Mucosal-associated invariant T (MAIT) cells are abundant in the lung and contribute to host defense against infections. During bacterial infections, MAIT cell activation has been proposed to require T cell receptor (TCR)–mediated recognition of antigens derived from the riboflavin synthesis pathway presented by the antigen-presenting molecule MR1. MAIT cells can also be activated by cytokines in an MR1-independent manner, yet the contribution of MR1-dependent vs. -independent signals to MAIT cell functions in vivo remains unclear. Here, we use Klebsiella pneumoniae as a model of bacterial pneumonia and demonstrate that MAIT cell activation is independent of MR1 and primarily driven by type I interferons (IFNs). During Klebsiella infection, type I IFNs stimulate activation of murine and human MAIT cells, induce a Th1/cytotoxic transcriptional program, and modulate MAIT cell location within the lungs. Consequently, adoptive transfer or boosting of pulmonary MAIT cells protect mice from Klebsiella infection, with protection being dependent on direct type I IFN signaling on MAIT cells. These findings reveal type I IFNs as new molecular targets to manipulate MAIT cell functions during bacterial infections.
Introduction
Unconventional T cells comprise several families of lymphocytes that rapidly sense the presence of microbes at mucosal and non-mucosal surfaces, functioning as sentinels of tissue integrity during infections. Among the families of unconventional T cells, mucosal-associated invariant T (MAIT) cells have recently emerged as central players of immunity in the airways, performing protective roles against bacterial and viral infections (Godfrey et al., 2019; Leeansyah et al., 2021; Provine and Klenerman, 2020; Ussher et al., 2018). MAIT cells carry a semi-invariant TCR that recognizes antigens presented by the ubiquitously expressed MHC class I–related molecule 1 (MR1; Kjer-Nielsen et al., 2012). Several bacteria and fungi produce antigens to stimulate MAIT cells, which include metabolites derived from the riboflavin synthesis pathway (Corbett et al., 2014). Consequently, MAIT cells are activated in vitro by many of these pathogens and contribute to protective immune responses in vivo against pulmonary infection by bacteria such as Francisella tularensis or Legionella longbeachae (Meierovics et al., 2013; Meierovics and Cowley, 2016; Wang et al., 2018, 2019). In addition to MR1-TCR–dependent activation, MAIT cells can also be activated in response to cytokines through MR1-independent mechanisms. This is the case during viral infections in which cytokines are sufficient to stimulate MAIT cells. For instance, during infection with dengue, hepatitis C, or influenza virus, IL-18 drives MAIT cell activation in synergy with IL-12, IL-15, and/or type I IFNs (van Wilgenburg et al., 2016). Despite the increasingly well-characterized functions of MAIT cells, the mechanisms driving their local activation and effector function in the tissues remain poorly defined. Moreover, whether MR1-independent signals are sufficient to control MAIT cell functions in vivo during bacterial infections has not yet been addressed.
Type I IFNs (including IFNα and IFNβ) are known for driving a potent immunomodulatory action combating viral infections, yet their functions during bacterial infection remain unclear (McNab et al., 2015). Type I IFNs have been shown to play a protective role in some bacterial infections (Escherichia coli, Helicobacter pylori), but they appear to be harmful in others (Listeria monocytogenes). The mechanisms by which IFNα/β promote host protection or susceptibility to bacterial pathogens are poorly defined and seem to be bacterium dependent (Kovarik et al., 2016). Thus, understanding the mechanisms by which type I IFNs contribute to the immune response against bacterial infections is of utmost importance. Type I IFNs are part of a complex crossregulatory network and can directly control the activation and function of a variety of immune and non-immune cells. IFNα/β can act directly on conventional CD4 and CD8 T cells and have stimulatory or inhibitory effects on T cell survival and proliferation, cytokine production, and memory formation (Crouse et al., 2015; Kuka et al., 2019). In the case of MAIT cells, type I IFNs contribute to MAIT cell activation during viral infection or adenovirus vector vaccine by acting in synergy with IL-18 (Provine et al., 2021; van Wilgenburg et al., 2016). Furthermore, in vitro experiments suggest that IFNα can synergize with TCR signals to increase MAIT cell effector functions (Lamichhane et al., 2020; Pavlovic et al., 2020). Nonetheless, whether type I IFNs modulate MAIT cell activation and/or functions during bacterial infections and the relevance of this regulation in vivo remain unknown.
Klebsiella pneumoniae is a Gram-negative opportunistic enterobacterium causing severe pneumonia, sepsis, and urinary tract infections (Bengoechea and Sa Pessoa, 2019). This bacterium has been cataloged as a global health threat due to the increasing prevalence of antibiotic-resistant isolates and has been identified by the World Health Organization as of critical priority for the development of new treatments (Theuretzbacher, 2017). Previous studies suggested that MAIT cells are activated by and contribute to the control of K. pneumoniae (Georgel et al., 2011; Le Bourhis et al., 2010), although the mechanisms underlying these processes remain unknown. Klebsiella encodes genes involved in riboflavin biosynthesis (Vitreschak et al., 2002), suggesting an MR1-dependent mechanism of MAIT cell activation. However, we unexpectedly found that Klebsiella drives activation and effector functions of murine and human MAIT cells through an MR1-independent mechanism. Accordingly, our experiments demonstrate that Klebsiella induces activation, tissue relocation, and a Th1/cytotoxic transcriptional program in MAIT cells, all of which are regulated by type I IFN, independently on MR1-TCR signals. Consequently, transfer of MAIT cells to immune-deficient mice or expansion of pulmonary MAIT cells protect mice from Klebsiella infection, with protection being dependent on direct type I IFN signaling on MAIT cells. Our data indicate that type I IFNs drive the activation and antimicrobial function of MAIT cells during bacterial pneumonia, contributing to host protection.
