MAIT cell activation and recruitment in inflammation and tissue damage in acute appendicitis

Clinical characteristics of pediatric patients with acute appendicitis and control participants

The pediatric cohorts comprised 48 patients with acute appendicitis and 37 control participants. There was no significant difference in age and sex ratios between the two groups (table S1). The levels of white blood cells, neutrophils, and monocytes were all significantly increased, while the lymphocyte count was significantly decreased in patients with APA (table S1 and fig. S1A). Gangrenous and perforated appendicitis accounted for most acute appendicitis in this study, followed by acute purulent appendicitis, simple acute appendicitis, and peri-appendicular abscess, respectively (fig. S1B). The vast majority of the patients received an appendectomy within 96 hours of the onset of symptoms (fig. S1C). The most common culturable bacteria from the appendix samples in this cohort were the gram-negative bacteria (GNB) family Enterobacteriaceae, particularly Escherichia coli (55%), followed by Klebsiella pneumoniae, and other Enterobacteriaceae. Pseudomonas aeruginosa and other GNB were less commonly cultured, with gram-positive bacteria being the least common (fig. S1D). In almost a quarter of cases, no bacteria were culturable (fig. S1D), and most of such patients appeared to suffer the milder form of APA, with very few gangrenous or perforated appendixes (fig. S1E).

MAIT cells decline in the circulation of patients with acute appendicitis

In light of the decrease of PB lymphocyte counts during acute appendicitis, further investigation was undertaken to determine which lymphocyte populations were affected (fig. S2A). Strikingly, T cell counts were strongly decreased in patients with APA compared with the control group (fig. S2B). In contrast, CD19⁺ B cell levels were unchanged in children with acute appendicitis (fig. S2C). The Uniform Manifold Approximation and Projection dimensionality reduction analysis indicated that the decrease in T cell populations was broad, with a profound decrease of the antimicrobial innate-like MAIT cells and γδT cells (Fig. 1A). Deeper analyses indicated that the numbers of MAIT cells, γδT cells, CD8⁺ T cells, and CD4⁻ CD8⁻ double-negative (DN) T cells were all significantly decreased (Fig. 1, B to E), whereas CD4⁺ T cell numbers were not affected (fig. S2D). However, within the T cell population, only the proportion of MAIT cells and γδT cells decreased in patients with acute appendicitis compared to control participants, whereas the proportion of other T cell subpopulations was unaffected (Fig. 1, F and G, and fig. S2, E to G).

MAIT cells and γδT cells belong to the innate-like antimicrobial T cell population and are poised to mediate rapid and robust responses against bacterial infections (16–18, 29, 53, 54). Given that the most common cause of acute appendicitis is bacterial infection leading to severely inflamed appendix and the consistent decrease in the numbers and frequency of MAIT cells and γδT cells, further investigations were performed to understand their potential roles in acute appendicitis. Circulating MAIT cell levels were lower in pediatric patients with the more severe appendicitis (Fig. 1H). In addition, the numbers of PB MAIT cells correlated negatively with neutrophil levels, as well as plasma levels of the inflammatory markers C-reactive protein (CRP) and D-dimer (Fig. 1, I to K). These associations were not observed for γδT cells and other T cell populations, except for CD4⁺ T cells (fig. S2, H to K, and table S2). The numbers of circulating MAIT cells significantly increased at 5 to 7 days after appendectomy compared to those just before the surgery (Fig. 1L), coinciding with the decline of CRP levels and pan-leukocyte, neutrophil, and monocyte counts (table S3). Of note, total lymphocyte counts had not yet significantly recovered at this early time point after appendectomy (table S3). These findings indicate that the circulating levels of MAIT cells decrease during APA and correlate with disease severity. This suggests a potential role of MAIT cells in the pathogenesis of appendicitis.

MAIT cells are enriched in the inflamed appendix and display altered expression of chemokine and homing receptors

