MAIT cells kill E. coli–infected cells and reduce cell-associated bacterial loads in infected cells (V体育官网入口)

To study the antimicrobial effector functions of MAIT cells, we developed an approach using the drug-sensitive E. coli clinical strain EC120S isolated from a patient suffering from a bloodstream infection [23]. We also developed a method for short-term PB MAIT cell expansion to provide the cell numbers with high purity required to perform such assays (S1A and S1B Fig). The E. coli strain EC120S infected HeLa cells efficiently as determined by pHrodo red-labeled live bacteria (S1C and S1D Fig). pHrodo becomes fluorescent in the acidic environment of cellular endosomal compartments and can therefore be used to evaluate E. coli internalization [24]. EC120S infected, replicated, and remained viable within HeLa cells’ intracellular compartments for at least 24 h post-infection (S1E Fig) V体育官网. MAIT cells rapidly degranulated and killed the tested epithelial cell lines (HeLa and A549) infected with E. coli EC120S (Fig 1A–1D; S1F and S1G Fig) through an MR1-dependent mechanism (Fig 1C and 1D).

Because MAIT cells achieved complete lysis (Caspase [Casp]3⁺ amine-reactive dead cell marker [DCM]⁺) of the infected cells by 24 h, we examined MAIT cell antimicrobial activity after 3 h, when the cell membrane of most infected cells undergoing apoptosis remained intact, as shown by staining with amine-reactive cytoplasmic dyes (Fig 1B and 1C). Notably, bacterial viability inside both HeLa and A549 cells was significantly reduced in the presence of MAIT cells, via an MR1-dependent mechanism (Fig 1E, S1H Fig). To assess whether the reduced bacterial load reflected true bacterial killing by MAIT cells, or simply the release of bacteria into the supernatants during the assay, the bacterial loads in the cell lysates alone, or in lysates containing infected-cells and cell-free supernatants (total lysates), were enumerated. In both conditions, the presence of MAIT cells decreased bacterial load, indicating direct bacterial control (Fig 1F). Contrary to MAIT cells, Vα7.2⁻ non-MAIT T cells were not able to control bacterial growth (S1I–S1K Fig). In addition, the capacity of MAIT cells to degranulate and induce apoptosis of infected target cells was superior to that of Vα7.2⁻ T cells (S1I and S1J Fig). We next investigated whether MAIT cells at resting state were similarly able to mediate antimicrobial activity. Consistent with previous studies [24,25], resting MAIT cells purified using the Vα7.2 beads were inefficient in degranulation and killing of HeLa cells infected with E. coli (Fig 1G and 1H; S2A Fig) or pulsed with the positive control MR1 ligand 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU; S2B and S2C Fig). Interestingly, resting MAIT cells also failed to control intracellular bacterial loads in E. coli–infected cells (Fig 1I). To determine whether the inability of resting MAIT cells to mediate antimicrobial activity was simply due to their weak cytotoxicity, purified MAIT cells were stimulated with IL-2 + IL-7 for 2 d to allow up-regulation of cytotoxic molecules [24,25]. These cytokine-activated MAIT cells were able to degranulate (Fig 1G; S2A Fig) and efficiently kill E. coli−infected cells (Fig 1H; S2A Fig). Similar or greater response was also observed following stimulation with the positive control MR1 ligand 5-OP-RU (S2B and S2C Fig). However, these newly activated MAIT cells still failed to efficiently control bacterial loads within infected cells, compared to the bacterial control exerted by MAIT cells that had been cultured for 15 d (Fig 1I). Interestingly, this temporal regulation of MAIT cell antimicrobial activity appeared to correspond to the differential expression of cytolytic proteins by MAIT cells (Fig 1J; S2D and S2E Fig). These results indicate that antigen-activated MAIT cells have the capacity to kill infected target cells in an MR1-dependent manner through Casp3 activation and progressively over several days develop the ability to control bacterial growth within infected cells.