Results and discussion
MR1-independent activation of MAIT cells during K. pneumoniae infection
While MAIT cells have been proposed to participate in the immune response against K. pneumoniae (Georgel et al., 2011; Le Bourhis et al., 2010), the signals driving MAIT cell activation in the tissues remain unknown. To investigate the mechanisms controlling the activation of pulmonary MAIT cells, we infected WT mice with K. pneumoniae (KP, strain Kp43816) through intranasal (i.n.) challenge with 5 × 104 live bacteria (Fig. 1 and Fig. S1). This strain is widely used to study the host response to Klebsiella as it recapitulates Klebsiella-triggered human pneumonia, inducing acute disease with fatal systemic spread at a low infectious dose (Chen et al., 2016a; Lawlor et al., 2005). Infected mice rapidly lost weight, and we recovered bacteria from the lungs of inoculated animals 24 h post-infection (hpi) and from distal organs (spleen and liver) at 48 hpi (Fig. S1, B and C). Phenotyping of pulmonary MAIT cells (defined as CD44+TCRβ+MR1(5-OP-RU)-Tetramer+; Fig. 1, A and B; and Fig. S1 A) indicated that these cells were readily activated in response to Klebsiella as we detected upregulation of the activation markers CD69 and CD25 (Fig. 1, A and B). MAIT cell activation was accompanied by an increase in MAIT cell frequencies and absolute numbers at 48 hpi (Fig. 1, A and B; and Fig. S1 D). We further phenotyped pulmonary MAIT cells at 48 hpi by examining the expression of transcription factors (T-bet and RORγt) and surface markers (Fig. S1, E and F). As previously described (Rahimpour et al., 2015), the majority of MAIT cells in the lungs of uninfected animals are T-bet−RORγt+ and lack expression of CD4 and CD8, and these proportions are not altered in response to Klebsiella infection (Fig. S1 F).
Since the K. pneumoniae genome contains genes encoding enzymes of the riboflavin pathway (Vitreschak et al., 2002), we investigated whether pulmonary MAIT cell activation was mediated by MR1-TCR engagement by injecting mice with αMR1 blocking antibody prior to infection (Fig. 1 C). Surprisingly, MR1 blocking did not affect MAIT cell activation, and MAIT cell absolute numbers (and frequencies), upregulation of CD69 and CD25, and lung bacterial burden (CFUs) were comparable for infected mice receiving αMR1 or isotype control (Fig. 1 C). Importantly, the same dose of αMR1 reduced activation markers in pulmonary MAIT cells after i.n. inoculation of mice with the MAIT cell ligand (5-(2-oxopropylideneamino)-6-D-ribitylaminouracil; 5-OP-RU), confirming its effectiveness (Fig. S1 G). Moreover, endogenous MAIT cell Nur77 (which is used as an indicator of TCR signaling; Moran et al., 2011) was not significantly upregulated in pulmonary MAIT cells in response to Klebsiella infection (vs. uninfected controls) or when comparing CD69+ vs. CD69− MAIT cells within infected animals (Fig. S1 H). However, we detected Nur77 upregulation in pulmonary MAIT cells after i.n. challenge with 5-OP-RU as well as in splenic cells after infection of mice with E. coli (Fig. S1 H). Thus, these data suggest that pulmonary MAIT cells are activated during K. pneumoniae infection thorough an MR1-independent mechanism.
Activation of MAIT cells during bacterial infections has been proposed to require TCR engagement, although during viral infections cytokines (such as IL-12 and IL-18) function in a TCR-independent manner to stimulate MAIT cells (van Wilgenburg et al., 2016). To investigate the signals controlling MAIT cell activation in response to Klebsiella, we cocultured in vitro murine T cells (enriched as B220− cells from spleen and lymph nodes) with RAW264.7 macrophages (which constitutively express surface MR1) in the presence/absence of Klebsiella for 3 h (Fig. 1 D and Fig. S1, I and J). As expected, Klebsiella induced the activation of MAIT cells (gated as CD44+TCRβ+MR1(5-OP-RU)-Tetramer+), as evidenced by CD69 upregulation, yet CD69 levels were unaffected by αMR1 blockade (1 μg/ml, Fig. 1 D). Moreover, blockade of other known MAIT cell-activating cytokines (IL-18, IL-12/23p40) alone, or in combination with αMR1, did not alter CD69 levels (Fig. 1 D). Similarly, MAIT cell CD69 was also unaltered when cells were cocultured with Klebsiella for a longer period of time (24 h) in the presence of a higher dose of αMR1 (10 μg/ml; Fig. S1 J). On the other hand, αMR1 blocked activation of MAIT cells in the presence of 5-OP-RU (Fig. 1 D, right), while αIL-18 or αIL-12/23p40 reduced MAIT cell CD69 in response to Staphylococcus aureus (SA, Fig. S1 J), confirming the functionality of these antibodies in our experimental settings. Comparable results were obtained while measuring activation of human MAIT cells in response to Klebsiella (Fig. 1 E). Coculture of human circulating CD8+ T cells (MAIT cells gated as CD3+CD8+CD161+Vα7.2+ cells) with autologous CD14+ monocytes in the presence of Klebsiella resulted in MAIT cell CD69 upregulation, which was independent of MR1 (1 μg/ml; Fig. 1 E). MAIT cell activation in response to Klebsiella was also unaltered by a higher dose of αMR1 (10 μg/ml), while this antibody efficiently blocked human MAIT cell CD69 upregulation in response to 5-OP-RU (Fig. 1 E, right).
To unequivocally address the requirement for microbial riboflavin metabolites in driving MAIT cell responses during Klebsiella infection, we constructed a K. pneumoniae mutant strain in which we deleted the ribD gene (KPΔribD; Fig. 1, F–H), which is essential for the production of 5-A-RU, the precursor of the MAIT cell antigens 5-OP-RU and 5-OE-RU (Corbett et al., 2014). We cocultured in vitro murine T cells (enriched as B220− cells from spleen and lymph nodes) with RAW264.7 macrophages in the presence of WT (KP, Kp43816) or KPΔribDKlebsiella for 3 h (Fig. 1 G). Strikingly, both WT and KPΔribD induced activation of MAIT cells to comparable levels as evidenced by CD69 upregulation (Fig. 1 G). Comparable results were obtained for human MAIT cells which when cultured with Kp43816 or the ribD mutant showed similar levels of CD69 induction (Fig. 1 H).
Collectively, these data suggest that K. pneumoniae mediates activation of murine and human MAIT cells primarily through an MR1-independent mechanism.
