking antibodies (Figures 3B, C, E, F).The expression of CD4 and CD8 by MAIT cells treated with IL-23 was consistent with previously published data (19), with the majority of MAIT cells being double negative for coreceptors, and minor populations of CD4 and CD8 single positive cells present in all treatment groups (Supplementary Figure 3C).Consistent with CpG-treated mice, we did not observe differences in the recruitment of non-MAIT ab T cells, total ab T cells or other immune cells between groups.(Supplementary Figures 1C, 2C).As expected, no gross tissue pathology was observed in any of the treated groups (Supplementary Figure 5), however mild tissue inflammation and infiltration of leukocytes were noticeable in the liver and to a lesser extent in the kidneys of mice boosted with IL-23-Ig, suggesting that none of the compounds elicited additional tissue stress (Supplementary Figure 5).A majority of expanded MAIT cells in both lungs and livers of mice stimulated with 5-A-RU+MG or 5-A-RU were MAIT17 based on their TF expression pattern (Figures 3D, G and Supplementary Figure 6B), consistent with MAIT cells expanded in mice stimulated with 5-OP-RU (13,19).
Collectively these data show that the administration of high doses of the MAIT cell antigen precursor, 5-A-RU, with or without the addition of MG, can promote MAIT cell accumulation similarly to 5-OP-RU when accompanied by inflammatory stimuli.

MAIT cells boosted with IL-23-Ig provide protective immunity to microbial challenge
Boosted MAIT cells can be protective during microbial infection while also exhibiting the capacity to modulate other immune cells in various mouse models of disease (13,15,17,20,21).To establish whether cells boosted using IL-23-Ig and MAIT cell boosting compounds is sufficient to provide protective immunity, we used a model where MAIT cells from boosted C57BL/6 mice were transferred into immunocompromised (Rag2 -/-gC -/-) recipient mice before microbial challenge (13,17).We chose to focus on IL-23-Ig boosted MAIT cells as this method provides the simplest adjuvant signal and microbial protection using CpG combo boosted MAIT cells has already been demonstrated (31).For challenge, we chose to use a lethal dose of the clinically relevant pathogen L. longbeachae, which has been extensively characterised previously using this model (17).MAIT cells from naïve C57BL/6 mice were boosted with IL-23-Ig HDI and 5-OP-RU, 5-A-RU+MG or 5-A-RU alone (Figure 4A) and sorted at day 6 as per the gating strategy in Supplementary Figure 8.The protective effect of the transferred boosted MAIT cells was evaluated by monitoring the survival following infection and the bacterial burden in the lungs of mice that survived to the experimental endpoint (Figure 4A).Efficiency of MAIT cell transfer was confirmed at the experimental endpoint (Figure 4B).
All mice without transferred MAIT cells succumbed to L. longbeachae infection by day 9 (Figure 4C), presumably by the uncontrolled growth of bacteria (17).By contrast, all but one of the mice with transferred MAIT cells from donor mice boosted with either 5-OP-RU, 5-A-RU+MG or 5-A-RU alone survived to the experimental endpoint (Figure 4C).All surviving mice recovered from the initial weight loss after infection to regain normal body conditions (Figure 4E).Interestingly, mice that had received transfer of 5-A-RU boosted MAIT cells lost the least weight and recovered earliest after infection (Figure 4E).The purity of transferred MAIT cells at the experimental endpoint (Figure 4B), confirmed the observed protection was solely mediated by the transferred MAIT cells, with no detectable contaminating non-MAIT ab T cells (Figure 4F).Although the surviving mice had fully recovered or exceeded their baseline weight, surprisingly, bacterial counts remained substantial in the lungs 24 days post-infection (Figure 4D).Further, to our surprise, and in line with the trend in weight recovery between treatment groups, mice that received 5-A-RU boosted MAIT cells had a lower bacterial load in the lungs at the experimental endpoint compared to the mice that had received transfer of 5-OP-RU or 5-A-RU+MG boosted MAIT cells, despite each group receiving equal numbers of transferred MAIT cells and having similar levels of MAIT cell accumulation in the lungs 24 days post-infection (Figure 4E).In summary, these findings reflect the potential for MAIT cells to provide acute protection against infection as well as demonstrate their limitations in providing sterile immunity in the absence of a complete immune system, as we have demonstrated previously (13,17).