Tissue-resident MAIT cells acquire cytolytic proteins with similar kinetics but of lower magnitudes

MAIT cells are abundant in the mucosal surfaces that serve as barriers to the external microbe-rich environments. To investigate the cytolytic capacity of MAIT cells in mucosa, tissue-resident MAIT cells were obtained from the nasopharyngeal (NP) mucosa of healthy individuals undergoing nasal polyp removals. CD69⁺ CD103⁺ tissue-resident NP MAIT cells (S2F Fig) overall expressed a similar set of cytolytic proteins, although at somewhat lower proportion and magnitude of cytolytic proteins at baseline and following in vitro culture when compared to those of matched PB MAIT cells (Fig 1K and 1L). However, the kinetics of cytolytic protein acquisition in NP MAIT cells appeared to be comparable to that of PB MAIT cells (Fig 1L; S2G Fig). Finally, NP MAIT cells degranulated and killed HeLa cells infected with E. coli EC120S (Fig 1M). These findings suggest that tissue-resident MAIT cells are able to mount cytotoxicity against bacteria-infected cells within the mucosal tissue environment.

MAIT cells mediate antimicrobial activity through the cytolytic protein-dependent pathway (VSports app下载)

Next, using PB MAIT cells, the role of cytolytic proteins in MAIT cell antimicrobial activity against cell-associated form of E. coli was investigated. Initially, the intracellular levels of GrzA, GrzB, GrzK, Gnly, and perforin (Prf) were measured using flow cytometry to determine the transfer of cytolytic effector proteins into E. coli–infected cells after co-culture with MAIT cells. With the exception of GrzK, all measured cytolytic proteins were detected in HeLa target cells (Fig 2A), matching their expression pattern in MAIT cells (Fig 2A; S2D and S2E Fig). The inhibition of cell-associated bacterial growth associated negatively with cell viability (Fig 2B) but positively with apoptosis (Fig 2C) and MAIT cell expression of GrzB (Fig 2D and 2E), Gnly (Fig 2F), and co-expression of GrzB and Gnly (Fig 2G). Because GrzB was readily detected inside E. coli–infected target cells co-cultured with MAIT cells (Fig 2A), we evaluated GrzB activity within the infected target cells using a fluorescent GrzB substrate (GranToxiLux) [26]. Active GrzB was detected in HeLa target cells infected with E. coli (Fig 2H), indicating that MAIT cells delivered GrzB into infected cells.

In order to assess MAIT cell use of the cytolytic protein pathway to control cell-associated bacterial growth, cytolytic protein activity was selectively inhibited by pharmacological inhibitors. Ethylene glycol tetraacetic acid (EGTA), a Ca²⁺-specific chelator that inhibits the release of cytolytic granules, was used in the presence of Mg²⁺ supplementation. EGTA strongly decreased MAIT cell degranulation as determined by CD107a expression (Fig 2I and 2J), reduced the delivery of cytolytic proteins (S2H Fig) and Casp3 activation in infected target cells (Fig 2I and 2K), and abolished the bacterial control by MAIT cells (Fig 2L). To investigate the role of GrzA and GrzB in MAIT cell antimicrobial activity, their activation was blocked using nafamostat mesylate (NAM) and N-acetyl-L-isoleucyl-L-α-glutamyl-N-[(1S)-2-carboxy-1-formylethyl]-L-threoninamide trifluoroacetate (Ac-IETD-CHO), respectively [27,28]. NAM had no effect on MAIT cell degranulation, Casp3 activation, and bacterial counts (Fig 2I–2L). In contrast, although Ac-IETD-CHO had no effect on MAIT cell degranulation (Fig 2I and 2J), it strongly abrogated Casp3 activation in infected target cells (Fig 2I and 2K) and impaired bacterial control (Fig 2L). To assess whether apoptosis of infected cells also contributed to bacterial control, Casp3 activation was blocked using the pan-Casp inhibitor valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone (Pro-VAD-FMK). The pan-Casp inhibitor had no effect on MAIT cell degranulation (Fig 2I and 2J) but severely diminished Casp3 activation in infected cells without affecting bacterial counts (Fig 2I–2L). This suggests that bacterial control by MAIT cells was independent of infected cell apoptosis. Finally, we assessed whether interferon (IFN)γ, tumour necrosis factor (TNF), and interleukin (IL)-17A, pro-inflammatory cytokines produced by MAIT cells (S2I Fig), played any role in MAIT cell bacterial control by blocking with neutralizing antibodies (Abs). Blocking these cytokines did not significantly affect degranulation or Casp3 activation (Fig 2I–2K). IL-17A blockade, but not IFNγ or TNF blockade, only slightly diminished MAIT cell antimicrobial activity (Fig 2L). Altogether, these findings indicate that MAIT cells use the cytolytic protein pathway to kill bacterially infected cells, and this may play an important role in early control of cell-associated bacterial loads.