As circulating MAIT cells during acute appendicitis negatively correlated with disease severity, MAIT cells may have redistributed to the appendix. To assess this possibility, MAIT cell levels were determined in the more inflamed portions of the appendix compared to the less or noninflamed parts. Appendiceal tissue inflammation severity was assessed by gross and pathological examinations and the influx of CD14⁺ monocytes and CD66b⁺ neutrophils (fig. S3, A to D), as well as by the heat flow produced by the equivalent number of single cells released from each tissue part using an isothermal microcalorimetry methodology (fig. S3E). The number of MAIT cells, identified as MR1-5-OP-RU–reactive TCR Vα7.2⁺ T cells (fig. S3F), were significantly elevated in more inflamed parts of the appendix tissue compared with the less or noninflamed parts (Fig. 2, A and B). In contrast, total T cells, CD4⁺ T cells, CD8⁺ T cells, DN T cells, and γδ T cells were not significantly different between the two parts of the appendix (fig. S3, G to K). Furthermore, the proportion of MAIT cells was substantially higher in the more inflamed part of the appendix than in blood (Fig. 2C). Conversely, the proportions of total T cells, CD8⁺ T cells, DN T cells, and γδ T cells were significantly lower in both inflamed and less/noninflamed parts of the appendix compared with those of PB (fig. S3, L to O). Although the proportions of CD4⁺ T cells in both inflamed and less/noninflamed parts of the appendix were higher than those in the circulation, there was no difference between these appendix parts (fig. S3P). Collectively, the MAIT cell population in blood was dramatically reduced during APA but enriched in the inflamed parts of the appendix. No such pattern was observed in other T cell populations, suggesting that MAIT cells may participate in the pathogenesis of acute appendicitis in children.

Elevated MAIT cell numbers in the inflamed portions of the appendix may reflect enhanced local tissue-resident MAIT cell proliferation from stimulation by riboflavin-autotrophic bacteria infecting the appendix (fig. S1D). However, there was a lack of Ki-67 expression by MAIT cells in blood, as well as both inflamed and noninflamed portions of the appendix (fig. S3, Q and R), suggesting that the increase in appendiceal MAIT cell numbers in the inflamed parts was less likely due to the proliferation of local tissue-resident MAIT cells. To address the possibility that this accumulation was due to PB MAIT cells homing to the inflamed appendix, expression of tissue-homing receptors was examined in the PB MAIT cells of APA and control groups. There was a significant reduction in MAIT cells expressing the chemokine receptors CCR2 and CXCR6, while there was no difference in the proportion of MAIT cells expressing CCR5 and CCR6 (Fig. 2, D and E). Unexpectedly, MAIT cells from children with acute appendicitis expressed significantly more CCR1 and CCR4 than those of control children (Fig. 2, D and E). This was intriguing as CCR1 and CCR4 are not commonly expressed by MAIT cells under a steady state in healthy humans (55). Although there was a significant difference in CCR3-expressing MAIT cells between the cohorts, the absolute levels of expression were very low (<5%) (Fig. 2, D and E). Of note, MAIT cell expression of the gut-homing integrin α4β7 was decreased in the blood of children with acute appendicitis, whereas expression of the tissue residency marker CD103 was elevated (Fig. 2, F and G). The circulating MAIT cells expressing the integrin αEβ7, the receptor for gut epithelial cells adhesion molecule E-Cadherin, were not significantly affected (Fig. 2, F and G). These findings indicate that circulating MAIT cells during APA have an altered chemokine and tissue-homing receptor expression pattern, potentially due to the activation and redistribution of specific MAIT cell subsets into the inflamed appendix.

Subsequent experiments were carried out to explore whether the alteration in chemokine and tissue-homing receptor expression patterns observed in circulating MAIT cells was reflected in their appendiceal counterparts. MAIT cells from the more inflamed appendix expressed more CCR1, CCR4, and CXCR6 than those from the less/noninflamed appendix (Fig. 2, H and I). Compared to those from circulation, MAIT cells from the less/noninflamed part of the appendix expressed significantly more CCR4 (Fig. 2, H and I). In contrast, MAIT cells from the more inflamed part of the appendix expressed more CCR4, CCR6, and CXCR6 (Fig. 2, H and I). CCR2 expression was significantly lower in appendiceal MAIT cells from either part when compared to those in circulation (Fig. 2, H and I). No significant differences were observed in CCR3 and CCR5 expression levels across PB and appendiceal tissues (Fig. 2, H and I).

Expression of integrin β7, CD69, and CD103 in MAIT cells was higher in less/noninflamed parts of appendiceal tissues than in PB (Fig. 2, J and K). However, expression of these gut homing and tissue residency markers were significantly lower in the more inflamed part than in the less/noninflamed part and, in most cases, were comparable to those of circulating MAIT cells (Fig. 2, J and K). These observations suggest a dilution effect of tissue-resident appendiceal MAIT cells by circulating MAIT cells that may have trafficked to the more inflamed parts of the appendix. Together, the altered expression patterns of chemokine receptors, gut-homing integrins, and tissue residency markers in appendiceal MAIT cells support the notion that circulating MAIT cells may have trafficked to the inflamed appendix during APA.