Klebsiella induces a dominant type I IFN signature in MAIT cells
Next, we investigated the signaling pathways driving activation of MAIT cells in response to Klebsiella. To obtain an unbiased overview of the phenotype and properties of MAIT cells in response to infection, we isolated human MAIT cells (from peripheral blood mononuclear cells [PBMCs] from four healthy donors) after overnight culture in the presence/absence of Klebsiella and analyzed their gene-expression profile by RNA sequencing (RNAseq). These analyses revealed 401 genes that were differentially expressed (DEG) in MAIT cells ± Klebsiella (adjusted P value <0.01, log2 fold change >0.5), with 255 upregulated and 146 downregulated genes (Fig. 2 and Fig. S2). As expected, MAIT cells were markedly activated by Klebsiella as analyses of genes involved in early T cell activation confirmed the upregulation of transcripts such as CD69 or IL2RA (CD25; Fig. 2, A and B; and Fig. S2). Gene set enrichment analyses (GSEA) showed enrichment of an IFNα response signature induced in MAIT cells in response to Klebsiella (IRF1, IRF7, IFIT1, and MX1; Fig. 2 B). In line with this, gene ontology (GO) analysis using PANTHER analysis tools demonstrated that Klebsiella induced a significant enrichment for genes related to “type I IFN signaling pathway” as well as “response to virus” (Fig. S2 A). Thus, Klebsiella induces a dominant type I IFN signature in MAIT cells.
To further investigate the effect of type I IFNs in MAIT cell activation, we compared the transcriptional program acquired by MAIT cells in response to Klebsiella with that of cells stimulated with IFNα or influenza virus (Fig. 2, C and D; and Fig. S2, B–E). Venn diagrams highlight the overlapping and unique transcriptional signatures elicited by these three stimulations (Fig. 2 C). Strikingly, from the 401 DEG induced by Klebsiella, more than 40% of genes (168) were also induced after stimulation with IFNα and 124 genes were common for the three stimulations. As expected, most of these common genes are IFN-related genes, and GO enrichment analysis revealed a dominant type I IFN signaling signature (Fig. S2 B). On the other hand, 210 genes appeared exclusively altered by K. pneumoniae, being substantially enriched in genes involved in “cellular response to TNF” including those related to TNF-dependent signaling (CD137 [TNFRSF9], CSF1, and TNFAIP2) and members of the TNF superfamily (CD40L [TNFSF5], GITR [TNFRSF18], and LTA [TNFB]; Fig. 2 D and Fig. S2 E).
We next compared the transcriptional program induced in MAIT cells by Klebsiella, with two publicly available datasets of MAIT cells stimulated with the MAIT cell antigen 5-OP-RU (Hinks et al., 2019; Lamichhane et al., 2019; Fig. 2 E). We detected 170 DEGs that were induced by Klebsiella and also found to be altered in response to 5-OP-RU in one or both datasets. GO analyses of these “common” DEGs showed an enrichment on “cytokine-mediated signaling” pathways, while DEGs only induced by Klebsiella but not 5-OP-RU (231) retained enrichment in IFN signatures. Also, while MAIT cells have been shown to acquire a “tissue-repair” signature in response to TCR stimulation (Hinks et al., 2019; Lamichhane et al., 2019; Leng et al., 2019), we didn’t detect a significant enrichment of tissue repair–related genes in response to Klebsiella (Fig. S2 F).
Given the dominant type I IFN signature induced by Klebsiella, we investigated whether type I IFNs are the main drivers controlling MAIT cell responses during infection. Indeed, in in vitro cocultures using human (Fig. 2 F) or pulmonary murine (Fig. 2 G) cells, blocking type I IFN signaling with the IFNα/β inhibitor B18R (human) or IFNα/β receptor (IFNAR) blocking antibody (αIFNAR, mouse) reduced CD69 upregulation in MAIT cells in response to Klebsiella. This process was independent of MR1 as it was not blocked by αMR1 incubation, while this antibody efficiently blocked 5-OP-RU–mediated CD69 upregulation in murine pulmonary MAIT cells (Fig. S2, G and H). Similar results were obtained in cocultures performed with RAW264.7 macrophages and MAIT cells isolated from mice lacking type I IFN receptor (IFNAR-KO), which show strongly impaired Klebsiella-mediated CD69 upregulation (Fig. 2 G, right), indicating a cell-intrinsic effect for type I IFN signaling in regulating MAIT cell responses. Moreover, type I IFNs also controlled MAIT cell-CD69 upregulation in vivo (Fig. 2 H). Blocking IFNAR in vivo by injection of a blocking antibody to WT mice prior to Klebsiella infection also resulted in significantly lower CD69 levels vs. isotype-treated mice (Fig. 2 H).
Next, we investigated the cell-intrinsic effects of type I IFNs on MAIT cells. Signal transducer and activator of transcription 1 (STAT1) is a critical component of IFN-I signaling, regulating the expression of hundreds of IFN-regulated genes, and it is transcriptionally induced in MAIT cells in response to Klebsiella or IFNα (Fig. 2 D). Thus, we analyzed the phosphorylation of STAT1 in sorted murine pulmonary MAIT cells in response to incubation with IFNα or with the supernatant of Klebsiella-infected RAW264.7 macrophages. Both type I IFN and RAW-supernatant induced a time-dependent phosphorylation of STAT1 in MAIT cells, which was evident already at 15–30 min of stimulation and sustained after 2 h (Fig. 2 I and Fig. S3 A). In line with this, incubation of sort-purified MAIT cells with IFNα was sufficient to induce CD69 upregulation of both murine and human MAIT cells (Fig. 2 J).
Altogether, these data demonstrate that Klebsiella induces a dominant type I IFN signature in MAIT cells, and type I IFNs are key drivers of MAIT cell responses during infection. Nonetheless, we cannot discard that other signals may also contribute to MAIT cell activation during Klebsiella infection. For example, RNAseq analyses detected a signature for “cellular response to TNF” induced by Klebsiella (Fig. 2 B), and TNF has been shown to contribute to MAIT cell activation in response to adenovirus-based vaccines (Provine et al., 2021).
Intrinsic type I IFN signaling controls MAIT cell effector functions
Next, we investigated the MAIT cell effector functions induced by Klebsiella and the mechanisms regulating those functions. A recent single-cell transcriptomic atlas of human peripheral blood MAIT cells revealed their phenotypical and functional heterogeneity covering a broad range of homeostatic, effector, helper, tissue-infiltrating, regulatory, and exhausted phenotypes (Vorkas et al., 2022). Comparison of the transcriptomic signatures of MAIT cells activated with Klebsiella or IFNα revealed that these stimuli induced a significant enrichment in MAIT1-cytotoxic/effector signature in MAIT cells (Fig. 3 A), which included cells with a combined granzyme and Th1-helper associated phenotype (Vorkas et al., 2022).