MAIT cells recognize and mediate antimicrobial activity against carbapenem-resistant VSports - E. coli clinical strains

Next, we tested the hypothesis that MAIT cells maintain antimicrobial activity against carbapenem-resistant E. coli (CREC) clinical isolates. To address this, we first investigated whether CREC strains EC234, EC241, EC362, and EC385 (S1 Table) were riboflavin autotrophs by culture in riboflavin assay medium (RAM), a riboflavin-deficient broth. All CREC isolates grew in both RAM and nutrient-rich lysogeny broth (LB) at comparable rates, and the addition of external riboflavin did not influence their growth (S3A–S3H Fig). Moreover, the expression of ribA, the gene encoding the first enzyme of the riboflavin biosynthesis pathway, was similar among the clinical isolates (S3I Fig). These findings confirmed that the CREC clinical isolates were riboflavin-synthesis competent, a requirement for generating the riboflavin-related bacterial metabolite antigens [7].

Next, the ability of CREC to stimulate MAIT cells was tested by incubating peripheral blood mononuclear cells (PBMCs) with fixed bacteria for 24 h. All CREC strains induced degranulation and expression of GrzB, IFNγ, TNF, and IL-17A by MAIT cells at comparable levels (Fig 3A; S3J Fig). Furthermore, MAIT cells displayed similar patterns of polyfunctionality against these diverse strains, although some differences in response profiles were noted (S3K Fig). These variations occurred despite similar uptake of the bacteria by the PBMCs (S3L Fig). MAIT cell responses to CREC were predominantly MR1-dependent although this differed somewhat between the strains (Fig 3B).

To assess the capacity of MAIT cells to kill CREC-pulsed cells, HeLa or 293T-hMR1 cells were pulsed with strains EC234, EC241, EC362, or with the drug-sensitive EC120S. In all cases, MAIT cells recognized and killed HeLa cells (Fig 3C and 3D) and 293T cells stably transfected with human MR1 (293T-hMR1) (S3M–S3O Fig) pulsed with CREC strains at a comparable levels, though some variations were also noted. To investigate whether MAIT cells can mediate antimicrobial activity against CREC strains, we selected strain EC241 because it most efficiently enters and is internalized by HeLa cells (S3P Fig). MAIT cells induced Casp3 activation in HeLa cells infected with EC241 in an MR1-dependent manner (Fig 3E), consistent with results using the drug-sensitive strain EC120S (Fig 1). Furthermore, MAIT cells significantly reduced the CREC cell-associated bacterial loads in an MR1-dependent fashion (Fig 3F). Taken together, MAIT cells recognize CREC, mediate killing of infected cells, and reduce cell-associated bacterial loads in an MR1-dependent manner.

"V体育平台登录" MAIT cells secrete high levels of cytolytic proteins and mediate antimicrobial activity against free-living CREC in the surrounding milieu

We next investigated whether MAIT cells release cytolytic proteins into the surrounding milieu in response to MR1-restricted cognate recognition of cells infected with a riboflavin-synthesis competent strain of E. coli. MAIT cells secreted high levels of GrzA, GrzB, and Gnly into the supernatants already after a short 3 h stimulation with E. coli–infected cells (Fig 4A). Because some cytolytic proteins have direct antimicrobial activity [29,30], the MAIT cell secretome ability to inhibit free-living, extracellular E. coli growth was investigated. To collect bacteria-free MAIT cell secretomes following TCR-stimulation, 293T-hMR1 cells were used as antigen-presenting cells and pulsed with the synthetic antigen 5-OP-RU [31] and co-cultured with Vα7.2 bead-purified MAIT cells (S4A Fig). Live E. coli strains were then incubated with supernatants obtained from these co-cultures (S4A Fig). MAIT cells strongly degranulated their cytolytic protein content following MR1-restricted TCR triggering, and these proteins accumulated at high levels in the supernatants after 24 h of co-culture (S4B and S4C Fig). To assess whether MAIT cell secretomes induce damage on extracellular bacteria, the cell-impermeable nucleic acid dye SYTOX Green was used [32] (S4D Fig). To allow detection of bacteria versus debris by flow cytometry, bacteria were prestained with the cell permeable SYTO 62 just before fixation (Fig 4B). In the presence of MAIT cell supernatants, the CREC strains EC234 and EC362 suffered an increase in membrane permeability, as seen by the significant increase of SYTOX Green and SYTO 62 staining when compared to the control supernatants (Fig 4B and 4C, S4E and S4F Fig). Finally, the bacterial counts of the E. coli strains exposed to the MAIT cell secretomes were significantly reduced, although not totally suppressed (Fig 4D). These results indicate that the MAIT cell secretomes display antimicrobial activity against extracellular E. coli by markedly increasing bacterial cell permeability.