Elevated proinflammatory cytokines in the plasma and appendix tissue are associated with chemokine receptor expression alterations in MAIT cells

Acute bacterial appendicitis often leads to elevated levels of various proinflammatory cytokines (56). MAIT cells are highly responsive to multiple proinflammatory cytokines (25, 26, 57–59). Therefore, we measured the levels of various proinflammatory cytokines in the plasma and appendiceal tissues in patients with APA, with particular attention to cytokines that have substantial effects on MAIT cells, such as interleukin-1β (IL-1β), IL-7, IL-12, IL-18, and IL-23. Several of these MAIT cell–activating cytokines were elevated in the plasma, as well as in the inflamed portions of appendix tissues, including IL-1β, IL-18, and IL-23 (Fig. 3, A to H). Short-term in vitro culture of MAIT cells from healthy donors with IL-18 in the presence of IL-12, a combination commonly used to activate MAIT cells, or with IL-1β and IL-23, led to the alteration of chemokine receptor expression similar to those observed in appendicitis (Fig. 2, D, E, H, and I), including up-regulation of CCR1, CCR3, CCR4, and CXCR1; down-regulation of CCR2; and no apparent effect on CCR5 (Fig. 3, I to L). The chemokine receptor alterations were also seen and enhanced during longer-term in vitro culture of MAIT cells from healthy donors with these cytokines in the presence of the MR1 ligand and MAIT cell antigen 5-OP-RU (Fig. 3, I to L), which is produced by most GNB commonly found infecting the inflamed appendix (fig. S1D). The up-regulation of CCR4 was particularly notable (Fig. 3, I to L) and mimicked the pattern observed in circulating and appendiceal MAIT cells from patients with appendicitis (Fig. 2, D, E, H, and I), suggesting that CCR4 may play a significant role in MAIT cell migration during appendicitis. The in vitro treatments, on the other hand, led to the decrease of CCR6 and CXCR6 expression by MAIT cells from healthy donors (Fig. 3, I to L), which were not observed in appendicitis (Fig. 2, D, E, H, and I). Nevertheless, the overall MAIT cell chemokine receptor expression patterns following in vitro treatments with these proinflammatory cytokines were comparable to those observed in appendicitis (Fig. 2, D, E, H, and I). This suggests that such cytokines may prime MAIT cell tissue–homing capacity toward the infected and inflamed appendix.

Inflammatory cytokines promote efficient MAIT cell migration to PDAOs

Next, we established the PDAO to test the hypothesis that MAIT cells migrate to the infected and inflamed appendix. Intestinal crypts were isolated according to established protocols and embedded in a dome of Matrigel extracellular matrix (ECM) (60). Such PDAOs were composed of intestinal epithelial cells with “apical-in” and “basal-out” polarity as determined by filamentous (F-) actin and EpCAM stainings (Fig. 4, A and B), thus recapitulating the proper and original polarity of the appendix tissue microenvironment. While undifferentiated, the appendix-derived organoids display the typical cystic appearance (Fig. 4A). However, after applying the differentiation protocol, organoids developed tubular structures resembling the intestine with mature enterocytes as determined by Krt20 staining (Fig. 4, A and B). To test whether there was a preferential migration of PB MAIT cells to PDAO, we cocultured the organoids with peripheral blood mononuclear cells (PBMCs) that were either untreated or prestimulated with IL-12 + IL-18 or IL-1β + IL-23. PBMCs were cultured outside the ECM domes and allowed to migrate into the ECM domes and organoids that were pre-embedded with E. coli (Fig. 4C, fig. S3S), the most common bacteria found in the inflamed appendix (fig. S1D). Using this methodology, there was a clear MAIT cell enrichment within the organoid-containing ECM domes when compared to empty ECM domes, nonmigrated PBMC outside the ECM dome, or the original PBMC population (Fig. 4, D to G). There were significantly more MAIT cells migrated into the ECM domes when E. coli–infected organoids were present, indicating preferential MAIT cell migration to the intestinal organoids (Fig. 4H). The different proinflammatory cytokines did not substantially influence MAIT cell enrichment within the ECM domes containing infected organoids (Fig. 4, D to G). However, there were significantly more migrating MAIT cells into the ECM domes containing E. coli–infected organoids in IL-12 + IL-18– or IL-1β + IL-23–stimulated PBMCs than in unstimulated PBMCs (Fig. 4H).