To investigate the type I IFN–dependent effector functions of MAIT cells, we analyzed the production of effector molecules by culturing sorted murine pulmonary MAIT cells with RAW264.7 macrophages in the presence/absence of Klebsiella (Fig. 3, B and C). In response to Klebsiella, MAIT cells produced IFN-γ, Granzyme B, and to a lesser extent IL-17A (Fig. 3 B). Secretion of these three effectors was independent of MR1 (not affected by αMR1 antibody), yet secretion of Granzyme B and IFN-γ (but not IL-17) were regulated by type I IFNs. Accordingly, Granzyme B and IFN-γ secretion induced by Klebsiella was significantly reduced by an αIFNAR blocking antibody, as well as in cocultures performed with IFNAR-KO MAIT cells (Fig. 3 C). This suggests that intrinsic type I IFN signaling controls MAIT cell effector functions. In line with this, IFNα is sufficient to induce Granzyme B production by sort-purified pulmonary murine MAIT cells (Fig. 3 D). We also measured IFN-dependent secretion of cytokines by pulmonary MAIT cells in vivo after infection with Klebsiella (Fig. 3 E). MAIT cell–Granzyme B secretion was reduced when WT mice were injected with αIFNAR blocking antibody prior to infection as well as in infected IFNAR-KO mice, while IL-17 production was independent of IFNAR (Fig. 3 E). Comparable results were obtained for human MAIT cells when CD8+T cells were cocultured with autologous monocytes, as secretion of effector molecules was blocked by incubation with B18R (Fig. 3 F). Importantly, mutant KPΔribD induced secretion of cytokines (and CD69 upregulation) by murine (Fig. 3 G and Fig. S3 B) and human (Fig. 3 H) MAIT cells at comparable levels with those induced by Kp43816 (KP), further supporting MR1-independent mechanisms as main drivers of MAIT cell effector functions in response to Klebsiella. Collectively, these data indicate that type I IFNs are key regulators of MAIT cells during Klebsiella infection. The type I IFN effects are—at least in part—due to intrinsic IFN signaling on MAIT cells as it directly contributes to controlling their activation and effector functions. Thus, in response to Klebsiella, MAIT cells become potent effectors by sensing inflammation independently of cognate antigen.
Type I IFNs regulate MAIT cell–dependent control of pulmonary Klebsiella
While our experiments confirm a key role for type I IFNs in regulating MAIT cell functions, the relevance of MAIT cell–intrinsic IFNAR for host protection remains unknown. To address whether IFNAR expression by MAIT cells confers protection against Klebsiella, we combined a variety of approaches to restrict IFNAR expression/depletion to MAIT cells in vivo (Fig. 4, A–C). First, we adapted a previously published approach (van Wilgenburg et al., 2018; Wang et al., 2018) by adoptively transferring pulmonary WT or IFNAR-KO MAIT cells into TCRα-KO mice (Fig. 4 A). Residual contaminating conventional T cells were depleted in recipients by injection of αCD4 and αCD8 antibodies. Mice were rested for 2 wk, to allow for MAIT cell expansion, and subsequently infected with Klebsiella. Strikingly, adoptively transferred MAIT cells contributed to the control of Klebsiella infection, as we recovered significantly lower CFUs from the lungs of mice receiving WT MAIT cells vs. control (untreated) animals (Fig. 4 A). However, the protective effect of MAIT cells was abrogated when the transferred cells were deficient in IFNAR, indicating that the protective role of MAIT cells is dependent on intrinsic IFNAR signaling (Fig. 4 A). As type I IFN signaling-deficient mice are susceptible to K. pneumoniae infection (Ivin et al., 2017), we asked whether MAIT cells are sufficient to control bacterial infection in an IFNAR-deficient context. To investigate this, we adoptively transferred WT pulmonary MAIT cells into IFNAR-KO recipients prior to infection with Klebsiella (Fig. 4, B and C). As previously described, IFNAR-KO mice showed increased bacterial loads in the lungs after infection in comparison with WT controls (Fig. 4 C). Strikingly, transfer of WT MAIT cells into IFNAR-KO recipients was sufficient to reduce bacterial loads (Fig. 4 C), confirming their protective role during Klebsiella infection. On the other hand, transfer of MAIT cells into MR1-KO recipients also conferred protection as evidenced by reduced body-weight loss and bacterial burden after infection, suggesting an MR1-independent mechanism driving MAIT cell protection in this setting (Fig. 4 D). Thus, these data indicate that type I IFN–mediated activation of MAIT cells drives their protective role during K. pneumoniae infection.
While MAIT cells are abundant in humans, in specific pathogen–free C57BL/6 mice, baseline frequencies of MAIT cells are very low. However, pulmonary MAIT cells can be expanded in mice via vaccination with 5-OP-RU in the presence of TLR agonists (Chen et al., 2016b; Hinks et al., 2019). Thus, we investigated whether MAIT cells contribute to the control of pulmonary Klebsiella infection and the relevance of type I IFNs in this system. To test this, we i.n. administered WT mice with 5-OP-RU+LPS and infected mice with K. pneumoniae a week later. As controls, we injected mice with either vehicle (PBS) or LPS (Fig. 4 E and Fig. S3 C). Administration of 5-OP-RU+LPS induced a strong increase in the total number of MAIT cells accumulated in the lung, which was not observed in mice receiving LPS alone or vehicle control (Fig. S3 C). MAIT cells from 5-OP-RU+LPS-treated mice showed comparable levels of cytokines, transcription factor expression, and surface markers, but increased levels of CD69 in comparison to MAIT cells from control (PBS treated) mice (Fig. S3 D). Strikingly, MAIT cell expansion resulted in protection from Klebsiella infection as evidenced by prevention of weight loss, decrease in bacterial CFUs, and decreased lung inflammation in 5-OP-RU+LPS-treated mice (Fig. 4, E and F). This protection was not observed in mice treated exclusively with LPS, which showed comparable numbers of MAIT cells, weight loss, bacterial CFUs, and lung pathology as control (untreated) mice (Fig. 4, E and F). To define the mechanisms mediating the MAIT cell–dependent control of Klebsiella, we injected mice with blocking antibodies prior to infection (Fig. 4 G). We found that the MAIT cell protective effect was independent of MR1 but dependent on type I IFN (Fig. 4 G). Accordingly, injection (prior to infection) of αIFNAR blocking antibody or αSiglecH (which reduces type I IFN secretion by plasmacytoid dendritic cells; Blasius et al., 2006) resulted in increased bacterial loads recovered from the lungs of infected mice (Fig. 4 G). Furthermore, the protective MAIT cell function was also lost when we expanded MAIT cells (5-OP-RU+LPS as above) in IFNAR-KO mice as we recovered increased bacterial counts in the lungs of these animals vs. WT controls (Fig. 4 H). Thus, expansion of MAIT cells is sufficient to confer protection from Klebsiella infection through a mechanism dependent on type I IFN but independent of MR1.