To determine whether the suppression of extracellular bacteria was due to a bacteriostatic effect or true killing of the bacteria, a time-kill analysis [33] of live bacteria was performed using CREC strain EC362 with various starting inocula (S4G Fig). At the original high starting inoculum of 10⁵ CFU/mL, MAIT cell secretomes mildly but significantly inhibited the growth of EC362 at 2 h (S4G Fig), in agreement with the previous findings. However, the bacteria rapidly rebounded and grew over a 24 h period (S4G Fig). Interestingly, when the starting inocula were reduced to 10⁴ and 10³ CFU/mL, MAIT cell secretomes displayed stronger bacterial growth inhibition and true bacterial killing, as reflected by the drop in bacterial counts, lasting longer with lower starting inocula (S4G Fig). In summary, these findings suggest that the antimicrobial activity of MAIT cell secretomes against extracellular bacteria in this static culture system was influenced by the starting inoculum effect.

V体育官网入口 - MAIT cell secretomes enhance the bactericidal activity of carbapenems against extracellular CREC

Following the observation that MAIT cell secretomes can directly alter bacterial membrane permeability (Fig 4B–4D), we asked whether the MAIT cell secretome could enhance the bactericidal activity of carbapenems against CREC strains. To test this, CREC strains EC234 and EC362 were incubated in the presence of MAIT cell supernatants with titrated concentrations of the carbapenem antibiotics imipenem, ertapenem, and meropenem. Greatly enhanced permeability and damage to the bacteria was evident by the appearance of bacteria stained with high intensities for SYTOX Green (SYTOX Green^bright), in comparison with bacteria cultured in either MAIT cell supernatants or carbapenem antibiotics alone (Fig 4E and 4F; S4H and S4I Fig). Fluorescence intensity of SYTO 62 was also increased in bacteria treated with a combination of MAIT cell secretome and imipenem (S4J and S4K Fig).

Next, effects on the live bacterial counts were evaluated using a time-kill method [33]. Notably, MAIT cell secretomes in combination with imipenem killed the highly carbapenem-resistant E. coli strain EC234 (minimum inhibitory concentrations [MICs] to all carbapenems ≥32 μg/mL; S2 Table), as well as the XDR strain EC362 harboring further resistance to colistin (S1 Table). In both cases, imipenem concentrations were well below the strains’ respective MICs (Fig 4G and 4H). Growth rates of strains EC234 and EC362 cultured in MAIT cell supernatants in the presence of titrated concentrations of imipenem, were significantly slower compared to their respective controls (S5A and S5B Fig). These growth delays occurred at imipenem concentrations up to 4-fold lower than the MIC for EC234 and up to 16-fold lower than the MIC for EC362 (S5A and S5B Fig). Moreover, the MAIT cell secretome decreased the imipenem concentrations required to fully suppress the growth of strains EC234 and EC362 by at least 2- and 4-fold, respectively (S5C Fig). In the presence of imipenem, the antimicrobial activity of MAIT cell secretomes appeared to be superior to that of non-MAIT Vα7.2⁻ T cells from the same donors (S5D Fig). Finally, we evaluated whether the antimicrobial activity was derived from the MAIT cells themselves or from the dying 293T-hMR1 cells used as antigen-presenting cells in the co-culture system. Firstly, supernatants of Vα7.2 bead-purified and expanded MAIT cells cultured without 293T-hMR1 cells still displayed antimicrobial activity in the presence of imipenem against the XDR strain EC362 (S5E and S5F Fig). Secondly, blocking the apoptosis of 293T-hMR1 cells caused by MAIT cell cytotoxicity with the pan-Casp inhibitor Pro-VAD-FMK caused no significant loss of antimicrobial activity in the presence of imipenem against strains EC234 and EC362 (S5G and S5H Fig). Taken together, these findings suggest that there is a synergy between MAIT cell secretome antimicrobial activity and carbapenems, thereby enhancing the in vitro bactericidal activity of carbapenems against extracellular CREC strains.