To visualize MAIT cell migration into E. coli–infected appendix organoids by confocal microscopy, the next set of experiments was conducted using expanded MAIT cells with a purity typically >85%, as previously described (17). These IL-2 + IL-7–supplemented expanded MAIT cells had comparable chemokine receptor expression profiles to those of IL-12 + IL-18 and IL-1β + IL-23–stimulated MAIT cells (fig. S4, A and B). Fluorescently labeled MAIT cells migrated into the organoids (fig. S4C), and such migration appeared directed toward the organoids with relatively fewer MAIT cells meandering within the ECM dome distal from the organoids or when the organoids were absent (Fig. 5A and fig. S4D). Further investigation by staining the organoids with an epithelial marker E-Cadherin indicated that these migrated MAIT cells appeared to associate with the organoids (Fig. 5B and fig. S4E). In the next set of experiments, expanded MAIT cells were rested for 2 days in the absence of exogenously added cytokines and then restimulated with IL-12 + IL-18, IL-1β + Il-23, or IL-2 + IL-7. In agreement with previous results (Fig. 4H), expanded MAIT cells restimulated with proinflammatory cytokines seemed to exhibit stronger migration into the appendix organoids than MAIT cells without prestimulation (Fig. 5C and fig. S4F). MAIT cells restimulated with IL-2 + IL-7 appeared to have the greatest capacity to migrate into the appendix organoids (Fig. 5C and fig. S4F), consistent with the ability of IL-7 to enhance MAIT cell functionality (17, 22, 57, 59). To ensure that the differences in migratory capacity of MAIT cells were not due to heterogeneity of the PDAO, the organoids were stained with the intracellular marker of mature enterocytes Krt20 following the live confocal microscopy (Fig. 5D). On the basis of Krt20 expression, the PDAO appeared relatively homogenous (Fig. 5D and fig. S4G), suggesting that the differences in MAIT cell migratory capacity was more likely due to the influence of the cytokines, in line with our previous results shown in Fig. 4H.

MAIT cell migration into PDAOs depends on CCR1, CCR2, and CCR4

To investigate the mechanism underlying MAIT cell migration into the infected and inflamed appendix, the plasma of appendicitis patients was first screened for inflammatory mediators, particularly the ligands of the chemokine receptors whose expression by MAIT cells was altered in appendicitis (Fig. 2, D, E, H, and I, and fig. S5A). There was a significant increase in the plasma of appendicitis patients of CCL2, CCL3, CCL8, CCL17, CCL20, CCL22, and CXCL8 (Fig. 6A), while there was no difference in the levels of CCL4, CCL5, CCL7, CCL11, CCL13, CXCL16, and CCL24 when compared to those of healthy controls (fig. S5B). We subsequently measured these inflammatory chemokines in the appendix tissues. Unexpectedly, similar chemokine profiles were observed in more inflamed appendix tissues when compared to less inflamed appendix, i.e., higher levels CCL2, CCL3, CCL8, CCL17, CCL22, CCL20, and CXCL8, but no difference in CCL4, CCL5, CCL7, CCL11, CCL13, CCL24, and CXCL16 levels (Fig. 6B and fig. S5C). These findings suggest that elevated inflammatory chemokines in the plasma were derived from the inflamed appendix tissues, and circulating MAIT cells may be recruited to the inflamed appendix via CCR1, CCR2, CCR3, CCR4, CCR5, CCR6, CXCR1, and/or CXCR6-dependent mechanisms.

Next, PDAOs were cultured to provide inflamed appendiceal secretomes to evaluate the specific chemokine receptors that may mediate MAIT cell migration to the inflamed appendix. To maximize the resemblance of the PDAO with the inflamed primary appendix tissues, only the secretomes from passage 0 PDAOs were used in subsequent experiments (Fig. 6C). Chemokine profiling of the PDAO secretomes revealed comparable profiles when compared to the inflamed primary appendix tissues as well as plasma of patients with appendicitis, with detectable levels of CCL2, CCL3, CCL8, CCL17, CCL22, CCL20, and CXCL8 (Fig. 6D). Of note, CCL20, CCL22, and CXCL8 levels were high in the PDAO secretomes (Fig. 6D). While the levels of CCL5, CCL7, CCL11, and CCL24 were relatively low and comparable to the patterns exhibited by the inflamed primary appendix tissues, PDAO secretomes in contrast had high levels of CCL4, CCL13, and CXCL16 (Fig. 6D). Thus, all of the chemokines that were significantly elevated in primary appendix tissues and plasma were also detected in the PDAO secretomes (fig. S5D). Almost half of the measured chemokines that were not elevated in primary appendix tissues and plasma were also not detected or only present at low levels in the PDAO secretomes (fig. S5D). Taken collectively, the chemokine profiles of the PDAO passage 0 secretomes were similar to those of inflamed primary appendix tissues and plasma of the patients, suggesting a robust PDAO model to evaluate MAIT cell migratory capacity during acute appendicitis. This was next investigated using a transwell-based MAIT cell migration assay and the PDAO secretomes as the chemoattractant (fig. S5E). Because of limited amounts of supernatants, initial experiments were performed to evaluate MAIT cell migratory capacity toward pooled appendix-derived organoids supernatants. MAIT cells rapidly and strongly migrated toward PDAO supernatants but not to the control organoid growth media (OGM) or the MAIT cell expansion media (Fig. 6, E and F), in agreement with the observation that MAIT cells migrated toward the PDAO (Figs. 4, D to H, and 5). Next, to examine which chemokine receptors may contribute to MAIT cell migration during acute appendicitis, MAIT cells expanded in the presence of the MR1 ligand 5-OP-RU and the proinflammatory cytokines IL-12 and IL-18 were pretreated with various chemokine receptor antagonists before the transwell migration assay using PDAO passage 0 supernatants from a single patient with appendicitis. Antagonists for the chemokine receptors CCR1, CCR2, and CCR4 significantly reduced MAIT cell migration to the PDAO supernatants (Fig. 6, G and H). In contrast, antagonists of CCR3, CCR5, CCR6, CXCR1, and CXCR6 had no effects on MAIT cell migration (Fig. 6, G and H). These findings collectively indicate that MAIT cells may migrate to the infected and inflamed appendix via predominantly CCR1-, CCR2-, and CCR4-dependent mechanisms.