Study design
Our study aimed to determine if MAIT cell accumulation could be achieved in mice treated by commercially available compounds (5-A-RU±MG) with alternative co-stimuli from infection, administration of synthetic TLR9 agonists (CpG combo) or cytokine (IL-23-Ig).To this end we assessed the capacity of boosted MAIT cells to combat a clinically significant bacterial pathogen.We also sought to address concerns over the potential off-target impact of MAIT cell boosting regimens on tissues within the mice and in other immune cell subsets.Various animal mouse models and flow cytometry of mouse organs was used to enumerate MAIT cells, other ab T cells and other immune cell subsets.Mouse group numbers were selected for sufficient statistical power with one-way ANOVA.

Mice
Mice were bred and housed under specific pathogen free conditions in the Melbourne Bioresource Facility at the Peter Doherty Institute for Infection and Immunity.C57BL/6 (6-12 weeks of age) weeks of age and Rag2 -/-gC -/-(6-17 weeks of age) mice were used in experiments with the approval of the University of Melbourne's Animal Ethics Committee (10201, 1814616 and 23211).Mice were killed at either the humane endpoint or at the experimental endpoint, as per ethics approval.Investigators were not blinded for substance administration, monitoring for endpoint survival, or subsequent analyses.

Choice of compound dosage
For each boosting method (involving infection, CpG or IL-23 mediated co-stimulation) we used our previously established optimal doses of 5-OP-RU (50 pmol for administration with S. Typhimurium DRibD/H, 2 mmol for CpG combo and 200 pmol for IL-23-Ig) (13)(14)(15).We based the relative doses of 5-A-RU and 5-OP-RU in each experiment on matched MAIT cell accumulation observed in vivo.Matched accumulation was shown to be at an approximate 5-A-RU:5-OP-RU ratio of 650:1, leading us to choose a dose 650 times that of 5-OP-RU.The optimal ratio of 5-A-RU : MG to produce 5-OP-RU is 1:3.4,established in cell and MR1-free optimisation assays previously in the lab (unpublished data).A matched dose of 5-A-RU ± MG was used for downstream experiments after initial titration (Supplementary Figure 9): 1. Together with 2 × 10 7 CFU S. Typhimurium DRibD/H (IT, 50 mL): 5-OP-RU (50 pmol); 5-A-RU (32.5 nmol)+MG (110.5 nmol) and 5-A-RU (32.5 nmol), then 3 more times with individual compounds alone at days 1, 2 and 4 (Figure 1A).To prepare the 5-A-RU+MG premix, 5-A-RU was incubated with MG at a ratio of 1: 3.4 for 1 hour in the dark at room temperature in PBS.Compounds or premix were made in a master mix and were aliquoted and stored at -80°C for future use.

Bacterial strains and infection/inoculation of mice
S. Typhimurium BRD509 and S. Typhimurium HW501 (S.Typhimurium DRibD/H) inoculums were prepared as described previously (15,24,25).In brief, the bacteria were cultured statically overnight in Luria Bertani (LB) broth containing 50 mg/mL streptomycin for S. Typhimurium BRD509, 30 mg/mL kanamycin and 50 mg/mL streptomycin with supplementation of riboflavin 20 mg/mL, for S. Typhimurium DRibD/H.Bacteria were then reinoculated for a further 2-4 hours with further static culture and allowed to reach OD 600 of 0.4-0.6.Bacteria inoculums were made with the estimation that 1 OD600 = 5 x 10 8 CFU/mL.Once made, the inoculum was then prepared using PBS ± compound(s) were added to the bacteria at the indicated concentrations, for inoculation of mice.A sample of each inoculum was retained and plated onto Luria Agar with antibiotics and supplementation of riboflavin (for S. Typhimurium DRibD/H) for CFU verification.
L. longbeachae NSW150 was grown at 37 °C in buffered yeast extract broth supplemented with 30-50 mg/mL of streptomycin overnight shaking at 180 rpm.For the infecting inoculum, bacteria were re-inoculated in prewarmed media for a further 2-4 hours culture to reach a log growing phase (OD 600 ∼0.4).With the estimation that 1 OD600 = 5 × 10 8 CFU/mL, bacteria were washed and diluted in PBS supplemented with 2% BYE for intranasal (IN) delivery to mice to a concentration of 2 × 10 4 CFU/mL.A 50 mL inoculum (1500 CFU per mouse) was instilled through the nose of the mouse while unconscious.A sample of inoculum was spread onto buffered charcoal yeast extract agar plates with streptomycin for verification of bacterial concentration by CFU counting.