MAIT cell secretome antimicrobial activity correlates with the levels of secreted cytolytic proteins

We next revisited the time-kill data (Fig 4G and 4H) and analyzed possible associations between the amounts of cytolytic proteins present in the MAIT cell supernatants and the antimicrobial activity when combined with imipenem. In the bacterial cultures in which growth was detected, there was significantly less Gnly and GrzB in the MAIT cell secretome (Fig 5A and 5B). However, no significant difference in GrzA and Prf levels was observed (Fig 5A and 5B). Moreover, there were negative correlations between viable bacterial counts and Gnly or GrzB levels within the MAIT cell secretomes (Fig 5C–5F) but no correlations with the levels of GrzA or Prf (S6A–S6D Fig). Interestingly, Gnly levels in the supernatants correlated positively with the length of the lag-phase for strains EC234 and EC362 in the presence of imipenem (S6E and S6F Fig). Furthermore, the MAIT cell secretomes that contained the highest concentrations of Gnly had significantly greater and longer inhibition of EC234 and EC362 (S6G Fig). Taken together, these findings suggest that cytolytic proteins, particularly Gnly and GrzB, may significantly contribute to the MAIT cell secretome antimicrobial activity against extracellular CREC in the presence of imipenem.

Gnly and GrzB contribute to the antimicrobial activity of MAIT cell secretomes and potentiate the bactericidal action of carbapenems against CREC

Finally, the functional contribution of different cytolytic proteins to the MAIT cell secretome antimicrobial activity against CREC was examined. To enable this, the assay dynamic range was enhanced by using strain EC362, which was more sensitive to the inhibitory effects of the MAIT cell secretomes (Fig 4H), and lower concentration of imipenem (2 μg/mL) to reduce its residual inhibitory effects. Blocking MAIT cell degranulation using EGTA + Mg²⁺ (S6H Fig) nearly completely abolished the MAIT cell secretome-mediated bacterial inhibition (S6I Fig), whereas blocking GrzB activity alone had no observable effect (S6I Fig). Similarly, blocking the activity of IFNγ, TNF, or IL-17A during the MAIT-293T-hMR1 cell co-culture period had no significant effect on the antimicrobial activity of harvested supernatants (S6I Fig). Importantly, specific depletion of Gnly from the supernatants (S6J Fig) significantly reduced the MAIT cell secretome inhibition of bacterial growth (S6K Fig), as well as the capacity to induce bacterial cell permeability (S6L Fig). Gnly was detected intrabacterially by flow cytometry in both EC234 and EC362 incubated with MAIT cell supernatants (Fig 5G and 5H, S7A Fig). GrzB was also detected, but mostly in Gnly-containing bacteria (Fig 5I), whereas no GrzA or Prf was detected inside bacterial cells (S7A and S7B Fig). Notably, bacterial cells containing both Gnly and GrzB had higher membrane permeability than those containing only Gnly (Fig 5J and 5K). Short-term 30 min concomitant treatment with imipenem did not increase Gnly levels inside bacterial cells despite the increase in bacterial permeability and damage (S7C Fig). This suggests that the increase in bacterial death following incubation with MAIT cell supernatants and imipenem (Fig 4H) was possibly caused by the higher imipenem penetration into bacterial cells following MAIT cell secretome-mediated increase in bacterial cell permeability. Collectively, these findings indicate that Gnly and GrzB contribute to the antimicrobial activity of MAIT cell secretomes against extracellular E. coli and potentiate the bactericidal action of carbapenem antibiotics against CREC (Fig 6).

Ethics statement

Human samples were collected after informed written consent was obtained from all donors in accordance with study protocols conforming to the provisions of the Declaration of Helsinki. Ethical approval was obtained from the National University of Singapore Institutional Review Board (NUS-IRB reference codes B-15-088 and H-18-029) and Singapore General Hospital Institutional Review Board (protocol 2019/2623).

Blood and tissue processing

PB was collected from healthy donors recruited at the apheresis unit, Bloodbank@HSA, Health Services Authority, Singapore. PBMCs were isolated by standard Ficoll-Histopaque density gradient separation (Ficoll-Histopaque Premium; GE Healthcare). After isolation, PBMCs were cryopreserved in liquid nitrogen until further use or immediately used for MAIT cell purification.