MAIT cells from inflamed appendices display activated, inflammatory, and cytotoxic profiles

Because MAIT cells release potent antimicrobial, cytotoxic, and proinflammatory effector proteins during gastrointestinal infections caused by many GNB [reviewed in (2)], commonly found during appendicitis (fig. S1D), we next explored their potential inflammatory roles during APA. Circulating MAIT cells from children with acute appendicitis expressed higher levels of the activation markers CD38 and human leukocyte antigen (HLA)–DR (Fig. 7, A and B). Moreover, CD38 and HLA-DR levels were higher in MAIT cells derived from more inflamed parts of the appendix compared with less/noninflamed parts or circulation of the same patients (Fig. 7, C and D). Subsequent evaluation of MAIT cell effector functions following stimulation with fixed E. coli (17, 22, 24, 35, 57, 61) showed higher levels of degranulation (CD107a), interferon-γ (IFN-γ), tumor necrosis factor, IL-17A, and IL-22 by MAIT cells from more inflamed parts of appendix than less/noninflamed parts (Fig. 7, E to I). Of note, there were also significantly higher levels of IFN-γ and IL-17A produced by ex vivo cultures of the more inflamed appendix parts (Fig. 7, J and K). These findings suggest that MAIT cells from patients with appendicitis, especially those from the inflamed appendix, were activated and produced enhanced levels of cytotoxic and proinflammatory mediators following encounters with GNB within the infected appendix.

MAIT cells migrate to bacteria-infected appendix-derived organoids and mediate MR1-dependent organoid cell death and structural destruction

Having established that MAIT cells migrate to the infected and inflamed appendix by using the PDAO models, as well as their inflammatory and cytotoxic profiles in the appendixes of patients with appendicitis (Fig. 7, A to K), we next explored whether migrated MAIT cells could promote appendiceal tissue damage and may thus participate in the pathogenesis of the disease. Using a 2D cell culture of PDAO that were dissociated into single cells from a 3D culture (fig. S5F), we enumerated MAIT cell cytotoxic capacity and organoid cell death using precise MAIT to organoid cell ratios not possible with the current 3D organoid technology. In addition, we used the MR1 ligand 5-OP-RU to allow rapid measurement of MAIT cell–mediated killing of organoid cells. In such experiments, MAIT cells [typical purity >85% (17, 62)] strongly degranulated (CD107a) following a short-term coculture with 2D organoid cells pulsed with 5-OP-RU but not when cocultured with organoid cells alone (Fig. 7L). Correspondingly, 5-OP-RU–pulsed organoid cells were rapidly and efficiently killed by MAIT cells, as indicated by active Caspase 3 expression (Fig. 7M). Neither 5-OP-RU nor MAIT cell coculture alone triggered organoid cell death (Fig. 7M). MR1-antibody severely inhibited MAIT cell degranulation and killing of 5-OP-RU–pulsed organoid cells (Fig. 7, L and M), indicating that antigen presentation by MR1 was essential. These findings indicate that PDAO could present the bacterial MR1 ligand 5-OP-RU and trigger MR1-dependent MAIT cell cytotoxicity and killing of the organoid cells.