Bacterial counts in infected lungs
Bacterial burden in mice was determined by counting CFU obtained from plating homogenised lungs on BCYE agar plates containing 30 mg/mL streptomycin and colonies counted after 4 days at 37 °C under aerobic conditions.

Constructs, compounds, immunogens, and MR1 tetramers
The IL-23-Ig construct was generously provided by Professor Burkhard Becher, Switzerland.Synthetic 5-A-RU was purchased from the Ferrier Research Institute, Wellington, New Zealand through Prof. Gavin Painter.Synthetic 5-OP-RU was prepared in house as described previously (3).MG was purchased from Sigma.MR1-5-OP-RU and MR1-6-FP monomers and tetramers were made in-house as described previously ( 2
IL-23-Ig was administered to mice as described previously (14).In brief, mice received an HDI of 2 mg of IL-23-Ig PEF-BOS plasmid in 1.6-1.8mL (8-10% of body weight) of TransIT-EE Hydrodynamic Delivery Solution (MIR 5340, Mirus Bio LLC, Madison, WI, USA) injected over a period of 10 seconds.Six-eight hours post-HDI, mice were administered with compounds (5-OP-RU, 5-A-RU or 5-A-RU +MG) intravenously in a volume of 200 mL.Mice received individual compounds again on day 2. Mice were monitored closely and were killed day 6 post-inoculation.Anti-MR1 antibody blocking was performed as described above (days 0-1, 0, 1 and 3).

Preparation of cells for flow cytometry
Mouse organs were harvested from mice killed by CO 2 asphyxiation.Single-cell preparation and MR1-5-OP-RU-tetramer staining for flow cytometry analysis were performed as described previously (24).In brief, lungs were perfused through the heart using 10 mL of cold Roswell Park Memorial Institute medium (RPMI) and were finely chopped before digestion using 3 mg/mL Collagenase III (Worthington, Cat#LS004182) with 5 mg/mL DNAse in RPMI (Gibco, Cat#21870-076) with 2% foetal calf serum (FCS) for 90 minutes at 37°C.Digested lung tissue was passed through a 70 mm strainer, pelleted and resuspended in 5 mL of a tris ammonium chloride solution (0.14 M NH 4 Cl (Sigma, Cat#A9434), 0.017 mM Tris (pH7.5, Sigma, Cat#T1503), then adjusted pH to 7.2 with 2 M HCl) for 5 minutes to lyse red blood cells, before neutralization with PBS with 2% FCS.Livers were perfused through the hepatic portal vein using 10 mL of cold RPMI and were passed through a 70 mm strainer before the separation of lymphocytes using Percoll gradient centrifugation (37% layered on top of 70%).Lymphocytes were removed, pelleted and red blood cells lysed as needed, as described with lungs, before resuspension in PBS with 2% FCS.Kidneys were finely chopped and digested, as with lungs using collagenase for 90 minutes.Samples were then passed through a 70 mm filter before being separated using Percoll gradient centrifugation as with livers.Lymphocytes were removed pelleted and lysed as with lungs before resuspension in PBS with 2% FCS.Spleens were collected and passed through a 70 mm filter, pelleted and lysed as described with lungs before resuspension in PBS with 2% FCS.Blocking Ab (26.5: anti-human MR1 MoAb), were prepared in-house.

Flow cytometry
Cell surface marker staining: Prior to MR1-5-OP-RU-tetramer staining, cell suspensions were blocked for non-specific staining using unlabelled MR1-6-FP-tetramer and anti-Fc receptor antibody (2.4G2; in house) for 15 minutes at room temperature.Cells were then stained with antibody and tetramer cocktail in PBS with 2% FCS containing 7-AAD (1:500) for 30 minutes at room temperature in the dark.Cells were analysed on an LSR Fortessa X20 with UV upgrade (BD Biosciences).
Intranuclear marker staining: For intranuclear staining, cells were stained in Fixable Viability Dye e780 (eBiosciences, San Diego, CA) for 30 minutes on ice before surface staining as above.Intranuclear staining was conducted using the Transcription Factor Buffer set (eBiosciences), as per the manufacturer's instructions.
Data acquisition and analysis: Data were acquired on a LSR Fortessa X20 with UV upgrade (BD Biosciences) and data acquisition and analysis were carried out using Diva and FlowJo version 10.9.0 (BD Biosciences) respectively.Gating strategies are shown in Supplementary Figures 8 and 10.

MAIT cell sorting, adoptive transfer, and infection challenge
MAIT cells enriched by various vaccination regimens with IL-23 were FACS sorted as previously described (17).Briefly, on day 6 of MAIT cell boosting, mice were killed, single-cell suspensions were prepared and live CD3+CD45+MR1-5-OP-RU tetramer+ cells were sorted using BD FACS Aria III (BD Biosciences).Next, 1 × 10 5 MAIT cells were injected into the tail veins of recipient mice which subsequently received 0.1 mg each of anti-CD4 (GK1.5) and anti-CD8 (53-6.7)intraperitoneally on days 9 and 12 (3 and 7 days post transfer to Rag2-/-gC-/-mice) to deplete residual conventional T cells.Mice were rested for 2 weeks post-cell transfer to allow expansion and settlement of the MAIT cell population prior to subsequent infection challenge.Mice were then infected with 1500 CFU of L. longbeachae IN and weighed daily and assessed for visual signs of clinical disease, including inactivity, ruffled fur, laboured breathing and hunched behaviour.Animals that had lost ≥20% of their original body weight and/or displayed symptoms of pneumonia (deemed to have reached a humane end point) were humanely euthanised.Remaining mice were humanely euthanised at the experiment end time point (Day 24), and organs harvested for analysis of MAIT cells and bacterial burden.

Histopathology
Livers and kidneys were fixed in 10% neutral buffered formalin, embedded in paraffin; 5 mm sections were stained with Hematoxylin and Eosin and examined by a pathologist (blinded) under light microscopy for tissue histopathology and leukocyte cell infiltration.The histopathology scoring system used was for liver: 0 (Normal): Aggregates of extramedullary haematopoiesis within periportal regions, around central veins and occasionally within sinusoids. 1 (Mild): Small scattered infiltrates of leukocytes within hepatic parenchyma or periportal regions (<1 per 10X field). 2 (Moderate): Average 1-5 aggregates of leukocytes per 10X field.Infiltrates of leukocytes within up to 25% of lobules.3 (Severe):

Statistical analysis
Statistical tests were performed using GraphPad Prism (version 9.1 LaJolla, CA).Normal Distribution was determined using Shapiro-Wilk Normality tests.Normally distributed data were compared using one-way analysis of variance (ANOVA) tests.Data sets found not to be normally distributed were compared using Mann-Whitney U tests.

Discussion
MAIT cells are present at a relatively low frequency within most laboratory mouse strains (32) limiting their analysis in mouse models.Increasing MAIT cell numbers in mice to levels comparable to those in humans is a valuable approach when studying the role of MAIT cells in mouse models of infection or disease, as well as performing phenotypic and functional studies ex vivo.Several mouse MAIT cell boosting strategies have been developed to enable the greater study of MAIT cells, leading to key findings, such as both the protective and detrimental functional capacity of MAIT cells in controlling bacterial infections (13,14,17,18,20,33,34).These boosting strategies typically rely on the infection of mice with bacterial pathogens, or vaccination with 5-OP-RU which is not commercially available, difficult to produce and highly unstable (3).Here, we demonstrate three effective MAIT cell boosting strategies using commercially available 5-A-RU, the precursor to the potent MAIT cell antigens 5-OP-RU and 5-OE-RU (2,16).Notably, the amount of 5-A-RU we used to stimulate equivalent MAIT cell response was 2-3 orders of magnitude greater compared to 5-OP-RU, as previous research had shown that only 1% of 5-A-RU is converted to 5-OP-RU in aqueous media (3), suggesting 5-A-RU may have a poor conversion rate to known MAIT cell antigens in vivo.A potential consequence of using significantly higher quantities of 5-A-RU during boosting is the likely higher concentration of contaminants including unreacted 5-A-RU and other hypothetical adducts with RCS other than MG and G that may spontaneously form in vivo and should be considered within the context of downstream experiments.
Whether 5-A-RU is a physiological MR1 ligand that can drive MAIT cell activation has not been fully resolved.To our knowledge, there is no direct evidence that 5-A-RU has the capacity to bind to MR1.Notably, 5-A-RU is not predicted to form a Schiff base with MR1 (2, 4), which is a key feature of the potent MAIT cell antigens 5-OP-RU and 5-OE-RU (2,25,35).However, non-Schiff baseforming compounds are known to have the capacity to bind MR1 and some are weak agonists for MAIT cells (35,36).In vitro studies have demonstrated that the addition of 5-A-RU alone to cultured cells can stimulate MAIT cells, albeit to a lesser extent than with equivalent amounts of 5-OP-RU (2,37), which is consistent with the relative accumulation of MAIT cells using our 5-A-RU/5-OP-RU boosting in vivo.The ubiquitous presence of RCS, including MG and G, in cell cultures suggests that the observed MAIT cell activation results from the spontaneous conversion of 5-A-RU to adducts such as 5-OP-RU and 5-OE-RU.However, it is difficult to discern whether these are the only RCS that are capable of forming physiologically relevant adducts with 5-A-RU within the cell culture media.
In addition to appropriate antigen stimulation, MR1-dependent activation of MAIT cells in the periphery requires co-stimulation (13)(14)(15)19).However, a study using germ-free mice demonstrated that co-stimulation is not required to recapitulate MAIT cell expansion during development (37,38).Interestingly, Legoux et al., showed that administration of low doses of 5-OP-RU, but not equivalent molar amounts of 5-A-RU, could restore MAIT cell development within the thymus in germ-free mice (38), suggesting that host-derived metabolites could not react with 5-A-RU in vivo and contribute to MAIT cell activation/accumulation (37,38).Importantly, that study used low, matched doses of 5-A-RU and 5-OP-RU, which would likely result in unequal concentrations of MAIT cell agonists in vivo due to the poor conversion efficiency of 5-A-RU to other adducts.Thus, further investigations into whether a high concentration of 5-A-RU may also support MAIT cell development should be considered.
Despite most laboratory mouse strains having a low frequency of MAIT cells, a mouse line, B6-MAIT CAST , with substantially increased frequencies of MAIT cells, was generated by crossing C57BL/6 mice with CAST/EiJ mice (39).B6-MAIT CAST mice have been proven to be useful in studying the fundamental biology of MAIT cells, given access to the increased MAIT cell numbers, especially within tissues and in naïve status (18,(40)(41)(42)(43).However, some animal disease models involve genetic backgrounds other than C57BL/6 and it is time-consuming and costly to breed mice to a different genetic background and maintain all control mouse lines (e.g.MR1 -/-).More importantly, B6-MAIT CAST MAIT cells do not perfectly reflect the MAIT cell phenotype present in wild-type C57BL/6 mice.For instance, B6-MAIT CAST MAIT cells produce type 2 cytokines and do not express the full complement of chemokine receptors when compared to MAIT cells derived from wild-type mice (39,44).Thus, the MAIT cell boosting strategies presented here provide an alternative to genetically augmented MAIT cells.
Within the MAIT cell populations, we observed differences in CD4 and CD8 coreceptor expression as well as MAIT cell phenotype distribution depending on the co-stimuli provided.Infection with S. Typhimurium DRibD/H and administration of IL-23 skewed the population towards CD4, CD8 double negative MAIT cells as well as an increase in MAIT17; whereas CpG combo yielded a larger population of CD8 + MAIT cells and a bias towards MAIT1.MAIT1 cells are more enriched in the liver of mice where the majority of MAIT cells express CD8, while MAIT17 cells are more enriched at mucosal barriers where MAIT cells have a greater bias towards being coreceptor double negative (19).While not directly investigated here, this finding warrants further investigation for a potential role of coreceptor expression and MAIT cell functional phenotype (45,46).While we have demonstrated MAIT cell phenotype and function, we have not assessed the exhaustion status and longevity of MAIT cells using these three methods of boosting.Assessment of exhaustion markers such as PD-1, PD-L1 and Tim-3 (47) and the long-term persistence of boosted MAIT cells within mice warrants further investigation.
The optimised doses of individual components, particularly 5-A-RU, in the MAIT cell boosting regimens presented here are well tolerated by the mice and do not exceed the maximum dosage or number of inoculations previously reported for 5-OP-RU (18).The use of live bacteria, CpG and cytokine as adjuvants, has also been well established (14,15,20,27,48,49).Despite observing some signs of pathology in the liver and kidneys for each boosting method, the level of pathology with 5-OP-RU treated mice as well as with adjuvant treated (Nil) groups were similar, indicating that the use of 5-A-RU and 5-A-RU+MG are viable alternatives to 5-OP-RU in sterile and non-sterile boosting methods.
In summary, we demonstrate three alternative methods for boosting MAIT cells in vivo distinct from previously published methods.We demonstrate 5-A-RU±MG and various co-stimuli that are all commercially available and sufficient to induce significant accumulation of MAIT cells in vivo.Boosted MAIT cells display consistent functional phenotypes based on transcription factor and coreceptor expression (13,14,19).Critically, these boosted MAIT cells were demonstrated to be capable of conferring protection in an adoptive transfer infection model using L. longbeachae, a clinically relevant pathogen.Thus, we believe these MAIT cell boosting strategies will be helpful in expanding the breadth of research on MAIT cells in various mouse models of disease.
(Related to and Figures 2, 3   SUPPLEMENTARY FIGURE 8 (Related to all Figures, except Supplementary Figure 5): Cytometric gating strategy for murine MAIT cells.Murine MAIT cells were identified by gating on lymphocytes and then excluding doublets using forward and side scatter (height vs. area).7-AAD -CD45 + events (live lymphocytes) were gated, and ab T were identified as TCRb + CD19 -.Auto-fluorescent cells were excluded using violet channel 525/50 (autofluorescence exclusion gate) and any residual non-T cells were excluded based on negative MR1-5-OP-RU tetramer staining (total T cells).Non-MAIT ab T were gated as TCRb + MR1-5-OP-RU tetramer -, and MAIT cells were gated as MR1-5-OP-RU tetramer+ TCRb intermediate cells.Coreceptor expression of CD4 and CD8 was gated on non-MAIT, MAIT and total T cell populations.MAIT1 (T-bet high, RORgT low) and MAIT 17 (T-bet high or low, RORgT high) cells were separated based on T-bet and RORgT expression.This gating strategy was used for all experiments.
): Histopathology of liver and kidneys in MAIT cell boosting strategies: Histopathology of liver and kidneys in MAIT cell boosting strategies: Graphs showing the histopathological score in the livers (G, H) and kidneys (I, J) of mice treated with CpG combo (G, I) or IL-23 (H, J).Coloured graph bars represent average of histopathology scores.Each dot represents one mouse (n=2-3).(E-J) Representative hematoxylin and eosin (H&E) staining images of liver and kidney samples as per histopathology grading scores (Bar=100 mm, bottom-right in each image): (E) Normal mouse liver with no inflammation present.(F) Mouse liver with grade 1 inflammatory lesions.Low numbers of infiltrates of mononuclear leukocytes (arrow) are present within the parenchyma.(G) Mouse liver with grade 2 inflammatory lesions.Moderate numbers of infiltrates of mononuclear leukocytes (arrows) are present within the parenchyma.(H) Mouse liver with grade 3 inflammatory lesions.Frequent infiltrates of mononuclear leukocytes (arrows) are present throughout the parenchyma, with a vascular thrombus (asterisk) also present.(I) Normal mouse kidney displaying no histological lesions.(J) Mouse kidney displaying mild focal interstitial infiltrates of mononuclear leukocytes (arrow).

SUPPLEMENTARY FIGURE 6 (
Related to and Figures 2, 3): MAIT cell functional phenotyping of MAIT1 and MAIT17 by expression of hallmark transcription factors RORgT and T-bet.Flow cytometry plots of MAIT cells from indicated organs expressing RORgT and Tbet are depicted from mice treated with CpG (A) and IL-23-Ig (B), with or without MAIT stimulating compounds as indicated.MAIT1 (T-bet high, RORgT low, left gate) and MAIT17 (T-bet high or low, RORgT high, right gate) were shown in the gate.The numbers indicate the abundance as a percentage (%) of total MAIT cells.

SUPPLEMENTARY FIGURE 7 (
Related to Figure 3): MAIT cell enumeration in various tissue from boosting with IL-23-Ig with and without MAIT cell boosting compounds.(A) Representative flow cytometry plots of MAIT cells from the liver, lungs, kidney, spleen, and blood of mice treated with IL-23-Ig plus various compounds and controls.MAIT cells were gated and MAIT cell frequency as a percentage of total ab T cells shown.Absolute number (B-D) and frequency (E-G) of MAIT cells as a percentage of total ab T were displayed from indicated groups of mice: Blood: (B, E); Kidney: (C, F); Spleen: (D, G).Data show mean ± SEM and dots represent individual mice (n= 3).Statistical significance is indicated by ns (≥0.05) * (p<0.05),** (p<0.01),*** (p<0.001);**** (p<0.0001).One-way ANOVA with Tukey correction was performed on log-transformed data or percentage data.