Anonymized matched blood and NP tissue samples were collected from healthy individuals from Singapore General Hospital; Singapore. PBMCs were isolated by density gradient centrifugation on Lymphoprep (Axis Shield). PBMCs and NP tissue biopsies were kept frozen in 10% DMSO (Sigma-Aldrich), 90% Fetal Bovine Serum (Hyclone). Before initiation of tissue-resident MAIT cell culture, frozen tissue biopsies were thawed, washed, and disinfected with 5× antibiotic-antimycotic solution (Gibco) supplemented-RPMI-1640 (Hyclone). Tissues were digested using human tumor dissociation kit (Miltenyi Biotec), then passed through nylon mesh to obtain single-cell suspensions.

MAIT cell expansion

MAIT cells were expanded using 2 different protocols in this study. For the first set of expansion, MAIT cell were isolated from freshly isolated PBMCs by positive selection using the 5-OP-RU-loaded human MR1 tetramer-PE using magnetic-activated cell sorting (MACS) and anti-PE microbeads (Miltenyi). MAIT cells (>95% purity) were cultured 6 to 7 d in serum-free and xeno-free ImmunoCult-XF T cell expansion medium (STEMCELL Technologies) in the presence of 1:100 ImmunoCult human CD3/CD28/CD2 polyclonal T cell activator (STEMCELL Technologies), 10 ng/mL recombinant human (rh)IL-7 (R&D Systems), 50 μg/mL gentamicin (Gibco), and 100 μg/mL normocin (Invivogen). In selected experiments, MAIT cells were purified using the anti-Vα7.2 beads as described [24], then either immediately used or cultured for 2 or 15 d in the presence of 5 ng/mL rhIL-2 (Peprotech) and 10 ng/mL rhIL-7 before use.

For the second set of expansion, cryopreserved PBMCs were thawed and cultured in ImmunoCult-XF T cell expansion medium supplemented with 8% (v/v) xeno-free CTS Immune Cell Serum Replacement (ThermoFisher Scientific), 5 ng/mL rhIL-2 (Peprotech), 10 ng/mL rhIL-7, 50 μg/mL gentamicin, and 100 μg/mL normocin. PBMCs were stimulated with 10 nM 5-OP-RU on day 0, 5, and 10, and the culture media was replenished every 2 to 3 d. On day 11, viable cells were isolated by Ficoll-Histopaque density gradient centrifugation. On day 15 through 17, cells were checked for MAIT cell purity and numbers by flow cytometry. Cells were immediately used for subsequent assays when MAIT cell purity >70%. Unless otherwise indicated, all MAIT cell in vitro expansion used the second set of expansion protocol.

"VSports在线直播" Bacterial cultures

The E. coli strains 1100–2 and BSV18 were obtained from the Coli Genetic Stock Center, Yale University; the DH5α strain was obtained from New England Biolab. Carbapenem-resistant isolates were identified from the Singapore General Hospital microbiological database and retrieved from the archived bacteria repository. Further details on strain identification and determination of resistance and susceptibility profiles can be found in S1 and S2 Tables and S1 Text.

For MAIT cell stimulation assays, all E. coli strains were grown overnight at 37 °C in Luria (lysogeny) broth (LB) with shaking as described [65]. Overnight cultures of E. coli were then subcultured 10-fold in LB and incubated at 37 °C with shaking until OD₆₀₀ = 0.5. In selected experiments, growth curves were monitored by reading absorbance at 600 nm in a microplate reader with discontinuous shaking for 18 h at 37 °C (Cytation 5, BioTek Instruments).

Preparation of MAIT cell antigen 5-OP-RU

The MAIT cell antigen 5-OP-RU was synthesised according to a previously published procedure [31]. It is stable in DMSO solutions but converts rapidly to much less active lumazine in aqueous media, the exposure time to which should be minimized as much as possible to maximize activity. All 5-OP-RU working solutions were diluted from the DMSO stocks with appropriate culture medium immediately prior to any functional assay.

MAIT cell functional and antimicrobial activity assays

MAIT cells within bulk PBMCs, identified as MR1-5-OP-RU⁺ Vα7.2⁺ CD161^hi CD3⁺ T cells, were activated using formaldehyde-fixed E. coli strains as indicated for 24 h as previously described [24]. In selected experiments, 20 μg/mL MR1 blocking mAb (26.5, Biolegend) or IgG2a isotype control (MOPC-173, Biolegend) were used [24]. In all activation assays, monensin (Golgi Stop, BD Biosciences) were added for the last 6 h of incubation.

MAIT cell cytotoxicity assay was performed as previously described [24,25]. Briefly, HeLa cells or human 293T cells stably transfected with human MR1 (293T-hMR1) as indicated were incubated in complete RPMI medium with formaldehyde-fixed E. coli at the microbial dose of 30 for 3 h before the addition of expanded MAIT cells at the effector to target cell ratio 5:1. A total of 2 nM 5-OP-RU was added as a positive control for MAIT cell killing of target cells. Anti-CD107a-BUV395 at 1:200 was added at the beginning of the assay to detect MAIT cell degranulation. After 24 h of co-culture, cells were stained to detect target cell apoptosis using anti-active Casp3 mAb (BD Biosciences).

MAIT cell antimicrobial activity assay was performed as described [66]. Briefly, the bacteria subcultures were resuspended in RPMI without serum and antibiotics (ASF-RPMI). Adherent target epithelial cell lines were infected with live E. coli for 3 h at a microbial dose of 30 for strain EC120S and 3 for strain EC241 at 37 °C / 5% CO₂. Infected cells were washed extensively with complete RPMI medium supplemented with 200 μg/mL gentamicin (Gibco) and further incubated for 1 h at 37 °C / 5% CO₂ to kill extracellular bacteria, then washed extensively with ASF-RPMI. Expanded MAIT cells were labeled with 1 μM CellTrace Violet (CTV) dye (Thermo Fisher Scientific) before co-cultured with HeLa cells at an E:T ratio of 5:1 and incubated for 3 h at 37 °C / 5% CO₂. For the live bacteria enumeration, a duplicate set of experimental wells were done in parallel. Supernatants from the first set were collected, followed by adherent cell lysis with 0.1% (v/v) Triton-X for 10 min at RT. Equivalent volume of LB broth were added to lysates and plated onto LB agar plates in duplicates, incubated at 37 °C for 18 to 24 h, and counted visually. For the second set, anti-CD107a-BUV395 (1:200) was added into the culture medium at the start of the assay to assess MAIT cell degranulation. Cell-free supernatants were harvested, snap frozen in liquid nitrogen, and stored at −80 °C until further use. Adherent cells were harvested using trypsin-EDTA (Gibco) and stained for flow cytometry as indicated.

In selected experiments, MAIT cells or infected HeLa cells were treated with various pharmacological inhibitors or mAbs before use in assays. Briefly, MAIT cells were pre-incubated for 1 h with 5 mM EGTA (Bioworld) supplemented with 1 mM MgCl₂ (Sigma-Aldrich), then diluted to 1 mM EGTA with 1 mM MgCl₂ in-assay, or with 10 μM nafamostat mesylate (Sigma-Aldrich), or 100 μM of the GrzB inhibitor II Ac-IETD-CHO (Merck) before co-culture with HeLa cells. HeLa cells were pretreated for 1 h before addition of MAIT cells with 10 μM CAS-BIND Pro Pan Caspase Inhibitor (Pro-VAD-FMK; Vergent Bioscience). For antibody-blocking experiments, 20 μg/mL anti-MR1 or IgG2a isotype control was added on HeLa cells 1 h prior to co-culture, or 10 μg/mL mAb to IFNγ (B27; Biolegend), TNF (MAb1; Biolegend), IL-17A (eBio64CAP17; Invitrogen) or IgG1 isotype control (MOPC-21; Biolegend) 15 min before co-culture.

GrzB activity detection

GrzB activity was measured with GranToxiLux PLUS! (GTL) Kit (Oncoimmunin, Inc). Briefly, HeLa cells were trypsinized and infected with E. coli as per MAIT cell antimicrobial assay and MAIT cells were co-cultured with HeLa cells at E:T ratio of 5:1. Co-cultured cells were centrifuged immediately and resuspended in GrzB substrate solution. The co-culture was done for 1 h at 37 °C, and cells were washed in the GTL Wash Buffer. Cells were stained with viability dye and immediately analyzed by flow cytometry.

Preparation of MAIT cell supernatants

293T-hMR1 cells were plated in a 96-well flat bottom plate at 37 °C / 5% CO₂ in complete RPMI medium. After overnight incubation, culture medium was replaced with the antibiotic-free ImmunoCult medium and 2 nM 5-OP-RU was added into each well for 2 h. Expanded MAIT cells were then further purified using the Vα7.2 beads as described [24] (MAIT cell purity >98%) and co-cultured with the 5-OP-RU-pulsed 293T-hMR1 cells at an E:T ratio of 10:1. Supernatants from the co-culture wells or 5-OP-RU-pulsed 293T-hMR1 cell control wells were collected 24 h after co-culture, clarified, and snap frozen in liquid nitrogen until further use. The cytokine and cytolytic protein contents of the supernatants were measured using the LEGENDplex human CD8/NK cell panel (Biolegend) as described [2].

MAIT cell secretome antimicrobial activity assay

Overnight bacteria subcultures were washed with PBS and resuspended in MAIT cell or control supernatants at 10⁵ CFU/mL and incubated at 37 °C in flat-bottom 96-well plates without shaking for 24 h. To enumerate live bacteria, bacterial suspensions were harvested at various time points as indicated, serially diluted in LB broth, plated in triplicates on LB agar plates, and incubated at 37 °C for 18 to 24 h and counted visually. In selected experiments, MAIT cell and control supernatants were spiked with imipenem monohydrate (Sigma-Aldrich) at various concentrations as indicated. A 3 log₁₀ reduction in bacterial counts from the baseline population over 24 h was considered as bactericidal.

To assess bacterial damage, the cell-impermeable SYTOX Green nucleic acid stain (Thermo Fisher Scientific) was added at 20 μM during the 2 h incubation of 10⁶ CFU/mL E. coli with MAIT cell or control supernatants, with or without carbapenems at the indicated concentrations. Dead bacteria controls were prepared by incubating the bacteria in 70% (v/v) ethanol for 30 min at RT, washed, then incubated with control medium supplemented with SYTOX Green. To allow identification of the bacteria versus debris by flow cytometry, the bacteria were stained with SYTO 62 red fluorescent nucleic acid stain (Thermo Fisher Scientific; 2.5 μM) during the last 15 min at 37 °C. Bacteria were then fixed with 1% formaldehyde for 20 min at 4 °C just prior to FACS acquisition. In selected experiments, to detect the intracellular cytolytic proteins in bacterial cells, 10⁵ CFU/mL of E. coli were cultured with control or MAIT cell supernatants in the presence of 20 μM SYTOX Green for 30 min at 37 °C without shaking. During the last 15 min of incubation, fluorochrome-conjugated mAbs against Gnly, GrzA, GrzB, and Prf (S3 Table), as well as SYTO 62 nucleic acid dye were added to the culture. Bacteria were then washed, fixed, and permeabilized with BD Cytofix/Cytoperm buffers (BD Biosciences), then restained with the same mAb cocktail for 30 min at 4 °C and washed once prior to FACS acquisition.

VSports - Flow cytometry analysis

Cell surface and intracellular staining for cytokines, cytotoxic molecules, and active Casp3 were performed as previously described [2]. Staining with the MR1 5-OP-RU and MR1 6-FP tetramers was performed for 40 min at room temperature (RT) [7] before proceeding to the surface and intracellular staining with other mAbs (S3 Table.) Samples were acquired on an LSRFortessa flow cytometer (BD Biosciences) equipped with 355, 405, 488, 561, and 640 nm lasers. Single-stained polystyrene beads (BD Biosciences) and the compensation platform in FACSDiva version 8.0.1 (BD Biosciences) or FlowJo software versions 9.9 and 10.5 (TreeStar) were used for compensation.

Statistical analysis

Statistical analyses were performed using Prism software version 8.3.0 (GraphPad). Data sets were first assessed for data normality distribution. Data presented as a heat map shows the mean, whereas data presented as line or bar graphs with error bars represent the mean and standard error. Box and whisker plots show median, the 10th to 90th percentile, and the interquartile range. Statistically significant differences between samples were determined as appropriate using the unpaired t test or Mann-Whitney’s test for unpaired samples, and the paired t test or Wilcoxon’s signed-rank test for paired samples. The Kruskal-Wallis one-way ANOVA, the Friedman test, ordinary ANOVA, the repeated-measures (RM) one-way ANOVA, or mixed-effects analysis followed by the appropriate post hoc test as indicated was used to detect differences across multiple samples. Correlations were assessed using the Pearson correlation or Spearman rank correlation for parametric or nonparametric data, respectively. Two-sided p < 0.05 were considered significant.