Last, we revisited our 3D PDAO model to evaluate whether MAIT cells can efficiently migrate to and promote structural destruction and death of PDAO cocultured with E. coli. PDAO coculture with E. coli alone (up to 3 days) did not trigger significant organoid structural destruction and death (Fig. 7N, fig. S5G). Observable MAIT cell migration into uninfected 3D organoids was noted within 3 days of coculture; however, this did not promote organoid destruction and death despite the infiltration of MAIT cells into the organoids (Fig. 7N, fig. S5G). In contrast, there appeared to be an increase in MAIT cell migration to PDAO cocultured with E. coli compared to uninfected organoids (Fig. 7N, fig. S5G). MAIT cell migration was accompanied by severe destruction of the PDAO as indicated by the loss of organoid structural integrity as well as organoid cell death as indicated by positive costaining of annexin V and propidium iodide (PI) (n = 5; Fig. 7N, fig. S5G). Furthermore, MAIT cell proximity to, adherence, and infiltration of the 3D organoids were associated with the structural destruction and cell death of the PDAO (Fig. 7N, fig. S5G).

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. For the pediatric participants, written informed consent was obtained from adult legal guardians of the children before enrolment. The study was approved by the Ethics Committees of Zhangzhou Municipal Hospital of Fujian Province (Ref. 2022KYB277) and Tsinghua Shenzhen International Graduate School (Refs. 202171 and 202295).

Participants

Pediatric patients were recruited at the division of pediatric surgery in Zhangzhou Municipal Hospital of Fujian Province, while adult healthy donors were enrolled at Tsinghua Shenzhen International Graduate School. In the acute appendicitis cohort, we included pediatric patients (<14 years old) and admitted for acute appendicitis but excluded those with chronic appendicitis, acute exacerbation of chronic appendicitis, or other comorbidities such as chronic infection and autoimmune diseases. In the healthy children control cohort, we included <14-year-old patients admitted for minor surgical treatments for noninflammatory and noninfectious conditions. Those with acute or chronic infections, autoimmune diseases, elevated CRP levels, and leukocytes were excluded. All participants were screened for and were negative for HBV, HCV, HIV-1/2, and syphilis before surgery.

Blood and tissue processing

The blood was drawn from patients immediately at presentation. Ethylenediaminetetraacetic acid dipotassium (EDTA K2) was used as an anticoagulant. Plasma was separated from the cell pellet after spinning whole blood and stored at −80°C. PBMCs were isolated via Ficoll-Histopaque density gradient separation and cryopreserved in liquid nitrogen. The whole appendix was removed from pediatric patients who underwent surgery for acute appendicitis, dissected into two segments according to the gross severity of inflammation, and immersed in cold RPMI 1640 medium supplemented with gentamicin (50 μg/ml, Gibco) and 1% of penicillin-streptomycin. The samples were sent with ice packs from the hospital to the laboratory and processed within 24 hours.

Generation and maintenance of appendix-derived intestinal organoids

Appendiceal tissues were dissected and opened longitudinally with sterile scalpels, and the mucosae were flushed extensively with cold phosphate-buffered saline (PBS). The mucosae were scraped off and minced into small pieces (<5 mm) using the scalpel. Intestinal crypts were isolated by incubating the tissue fragments in Gentle Cell Dissociation Reagent (GCDR, STEMCELL Technologies) for 30 min on ice with constant agitation and then centrifuged at 500g at 4°C for 5 min. The pellets were dissociated by vigorous pipetting in Dulbecco’s modified Eagle’s medium/nutrient mixture F-12 (DMEMF12, Gibco) supplemented with 1% bovine serum albumin (Sigma-Aldrich) (DMEMF12-BSA) passed through a 70-μm strainer and centrifuged at 500g at 4°C for 5 min. The crypts were counted using a hemocytometer and resuspended in cold Matrigel ECM solution composed of 50% Matrigel Growth Factor Reduced Basement Membrane Matrix (Corning) and 50% DMEMF12-BSA at a concentration of 20,000 crypts/ml. Fifty microliters of crypt suspensions in Matrigel ECM solution were plated in 24-well plates and incubated for 15 min at 37°C until the Matrigel ECM solution solidified and then cultured with IntestiCult OGM (STEMCELL Technologies) supplemented with Primocin (100 μg/ml, InvivoGen), gentamicin (50 μg/ml, Gibco), and 10 μM Y-27632 (for the first 3 days; STEMCELL Technologies) at 37°C, 5% CO₂. Cell media were changed and replenished every 2 to 3 days. In selected experiments, immature organoids (<7 days after passage) were differentiated for a further 5 to 7 days or until tubular structures developed in human IntestiCult Organoid Differentiation Medium (STEMCELL Technologies).

PDAO typically formed large cystic or budded structures in 10 to 14 days when they were ready for passaging. The medium was removed, and PDAO was incubated in GCDR at 37°C for 10 min, followed by manual dissociation of organoids into small fragments through thorough yet gentle pipetting loaded with 200-μl pipette tips. Organoid fragments were pelleted at 500g at 4°C for 5 min and washed once with DMEMF12-BSA. The crypts were then resuspended in cold Matrigel ECM solution at 20,000 crypts/ml and cultured in 24-well plates as described above.