Our data are in agreement with previous reports in which MAIT cell expansion with 5-OP-RU in combination with TLR ligands or IL-23 provides protection from bacterial infection (Wang et al., 2018, 2019). Similarly, adoptive transfer of pulmonary MAIT cells into immunodeficient mice, protects from infection with influenza or L. longbeachae (van Wilgenburg et al., 2018; Wang et al., 2018). Interestingly, whereas protection by transferred MAIT cells is dependent on MR1 during Legionella infection (Wang et al., 2018), in the case of Klebsiella and influenza (van Wilgenburg et al., 2018), MAIT cells are able to exert antimicrobial functions in the absence of MR1. The mechanisms by which MAIT cells promote protection from Klebsiella will require further investigation, but it is possible that they contribute to bacterial clearance either by direct secretion of cytolytic molecules or by indirectly controlling the activation and recruitment of other immune cells. For instance, in response to E. coli, release of granzymes and granulysin by human MAIT cells mediates control of both cell-associated and extracellular forms of the bacteria (Boulouis et al., 2020). In this setting, granulysin contributes to direct killing of extracellular bacteria, while Granzyme B reduces cell-associated bacteria. Although murine cells do not secrete granulysin, K. pneumoniae can survive within macrophages (Cano et al., 2015), thus MAIT cell–derived granzyme could contribute to kill cell-associated bacteria. Added to this, MAIT cell–derived IFN-γ has been shown to be protective during influenza or L. longbeachae infections (van Wilgenburg et al., 2018; Wang et al., 2018), and IFN-γ deficient mice are highly susceptible to Klebsiella infection (Moore et al., 2002), suggesting further possible mechanisms for anti-bacterial MAIT cell functions.
Type I IFNs modulate tissue location of MAIT cells during pulmonary Klebsiella infection
Given that MAIT cells limit bacterial expansion in the lungs and preserve lung pathology during infection, we investigated the effect of bacterial infection in MAIT cell spatial location in the lung (Fig. 5). First, we analyzed the location of MAIT cells in the lung by intravenously injecting fluorescently labeled anti-CD45 antibody (CD45-FITC) 3 min before tissue collection. This approach labels CD45+ cells located in circulation and leaves cells in the interstitial tissue and alveoli unlabeled (Salou et al., 2019). Analyses of pulmonary MAIT cells at steady-state demonstrated that around 30% of MAIT cells are located in the lung parenchyma and inaccessible to CD45 labeling. However, the proportion of unlabeled cells significantly increased to around 70% at 48 h after infection, suggesting that cells are relocating into the tissue (Fig. 5 A). Moreover, unlabeled (parenchymal, anti-CD45−) MAIT cells showed an increase in expression of the activation marker CD69, in comparison with circulating MAIT cells (intravascular, anti-CD45+; Fig. 5 B), suggesting local activation within the tissue.
Next, we investigated how Klebsiella infection influenced the location of pulmonary MAIT cells in situ. To do this, we took advantage of CXCR6-GFP mice as MAIT cells (as well as other unconventional T cells) express CXCR6 in the lung and can be identified in the tissues on the basis of GFP expression (Fig. S3, E and F). MAIT cells represent around 5% of the GFP+CD3+ population in CXCR6-GFP lungs, but after MAIT cell expansion (5-OP-RU+LPS as above) MAIT cells constitute ∼80% of GFP+CD3+ cells (Fig. S3 F). Thus, we expanded MAIT cells in the lungs of CXCR6-GFP mice and analyzed their tissue distribution by FLASH (fast light-microscopic analysis of antibody-stained whole organs; Messal et al., 2021). This approach enabled us to perform high-resolution immunofluorescence in 3D sections of lungs at steady-state as well as after Klebsiella infection. In the absence of infection, MAIT cells (GFP+CD3+, white arrows) appear widely dispersed along the alveolar parenchyma (in blue, podoplanin+ cells). However, 48 h after Klebsiella infection, MAIT cells relocate and accumulate in the peri-bronchial spaces (Fig. 5 C). Cell crowding around this region was quantified by calculating the average distance (μm) between all cells appearing in an image (Zepp et al., 2017). Distance between cells was significantly reduced after Klebsiella infection, confirming the targeted recruitment of MAIT cells to peri-bronchial spaces induced by Klebsiella (Fig. 5 C). Since type I IFNs control the expression of chemokines involved in lung homing of T cells (Padovan et al., 2002; Voillet et al., 2018), we hypothesized that MAIT cell relocation could also depend on type I IFN signaling. Indeed, MAIT cell clustering was prevented by injection of mice with αIFNAR blocking antibody prior to infection (Fig. 5 C). Thus, type I IFNs modulate tissue location of pulmonary MAIT cells during infection.
Concluding remarks
Our data identify type I IFNs as key drivers of MAIT cell functions during pulmonary infection with K. pneumoniae. Type I IFNs act on MAIT cells controlling their activation, effector functions, and induction of a Th1/cytotoxic transcriptional program, ultimately contributing to host protection. Thus, during Klebsiella infection, MAIT cells become potent effectors by sensing inflammation independently of cognate antigen. This behavior strongly resembles that of memory CD8+ T cells, whose effector functions (Kohlmeier et al., 2010; Soudja et al., 2012) and trafficking (Sung et al., 2012) can be regulated by type I IFNs in an antigen-independent manner. The cytokine-mediated activation of MAIT cells in response to Klebsiella contrasts with the MR1-dependent activation during infection with other bacteria such as F. tularensis or L. longbeachae (Meierovics et al., 2013; Wang et al., 2018, 2019). The distinct timeframes of the infections (2 d vs. 2 wk) and the different activation/recruitment of immune cells associated with the specific pathogens may contribute to these differences. Accordingly, it is likely that the timing and (local) concentration of type I IFN delivery accompanying different infections will critically regulate subsequent MAIT cell responses. In keeping with this, IFN-β treatment of human naïve T cells induces several temporal transcriptional waves that regulate the dynamics of T cell activation and differentiation (Sumida et al., 2022). Given the abundance of MAIT cells in humans and their rapid response to inflammatory signals, we propose that type I IFNs may serve as a new molecular target to manipulate MAIT cell functions during infections.
Materials and methods
Mice, cells, and chemicals
Mice (WT, IFNAR1-KO, TCRα-KO, MR1-KO on C57BL/6J background; and CXCR6-EGFP “knock-in” in FVB/N background—the latter kindly provided by Adrian Hayday at the Francis Crick Institute, London, UK) were bred under specific pathogen–free conditions at the Francis Crick Institute animal facility. Experiments were carried out using 7–10-wk-old mice with age and gender-matched between genotypes. All experiments have been approved by the Francis Crick Institute and the King’s College London’s Animal Welfare and Ethical Review Body, and UK Home Office under the project license P0BD71419, performed in accordance with the Animals (Scientific Procedures) Act 1986.
Healthy human PBMCs were isolated from leukocyte cones purchased from the NHS Blood and Transplant (UK) and obtained from fully anonymized donors with informed consent. PBMCs were flushed from the leukocyte cones using a syringe and isolated using density gradient centrifugation.
RAW264.7 cells (ATCC TIB-71) were cultured in DMEM (Gibco; 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 14.3 μM β-mercaptoethanol) at 37°C in 5% CO2.
5-OP-RU was freshly prepared by mixing 5-A-RU and methylglyoxal (Sigma-Aldrich) as previously described (Li et al., 2018).
Bacterial strains and growth conditions
K. pneumoniae (43816 strain, O1K2 serotype) cultures were grown in Luria-Bertani (LB) broth from glycerol stocks at 37°C overnight with shaking at 180 rpm. For the preparation of the inoculum for in vivo infections, bacteria were reinoculated in prewarmed LB medium for 1–2 h culture until reaching OD600 = 0.6 (with the estimation that 1 OD600 = 7 × 108 CFUs/ml). The bacterial pellet was washed and diluted in phosphate-buffered saline (PBS) for i.n. delivery to mice under isoflurane-induced anesthesia. Bacterial concentration in the inoculum was verified by counting CFUs grown on LB Agar plates.
To construct a KlebsiellaribD mutant (KPΔribD), we used λ-red-mediated homologous recombination, as described previously (Hancock et al., 2020). Homologous regions upstream and downstream of ribD were amplified via PCR using primers Kp43-ribD.2 (5′-GCATTGCCCTTTCTGTTTC-3′) and Kp43-ribD.3 (5′-GGAATAGGAACTAAGGAGGAAGGATGGGTAGTAAAGCG-3′) as well as Kp43-ribD.4 (5′-CCTACACAATCGCTCAAGACTGCCTGCATTTAACCACTG-3′) and Kp43-ribD.5 (5′-GCTTCTGCACCTTTGTTACC-3′), respectively, and they were sewn together with the kanamycin cassette amplified from pKD4 (Datsenko and Wanner, 2000) using primers cm.3a (5′-TCCTCCTTAGTTCCTATTCC-3′) and cm.4a (5′-GTCTTGAGCGATTGTGTAGG-3′). The PCR products were amplified with primers Kp43-ribD.2 and Kp43-ribD.5, and the PCR fragment was purified with a Qiagen MinElute Kit. Strain Kp43 was transformed with pKOBEG-gent, which encodes the λ phage redγβα operon (Balestrino et al., 2005) and was expressed under the control of the arabinose-inducible pBAD promoter. Briefly, Kp43816 was grown overnight and subcultured 1:100 in 200 ml of LB supplemented with gentamicin and 0.2% arabinose. Cells were grown at 37°C until reaching an OD600 of 0.4, cooled on ice, pelleted, resuspended in ice-cold 10% glycerol, and incubated on ice for 1 h. The cells were electroporated with 1 μg three-way-PCR DNA in a 2-mm cuvette with 2.5 kV, 200 Ω, and 25 μF. Transformed cells were selected via plating on LB agar supplemented with kanamycin and riboflavin at 30°C. The replacement of the WT allele by the mutant was confirmed via PCR using primers Kp43-ribD.1 (5′-TGCTTAATGGTAGCCAAACC-3′) and Kp43-ribD.6 (5′-CAGGAACTGGTATTCGACATC-3′). The pKOBEG-gent plasmid was cured from the mutant strain via overnight growth at 37°C. ribD is an essential gene in K. pneumoniae and therefore the mutant needs to be grown in the presence of riboflavin (10 ng/ml).
In vivo inoculations and bacterial infections
For infection experiments, mice were i.n. challenged with 5 × 104 CFUs of K. pneumoniae in 30 μl of sterile PBS. For in vivo antibody-mediated blocking experiments, mice were intraperitoneally (i.p.) injected with the following antibodies 24–48 h before infection: αMR1 (150 μg/mouse; clone 26.5; Biolegend, Mouse IgG2a), αIFNAR1 (200 μg/mouse; clone MAR1-5A3; BioXCell, Mouse IgG1), αSiglecH (150 μg/mouse; clone 440c; BioXCell, Rat IgG2b), or their respective isotype controls: Mouse IgG2a (150 μg/mouse; clone MOPC-173; Biolegend), Mouse IgG1 (200 μg/mouse; clone MOPC-21; BioXCell), and Rat IgG2b (150 μg/mouse; clone LTF-2; BioXCell). When indicated, MAIT cell expansion in vivo was performed through a repeated i.n. inoculation (3×) of 5-OP-RU (100 μM) and LPS (17.4 μg/mouse; Invivogen).
Bacterial loads were determined by counting CFUs after plating 100-fold dilution series of tissue homogenates obtained from bacteria-infected mice. Colonies were counted at 24 h.
Tissue processing, flow cytometry, and MAIT cell sorting
Mice were euthanized by cervical dislocation, and spleen, inguinal lymph nodes, liver, and lungs were harvested. Lungs, inguinal lymph nodes, and liver were finely chopped and digested with 1.5 mg/ml collagenase D (Sigma-Aldrich), 0.1 mg/ml DNase I (ITW Reagents), 0.6 mg/ml NADase (Sigma-Aldrich), and 5% FBS (Thermo Fisher Scientific) in PBS for 60 min at 37°C with gentle shaking. Cells were then passed through a cell strainer (40–70 μm) and washed with sterile PBS. For flow cytometric analysis, cell suspensions were depleted from red blood cells using the hypotonic buffer Tris-based amino-chloride (5 min, room temperature) prior to staining with the antibodies depicted below.
Flow cytometry staining was performed in FACS buffer (1% FBS, 1% BSA, 0.02% sodium azide) using the following antibodies from Biolegend unless specified otherwise: Anti-mouse antibodies: B220 (RA3-6B2), CD11b (M1/70), CD11c (N418), CD27 (LG.3A10), CD28 (37.51), CD69 (H1.2F3), CD25 (PC61), CD44 (IM7), CD137 (17B5), ICOS (7E.17G9; Invitrogen), PD-1 (29F.1A12), TCRβ (H57-597), TCR γ/δ (clone GL3), IL-17A (TC11-18H10.1), IFNγ (XMG1.2), Granzyme B (QA16A02), and pSTAT1(Ser727; A15158B). Anti-human antibodies were as follows: CD14 (HCD14), CD19 (HIB19), CD3ε (HIT3a), CD8α (HIT8a), CD161 (HP-3G10), Vα7.2 (3C10), CD69 (FN50), and Granzyme B (QA16A02). Dead cells were excluded from the analyses using Zombie fixable viability dye (Biolegend). For intracellular staining, cells were fixed and permeabilized using BD Cytofix/Cytoperm Solution kit (BD Biosciences) according to the manufacturer’s instructions. MR1(5-OP-RU), MR1(6-formyl pterin, Ac-6-FP), and CD1d(PBS-57)-loaded tetramers were provided by the National Institutes of Health (NIH) Tetramer Facility. The MR1 tetramer technology was developed jointly by James McCluskey, Jamie Rossjohn, and David Fairlie, and the material was produced by the NIH Tetramer Core Facility as permitted to be distributed by the University of Melbourne. Data were recorded using LSR-II or LSR Fortessa cytometers (BD Biosciences) and analyzed with FlowJo software (TreeStar).
For cell sorting of pulmonary murine MAIT cells, T cells were previously enriched in lung single-cell suspensions using αCD3ε microbeads (Miltenyi Biotec). Murine MAIT cells were sorted as Zombie−B220−CD11b−CD11c−TCRβ+MR1(5-OP-RU) tetramer+ cells.
Human MAIT cells were sorted from PBMCs as Zombie−CD14−CD19−CD3ε+CD8α+CD161+Vα7.2+ cells.
Adoptive transfer
For adoptive transfer experiments, pulmonary MAIT cells (WT or IFNAR-KO) were isolated from the lungs of mice after expansion with 5-OP-RU/LPS as described above. 2–3 × 105 MAIT cells were injected into the tail veins of TCRα-KO recipient mice. To control residual contamination of conventional T cells, mice were i.p. injected with 100 μg each of αCD4 (clone GK1.5; BioXCell, Rat IgG2b) and αCD8 (clone 53-6.7; BioXCell, Rat IgG2a) blocking antibodies on days 2 and 6. To allow full expansion of the MAIT population in the lungs, mice were rested for 14 d prior to infection with K. pneumoniae.
For adoptive transfer into IFNAR-KO or MR1-KO mice, 2–4 × 105 MAIT cells were injected into the tail veins of recipient mice. Mice were infected with Klebsiella 7 d after transfer.
Lung histopathological evaluation
Harvested lungs were fixed in 10% neutral buffer formalin before routine histological processing in paraffin wax. 3-μm-thick sections were cut and H&E staining was performed following standard protocols. H&E-stained slides were evaluated concurrently by two board-certified veterinary pathologists (A. Suárez-Bonnet and S.L. Priestnall) blinded to the experimental groups using an BX43 microscope (Olympus). Representative slides were scanned using Panoramic Scan II slide scanner (3D Histech). Presence of airway inflammation (cell infiltration of the airways, bacterial presence within inflamed regions) was examined in the sections. Additional tissue sections were immunohistochemically stained using a rabbit monoclonal antibody anti-F4/80 (clone EPR26545-166; Abcam) to highlight macrophages and semiquantitatively evaluated. Standard histopathological criteria were used for the assignment of the lung injury scores as follows: score 0 (no lesion, 1–4%), score 3 (minimal, 5–9%), score 6 (mild, 10–19%), score 9 (moderate, 20–50%), and score 12 (marked, >51%).
In vitro stimulation and coculture experiments
For in vitro stimulation experiments, MAIT cells were sorted from the lungs of mice (previously injected with 5-OP-RU+LPS to expand the MAIT cell population). Cells were incubated with IFNα (1 μg/ml; Biolegend) or the supernatant of RAW264.7 macrophages previously exposed to Klebsiella (for 2 h).
For coculture experiments with murine cells, T cells from spleen and inguinal lymph nodes (negatively selected using biotin anti-mouse B220 antibody [RA3-6B2] and Dynabeads Biotin Binder [Invitrogen]) were cocultured with RAW264.7 macrophages (1:2 cell ratio) in the presence of live K. pneumoniae (multiplicity of infection [MOI] 100) for 3 h. Alternatively, sorted pulmonary MAIT cells (as above) were cocultured with RAW264.7 macrophages (1:4 cell ratio) in the presence of fixed K. pneumoniae (100 bacteria per cell) for 18 h. For detection of intracellular cytokines, brefeldin A (Sigma-Aldrich) was added to the final 4 h of the coculture. In some experiments, the following blocking antibodies were added to the cultures: αMR1 (26.5; Biolegend, Mouse IgG2a), αIFNAR1 (MAR1-5A3; BioXCell, Mouse IgG1), αIL-12/23p40 (C17.8; Biolegend, Rat IgG2a), αIL-18 (YIGIF74-1G7; BioXCell, Rat IgG2a); or respective isotype control antibodies: Mouse IgG2a (clone MOPC-173; Biolegend), Mouse IgG1 (clone MOPC-21; BioXCell), Rat IgG2a (clone RTK2758; Biolegend), and Rat IgG2a (clone 2A3; BioXCell).
For in vitro experiments with human cells, CD8+ T cells and CD14+ monocytes were isolated from PBMCs using αCD8α and αCD14 microbeads (Miltenyi Biotec). Cells were cocultured in the presence of fixed K. pneumoniae (100 bacteria per cell) in a 2:1 cell ratio (CD14+:CD8+ cells) in complete RPMI medium (Gibco; 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 14.3 μM β-mercaptoethanol). For detection of intracellular cytokines, brefeldin A (Sigma-Aldrich) was added to the final 4 h of the coculture. In some experiments, αMR1 (1 μg/ml, clone 26.5; Biolegend, Mouse IgG2a) or vaccinia virus B18R protein (5 μg/ml; eBioscience) were added to the cultures.
RNAseq and data analysis
Human PBMCs (from four different healthy donors) were incubated during 20 h with fixed K. pneumoniae (100 bacteria per cell) or, alternatively, with IFNα2A (50 ng/ml; Sigma-Aldrich) or influenza A virus (H3N2, X31 strain, MOI 1). MAIT cells were sorted as Zombie−CD14−CD19−CD3ε+CD8α+CD161+Vα7.2+ cells, and RNA was extracted with NucleoSpin RNA XS kit (Macherey-Nagel) following the manufacturer’s instructions. Libraries were generated using NEBNext Ultra II Directional PolyA mRNA kit (New England Biolabs), barcoded, and run on an Illumina NovaSeq 6000 system generating 25 million single-end 75 bp reads per sample. Gene expression was quantified from raw FastQ files in the GRCh38 genome with Ensembl release-95 gene models using the nf-core/rnaseq pipeline (version 3.3; Ewels et al., 2020). Reads were trimmed for adapters with trimgalore (version 0.6.6) and aligned and quantified with RSEM/STAR (version 1.3.1). Data quality was inspected using FastQC (version 0.11.9), Picard (version 2.23.9), and RSeQC (version 3.0.1). Quantified gene expression tables were loaded into R (version 4.1.1) using the tximport package (version 1.20.0; Soneson et al., 2015). DEGs resulting from each treatment, while accounting for between-donor variation, were identified using the DESeq2 package (version 1.32.0; Love et al., 2014). An adjusted P value threshold of 0.01 and log2 fold change >0.5 were applied to provide lists of DEGs. Sequencing data have been deposited in the National Center for Biotechnology Information's Gene Expression Omnibus repository under the accession number GSE218601.
Microscopy
MAIT cells were expanded (with 5-OP-RU/LPS as described above) in CXCR6-EGFP mice which were subsequently infected with Klebsiella. Lungs were harvested and fixed overnight in 4% paraformaldehyde (Sigma-Aldrich). Once fixed, antigen retrieval, sample preparation, and staining were performed following the FLASH immunofluorescence protocol (Messal et al., 2021). Samples were stained with rabbit polyclonal αCD3ε antibody (Abcam), αPodoplanin/gp36 (PMab-1; Abcam), and goat polyclonal αGFP (Abcam) antibodies followed by donkey anti-rabbit IgG Alexa Fluor Plus 594 (Invitrogen), donkey anti-rat IgG DyLight 680 (Invitrogen), and donkey anti-goat IgG Alexa Fluor 488 (Invitrogen). Images were acquired by using an inverted Zeiss LSM 710 microscope. Cell clustering was calculated (Zepp et al., 2017) by measuring the average distance (μm) between CD3+GFP+ cells on at least 9–10 different images per condition by using Fiji/Image J software.
Statistical analysis and figure preparation
Statistical tests were performed using Prism GraphPad software (version 9.4.0). Comparison between groups was performed using Student’s t tests, one-way, two-way ANOVA tests, or Mann-Whitney tests as appropriate unless otherwise stated.
Some figures were drawn using pictures from Servier Medical Art. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/).
Online supplementary material
Fig. S1 shows data related to Fig. 1, including MAIT cell gating strategy and additional controls for αMR1 in vivo. Fig. S2 shows additional RNAseq analyses and data related to Fig. 2, and Fig. S3 shows data related to Fig. 2 (pSTAT1 time-course), Fig. 3 (quantification of MAIT cell responses to WT Klebsiella and KPΔribD), Fig. 4 (phenotyping of MAIT cells after expansion in vivo with LPS+5-OP-RU), and Fig. 5 (characterization of MAIT cells in CXCR6-GFP mice). Tables S1, S2, and S3 include DEGs obtained from RNAseq for human MAIT cells in response to Klebsiella, IFNα, or influenza (vs. control).
Data availability
Sequencing data have been deposited in the National Center for Biotechnology Information's Gene Expression Omnibus repository under the accession number GSE218601. All other data are available in the article itself and its supplementary materials.
Acknowledgments
We thank the NIH Tetramer Facility for provision of MR1 and CD1d tetramers.
This work was funded by the UK Biotechnology and Biological Sciences Research Council (grant to P. Barral BB/S005560/1 and grant to J.A. Bengoechea BB/T001976/1). The authors acknowledge technical support from the Biological Research Facility, Advanced Sequencing, Histopathology, Flow Cytometry, and Light Microscopy Platforms from The Francis Crick Institute (which receives its core funding from Cancer Research UK, the UK Medical Research Council, and the Wellcome Trust).
Author contributions: P. Barral conceptualized and supervised the study; J.C. López-Rodríguez conducted most of the experiments; S.J. Hancock constructed the ribD mutant and established growing conditions; P. Klenerman, J.A. Bengoechea, and A. Wack contributed to experimental interpretation and discussion; A. Suárez-Bonnet and S.L. Priestnall performed pathology analyses; C. Barrington performed RNAseq analyses; J. Aubé and K. Li synthesized and provided 5-A-RU; S. Crotta and A. Wack provided influenza stocks; P. Barral and J.C. López-Rodríguez wrote the paper, which was revised and edited by all authors.
References
Author notes
Disclosures: P. Klenerman reported having acted as a consultant/advisory board member for UCB, AZ, Infinitopes, and Biomunex. No other disclosures were reported.