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 ΔRibD/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 ΔRibD/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 ΔRibD/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 αβ T cells and absolute numbers in the lungs of mice infected with S. Typhimurium ΔRibD/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 ΔRibD/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 αβ T cell numbers or non-MAIT αβ 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, γδ 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 RORγT ( Figures 1C, F ) (15, 19). Similarly, mice infected with S. Typhimurium ΔRibD/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 ΔRibD/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.

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 co-stimulatory signals for MAIT cell activation in mice together with 5-OP-RU (13, 26–29). 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 αβ 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 MR1-dependent, 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 ΔRibD/H to provide co-stimuli, we saw no significant differences in the accumulation of non-MAIT αβ 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 ΔRibD/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.

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 αβ T cells, total αβ 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.

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 ^-/- γC ^-/-) 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 αβ 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).

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 αβ T cells and other immune cell subsets. Mouse group numbers were selected for sufficient statistical power with one-way ANOVA.

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 ^-/- γC ^-/- (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.

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 ΔRibD/H, 2 μmol for CpG combo and 200 pmol for IL-23-Ig) (13–15). 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 ):

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.

S. Typhimurium BRD509 and S. Typhimurium HW501 (S. Typhimurium ΔRibD/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 μg/mL streptomycin for S. Typhimurium BRD509, 30 μg/mL kanamycin and 50 μg/mL streptomycin with supplementation of riboflavin 20 μg/mL, for S. Typhimurium ΔRibD/H. Bacteria were then reinoculated for a further 2-4 hours with further static culture and allowed to reach OD₆₀₀ of 0.4-0.6. Bacteria inoculums were made with the estimation that 1 OD600 = 5 x 10⁸ 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 ΔRibD/H) for CFU verification.

Mice were infected intratracheally (IT) on day 1 with S. Typhimurium ΔRibD/H (2 × 10⁷ CFU) in 50 μL of PBS containing indicated compound(s). Mice received either: 5-OP-RU (50 pmol, 50 μL), 5-A-RU+MG premix (5-A-RU+MG: 32.5 nmol +110.5 nmol, 50 μL), 5-A-RU (5-A-RU 32.5 nmol, 50 μL), MG (110.5 nmol, 50 μL) or PBS (50 μL). Mice were given subsequent IT inoculations of compound(s) in 50 μL at days 1, 2 and 4. Mice were killed at day 6, and lungs were collected for flow cytometric analysis ( Figure 1A ). As S. Typhimurium BRD509 forms a persisting infection with a much lower dose (2 × 10⁶ CFU), different doses for S. Typhimurium ΔRibD/H and S. Typhimurium BRD509 were chosen to simulate similar sublethal infection; producing comparable CFU counts at day 6 post-infection (15).

L. longbeachae NSW150 was grown at 37 °C in buffered yeast extract broth supplemented with 30–50 μg/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 (OD₆₀₀ ∼0.4). With the estimation that 1 OD600 = 5 × 10⁸ CFU/mL, bacteria were washed and diluted in PBS supplemented with 2% BYE for intranasal (IN) delivery to mice to a concentration of 2 × 10⁴ CFU/mL. A 50 μL 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 burden in mice was determined by counting CFU obtained from plating homogenised lungs on BCYE agar plates containing 30 μg/mL streptomycin and colonies counted after 4 days at 37 °C under aerobic conditions.

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). B-class and P-class CpG (CpG combo) with the sequence: 5’T*C*G*T*C*G*T*T*T*T*G*T*C*G*T*T*T*T*G*T*C*G*T*TT*CG*T*CG*A*CG*A*T*CG*G*C*G*CG*C*G*C*C*G-3′ (*phosphorothioate linkage) non-methylated cytosine–guanosine oligonucleotides was purchased from Integrated DNA Technologies, Singapore.

MAIT bosting regimen and delivery timeline are depicted in Figures 1A , 2A and 3A . Briefly, CpG combo (10 nmol) and either 5-OP-RU, 5-A-RU+MG premix, 5-A-RU, MG, or PBS were combined and administered to mice intravenously on day 0. Mice received individual compounds alone three more times on days 1, 2 and 4. MR1 specific monoclonal antibody (clone 26.5) was injected intraperitoneally (IP) at a dose of 250 μg in 200 μL of PBS 4 times (at days 0-1, 0, 1 and 4).

IL-23-Ig was administered to mice as described previously (14). In brief, mice received an HDI of 2 μg of IL-23-Ig PEF-BOS plasmid in 1.6-1.8 mL (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 μL. 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).

Mouse organs were harvested from mice killed by CO₂ 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 μg/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 μm strainer, pelleted and resuspended in 5 mL of a tris ammonium chloride solution (0.14 M NH₄Cl (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 μm 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 μm 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 μm filter, pelleted and lysed as described with lungs before resuspension in PBS with 2% FCS.

Antibodies against murine CD4 (clone GK1.5, #552051, APC-Cy7, 1:200), CD19 (clone 1D3, #551001, PerCP Cy5.5, 1:200), CD45.2 (clone 104, #553772, FITC, 1:200), TCRβ (clone H57-597, #553174 and #553172, APC and PE, 1:200), CD44 (clone IM7, #612799, BUV737, 1:200) CD11b (clone M1/70, Cat #557657, APC-Cy7, 1:200), Ly6C (clone AL-21, Cat #563011, BV605, 1:200), CD103 (clone M290, Cat #740238, BUV395, 1:200), Ly6G (clone 1A8, Cat #560601, PE-Cy7, 1:200) were purchased from BD Biosciences (Franklin Lakes, NJ). Antibodies against murine CD8 (clone 53-6.7, #12-0081-83, PE, 1:800), T-bet (clone 4810, #25-5825-82, PE-Cy7 1:200), RORγT (clone B2D, #12-6981-80, APC, 1:200), CD49b (clone DX5, #25-5971-81, PE-Cy7, 1:200) were purchased from Thermo Fisher Scientific, eBioscience (San Diego CA). F4/80 (clone BM8, Cat #123116, APC, 1:200), MHCII (clone M5/114, Cat #107631, BV421, 1:200), CD11c (clone N418, Cat #117308, PE, 1:200), NK1.1 (clone PK136, Cat #108731, BV421, 1:200) were purchased from BioLegend (San Diego, CA).

Blocking Ab (26.5: anti-human MR1 MoAb), were prepared in-house.

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 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 × 10⁵ 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-/-γC-/- 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.

Livers and kidneys were fixed in 10% neutral buffered formalin, embedded in paraffin; 5 μm 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): Widespread foci of leukocytes, average >5 aggregates of leukocytes per 10X field (often also necrosis and/or thrombosis). For kidney: 0 (Normal): No significant lesions. 1 (Mild): Rare interstitial inflammatory infiltrates.

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.