PBMC and MAIT cell migration assays

The PDAOs were typically serially passaged at least five times in a medium containing broad-spectrum antibiotics and antifungals before coculture with the PBMCs to minimize residual bacterial contamination and inflammatory status of the organoids. The intact PDAOs were then carefully harvested from the ECM dome and cocultured with E. coli before being reembedded into the ECM dome. Briefly, mature PDAOs were released from Matrigel via gentle pipetting, resuspended in DMEMF12-BSA, and manually counted by macroscopic visualization of the organoid suspension. Then, five organoids were carefully collected from the organoid suspension for each group using 200-μl pipette tips and were left to settle under 1g for 5 min at room temperature. After removal of supernatants, these mature ogranoids were reembedded in Matrigel ECM solution alone or with formaldehyde-fixed E. coli DH5α (5 × 10⁷ CFU/ml) (77) as indicated, plated, and incubated for 15 min at 37°C until the Matrigel ECM solution became solidified. The ECM alone controls containing the same number of E. coli were included in all experiments as indicated. Organoids were checked and recounted under the microscope to ensure seeding number accuracy. Cell suspensions in MAIT cell medium containing 2 × 10⁶ cells/ml PBMCs or MAIT cells, as indicated, were then added to the organoid cultures. In selected experiments, MAIT cells were labeled with CellTrace Violet (CTV) (Thermo Fisher Scientific) as previously described (17), before being added to the coculture system. The cells were cocultured for 2 to 3 days at 37°C/5% CO₂ thereafter harvested for flow cytometric analyses or stained for live confocal microscopy.

For flow cytometric analyses, cells from both the outside and inside of the Matrigel ECM domes were harvested and dissociated into single-cell suspensions. Briefly, supernatants were collected, and the wells were extensively washed with fluorescence-activated cell sorting buffer [Ca²⁺- and Mg²⁺-free PBS supplemented with 2% fetal bovine serum (FBS) and 2 mM EDTA] while taking care to preserve the integrity of the domes. The washed fractions were combined with the supernatants and kept on ice. To harvest cells from within the Matrigel ECM domes, the domes were gently aspirated from the wells and then dissociated in prewarmed Trypsin-EDTA (0.25%; Gibco) by extensive pipetting and vortexing for 2 min, followed by addition of FBS at final concentration of 10% (v/v). Cellular suspensions from both the outside and the inside of the Matrigel ECM domes were then pelleted at 500g at 4°C for 5 min, followed by flow cytometric analyses.

In selected experiments, MAIT cell chemotaxis to organoid-derived supernatants was performed. Briefly, 10% (v/v) of pooled supernatants of mature PDAO cultures (>day 7 after passage) from multiple donors, or 5% (v/v) of PDAO passage 0 supernatants from a single patient with appendicitis in OGM were transferred to the bottom chambers of 6.5-mm Transwell plates with 5-μm pore polycarbonate membrane inserts (Corning). 5-OP-RU and IL-12 + IL-18–expanded MAIT cell suspensions were loaded to the insert with the density of 10⁶/ml in MAIT cell medium and incubated at 37°C/5% CO₂ for 3 hours and harvested for flow cytometric analyses. In selected experiments, MAIT cells were pretreated with 100 nM of various chemokine receptor antagonists (table S4) for 2 hours before the chemotaxis assay.

MAIT cell cytotoxicity assay in 2D and 3D cell cultures

The MAIT cell cytotoxicity protocols in 2D cell cultures were adapted from our previously described methodology (62). Briefly, mature cultures of PDAOs were harvested as described above and dissociated into single-cell suspension by incubating in GCDR at room temperature for 10 min followed by 0.05% Trypsin-EDTA digestion at 37°C for 5 min. Unviable organoid cells were removed using human dead cell removal kit according to the manufacturer’s instruction (MojoSort, BioLegend). Viable organoid cells were resuspended in MAIT cell medium, plated, and rested at 37°C/5% CO₂ for 1 hour, followed by addition of 5-OP-RU at a final concentration of 2 nM and incubated for 2 hours. Anti-MR1 monoclonal antibody (mAb, 26.5; BioLegend) or IgG2a isotype control (MOPC-173; BioLegend) was added and incubated for 1 hour, followed by the addition of CTV-labeled MAIT cells at an effector to target cell (E:T) ratio of 5:1. Anti–CD107a-PE (1:200) was added into the culture medium at the start of the assay to assess MAIT cell degranulation. After 3 hours of culture at 37°C/5% CO₂, cells were harvested and stained for flow cytometry as indicated. In the 3D experiment, mature PDAOs were retrieved from the Matrigel ECM domes, incubated with or without formaldehyde-fixed E. coli, reembedded in Matrigel, and cocultured with CTV-labeled MAIT cells for 2 to 3 days, as described above. The cytotoxicity of MAIT cells toward PDAO was visualized using confocal microscopy after staining with annexin V and PI.

Preparation of appendix tissue and appendix-derived organoid supernatants

The supernatants of mature culture of PDAO or from passage 0 of PDAOs were collected every 2 to 3 days during half-media changes. Supernatants were then pooled from multiple donors from different time points or the same PDAO donor from different time points as indicated, aliquoted, snap-frozen in liquid nitrogen, and stored in −80°C until further use. In selected experiments, 10⁶ cells derived from inflamed and less/noninflamed portions of appendix tissues were cultured in MAIT cell medium for 24 hours at 37°C 5% CO₂. Supernatants were collected, clarified, and stored in −80°C until further use.

Preparation of 5-OP-RU

The MAIT cell antigen 5-OP-RU was synthesized as a solution in dimethyl sulfoxide (DMSO) (and its concentration quantified by nuclear magnetic resonance spectroscopy) as previously reported (6, 78, 79). It is stable in DMSO solutions but converts rapidly to much less active lumazines in aqueous media, and the exposure time should be minimized as much as possible to maximize activity. All 5-OP-RU working solutions were diluted from the DMSO stocks with the appropriate culture medium immediately before any functional assay.

Flow cytometry

Cells were stained for surface and intracellular molecules in accordance with a protocol described previously (77). The antibodies were listed in table S4. The flow cytometry was performed in CytoFLEX (Beckman Coulter) equipped with 405-, 488-, and 638-nm lasers. The data were analyzed with Flowjo v.10 (BD Biosciences). MAIT cells were identified as Vα7.2⁺ CD161^hi CD3⁺ T cells or Vα7.2⁺ MR1–5-OP-RU⁺ CD3⁺ T cells as indicated. Cell numbers were enumerated using flow cytometry–based Precision Count Beads (BioLegend) according to the manufacturer’s instructions.

Immunofluorescence confocal microscopy

These protocols used the whole-mount staining of PDAO plated on 96-well or 24-well polystyrene tissue culture-treated plates (Corning). In brief, the PDAOs were first fixed with TF Fix/Perm (BD Biosciences) for 15 min at 37°C and washed carefully with TF Perm/Wash (BD Biosciences) to maintain the integrity of the Matrigel ECM domes. The PDAOs were subsequently stained for 30 min at 37°C in TF Perm/Wash buffer with fluorochrome-conjugated mAbs against EpCAM-1 and Krt20 as listed in table S4. The nuclei and actin filaments of organoids were concurrently counterstained with 4′,6-diamidino-2-phenylindole and phalloidin, respectively, in accordance with the manufactures’ instructions (table S4). In selected experiments, the PDAOs were live-stained with fluorochrome-conjugated mAb against E-cadherin, and the nuclei were counterstained using cell permeant SYTO21 nucleic acid stain for 30 min at 37°C in binding buffer (MAIT cell medium supplemented with 1 mM CaCl₂ and 1 mM MgCl₂). In the PDAO and CTV-labeled MAIT cells coculture experiments, the cocultures were carefully live-stained with annexin V and PI for 30 min at 37°C to assess MAIT cell–mediated killing of organoids. Nuclei were concurrently stained by using SYTO21. In selected experiments where PBMCs were used, the cocultures were carefully stained with mAbs against E-cadherin and CD3 for 30 min at 37°C in binding buffer. The images were acquired with a Nikon A1R+ laser scanning confocal microscopy equipped with 4×/0.2, 10×/0.45, and 20×/0.45 objectives (Nikon). Imaging was controlled using NIS-Elements software (Nikon) and processed with ImageJ 1.53a software (National Institutes of Health) (80). Integrated fluorescence density was measured in ImageJ by using the Threshold function to separate regions of interest from the background. Lines were drawn using the line tool in the ImageJ software to estimate potential areas of colocalization and define the region of interest for subsequent analysis for overlapping fluorescence intensities.

Statistical analysis

Statistical analyses were performed using Prism software version 10.0.2 (GraphPad). Datasets were first assessed for data normality distribution. Data presented as a heatmap show the mean, whereas data presented as line graphs with shaded error bars represent the mean and SE. Truncated violin plots represent median, the interquartile range, the minimum and maximum observation for each group, and the approximate frequency of data points. 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 analysis of variance (ANOVA), the Friedman test, ordinary ANOVA, the repeated-measures 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 Spearman rank correlation. Two-sided P < 0.05 was considered significant.

This PDF file includes:

Other Supplementary Material for this manuscript includes the following: