Fiber- and acetate-mediated modulation of MHC-II expression on intestinal epithelium protects from Clostridioides difficile infection

SUMMARY

Here, we explore the relationship between dietary fibers, colonic epithelium major histocompatibility complex class II (MHC-II) expression, and immune cell interactions in regulating susceptibility to Clostridioides difficile infection (CDI). We find that a low-fiber diet increases MHC-II expression in the colonic epithelium, which, in turn, worsens CDI by promoting the development of pathogenic CD4⁺ intraepithelial lymphocytes (IELs). The influence of dietary fibers on MHC-II expression is mediated by its metabolic product, acetate, and its receptor, free fatty acid receptor 2 (FFAR2). While acetate activation of FFAR2 on epithelial cells helps resist CDI, it does not directly regulate MHC-II expression. Instead, MHC-II is regulated by FFAR2 in type 3 innate lymphoid cells (ILC3s). Acetate enhances interleukin-22 (IL-22) production by ILC3s, which then suppresses MHC-II expression on the colonic epithelium. In conclusion, a low-fiber diet reduces acetate-induced IL-22 production by ILC3s, leading to increased MHC-II on the colonic epithelium. This change affects recovery from CDI by expanding the population of pathogenic CD4⁺ IELs.

In brief

Fachi et al. show that a low-fiber diet worsens CDI by increasing MHC-II in colonic epithelium, promoting pathogenic CD4⁺ IELs. Acetate from fibers enhances ILC3-derived IL-22 via FFAR2, suppressing MHC-II expression. This study highlights the links between gut microbiota, dietary fibers, acetate, and host immune responses in CDI susceptibility.

Graphical Abstract

INTRODUCTION

Clostridioides difficile is an opportunistic enteric pathogen and the leading cause of hospital-acquired gastrointestinal infections.¹ Normally, its colonization is prevented by the gut’s microbial community. However, C. difficile infections (CDIs) typically arise when antibiotics disrupt the normal gut microbiota, diminishing its ability to control C. difficile growth.^2–4 This disruption allows C. difficile to multiply and produce toxins, which severely damage the colonic epithelium, causing inflammation and symptoms that range from mild diarrhea to severe pseudomembranous colitis.^5,6 CDI represents a significant public health threat, especially in healthcare settings where antibiotic use is common.

The host defense against acute CDI primarily relies on innate immune responses that limit and repair the damage caused by C. difficile.^7,8 In mice, the early response is characterized by the rapid recruitment of neutrophils and innate lymphoid cells (ILCs), specifically group 1 (ILC1s) and group 3 (ILC3s), to the site of infection.^9–12 Neutrophils help control the bacterial burden through phagocytosis and the production of reactive oxygen species (ROS).^10,13,14 ILC1s enhance phagocytic activity and increase the expression of ROS-producing enzymes in the colonic mucosa through their production of interferon (IFN)-γ.¹¹ ILC3s help maintain the epithelial barrier by producing interleukin-22 (IL-22). This cytokine promotes the production of antimicrobial peptides and ROS by epithelial cells, and it also activates repair mechanisms.^11,12,15,16 T and B cells play a minimal role in resolving the acute phase of CDI but are more relevant in recurrent CDI. Immunoglobulins neutralize toxins both at the mucosal surface of the intestine and systemically.^17,18 IL-17A produced by γδ T cells protects newborn infants from CDI.¹⁹ CD4⁺ Foxp3⁺ regulatory T (Treg) cells promote the colonization of the intestine by healthy fecal microbiota that has been transferred to correct dysbiosis and resolve CDI.²⁰ In mice deficient in major histocompatibility complex class II (MHC-II), which lack CD4⁺ T cells, the production of toxin-specific antibodies and protection against recurrent CDI are reduced compared with wild-type (WT) mice.²¹

The intestinal microbiota provides colonization resistance against CDI through various mechanisms.²² It competes for growth-related metabolites like sialic acid and succinate,^23,24 produces bacteriocins that target C. difficile,²⁵ and it generates metabolites such as secondary bile acid derivatives,^26–28 tryptophan derivatives,²⁹ and short-chain fatty acids (SCFAs)^30,31 that inhibit C. difficile growth and enhance immune responses. Since many bacterial metabolites derive from the degradation of dietary compounds, diet can influence the incidence and severity of CDI. Naturally fiber-rich foods like grains, fruits, and vegetables are readily fermented by intestinal bacteria to produce SCFAs, which protect against CDI by supporting both epithelial cell resistance and immune response.^32,33 Conversely, a diet low in fiber alters the gut microbiota and its metabolic products, impairing intestinal health and increasing susceptibility to CDI.^34,35

In this study, we found that dietary fiber content impacts CDI by controlling the colonic epithelial expression of MHC-II and CD4⁺ intraepithelial lymphocytes (IELs) function. A low-fiber diet (LFiD) increased MHC-II expression on the colonic epithelium, which aggravated CDI by inducing pathogenic IFN-γ-producing CD4⁺ IELs that triggered excessive inflammation. Mechanistically, we showed that acetate, a byproduct of fiber metabolism, is essential for controlling MHC-II expression. Acetate acted on ILC3s, boosting IL-22 production, which then reduced MHC-II levels in the colonic epithelium.

RESULTS

Dietary fiber content modulates CDI severity

We investigated the impact of diets with varying fiber content on CDI. Mice were fed a standard chow diet (SD), a LFiD with 0% soluble fiber, or a high-fiber diet (HFiD) with 10% inulin. After 3 weeks, mice were treated with antibiotics and infected with 10⁸ colony-forming units (CFUs) of C. difficile (Figure 1A). LFiD resulted in more severe disease and delayed recovery from infection compared with SD, while HFiD reduced disease severity and promoted faster recovery, as indicated by survival rates (Figure 1B), weight loss (Figure 1C), and clinical score (Figure 1D). Additionally, the C. difficile burden in the feces was higher in LFiD mice and lower in HFiD mice compared with the SD group (Figure 1E).

Figure 1. Dietary fibers influence disease severity in a mouse model of CDI.

(A) Experimental design: mice were fed SD, LFiD, or HFiD, followed by antibiotics and C. difficile infection (10⁸ CFU).

(B–D) Survival rates (B), body weight (C), and clinical score (D) during CDI (n= 20).

(E) C. difficile abundance in feces after infection (n = 10).

(F) Colon and cecum length 4 d.p.i. (n= 4–5).

(G) Fecal lipocalin-2 levels 4 d.p.i. (n = 5).

(H) Frequency of CD11b⁺ Ly6G⁺ neutrophils and CD11b⁺ Ly6C⁺ inflammatory monocytes in the colonic LP 4 d.p.i. (n= 3–4).

(I) Representative H&E cecum and colon sections 4 d.p.i. Scale bars, 50–100 μm.

(J) Histopathological score of colonic sections 4 d.p.i. (n= 4–5).

(K) Ki67 staining for intestinal cell proliferation (n= 4).

(L and M) Goblet cell analysis by Alcian blue (AB) and periodic acid Schiff (PAS) staining of colon sections 4 d.p.i. (n = 4). Scale bars, 50 μm.

Error bars, mean ± SEM. Statistical analysis: t test or one-way ANOVA with Tukey’s post hoc test. *p < 0.05, **p < 0.01, ***p< 0.001.

We assessed colon and cecum shortening as indicators of tissue inflammation. 4 days post-infection (d.p.i.), LFiD mice showed more pronounced shortening than SD mice, whereas HFiD mice experienced less shortening (Figure 1F). These results were consistent with fecal lipocalin-2 levels (Figure 1G) and the infiltration of neutrophils and inflammatory monocytes in the colonic lamina propria (LP) (Figure 1H). Conversely, no significant differences were observed in macrophage and dendritic cell populations (Figure S1A). Histopathological examination of the cecum and colon 4 d.p.i. showed that LFiD exacerbated inflammation, edema, and epithelial barrier damage compared with SD. Conversely, the HFiD attenuated these effects, leading to better preservation of tissue integrity (Figure 1I). These observations were quantified using a histopathological score ranging from 0 to 30, evaluating 10 parameters scored from 0 (normal) to 3 (severe).³⁶ LFiD mice had higher score, indicating more severe disease, while HFiD mice showed reduced score compared with the SD group (Figure 1J). These findings were further supported by examining intestinal epithelial cell (IEC) proliferation within the crypts, which reflects epithelial responses following damage (Figures 1K, S1B, and S1C), and by assessing goblet cell frequency, another marker of tissue damage and inflammation (Figures 1L and 1M). In both cases, LFiD mice showed increased disease severity, whereas HFiD mice exhibited a mitigated response to CDI compared with SD mice.

We also examined the impact of dietary fibers on epithelial cell death 4 d.p.i. using terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) staining. LFiD mice exhibited significantly increased epithelial cell death in the cecum and colon compared with both SD and HFiD groups (Figure 2A). Quantification on a scale of 0 to 3, where 0 represents normal tissue and 3 indicates extensive cell death, confirmed that LFiD mice had greater epithelial cell death (Figure 2B). Further evidence of epithelial barrier disruption was obtained by measuring intestinal permeability to fluorescein isothiocyanate (FITC)-dextran (Figure 2C) and assessing bacterial translocation and systemic dissemination via 16S rDNA quantification in the liver, spleen, and mesenteric lymph nodes (mLNs) (Figure 2D). These results confirmed that LFiD impaired epithelial barrier integrity during CDI, while HFiD enhanced it.

Figure 2. HFiD enhances epithelial integrity and reduces MHC-II expression in colonic epithelial cells during CDI.

(A) Representative TUNEL staining of colon and cecum sections 4 d.p.i. Scale bars, 20 μm (40×) or 50 μm (20×).

(B) Scoring of TUNEL⁺ epithelial cells (n= 4).

(C) Intestinal permeability 4 d.p.i. measured by FITC-dextran assay (n = 3–4).

(D) Bacterial translocation in liver, spleen, and mLN 4 d.p.i. (n= 4).

(E) H2-Ab1 and Ciita mRNA levels in proximal colon during CDI (n= 3–7).

(F) Immunofluorescence of colon sections for MHC-II expression in epithelial cell adhesion molecule (EpCAM)⁺ cells 4 d.p.i. Scale bars, 20 μm. DAPI(4′,6-diamidino-2-phenylindole) stains nuclei.

(G and H) Representative fluorescence-activated cell sorting (FACS) plots (G) and quantification of frequency and gMFI of MHC-II expression (H) in colonic epithelial cells 4 d.p.i. (n = 3–4).

(I) FACS plots and frequency of CD3⁺ CD4⁺ T cells in epithelium and LP 4 d.p.i. (n= 3–4).

(J) mRNA levels of cytokines in proximal colon 4 d.p.i. (n = 4).

Error bars, mean ± SEM. Statistical analysis: t test or one-way ANOVA with Tukey’s post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant.

LFiD augments MHC-II expression on colonic epithelium and IFN-γ-producing CD4⁺ IELs during CDI

Recent reports suggest that MHC-II expression in colonic epithelial cells is induced by inflammation.^37–39 Consistent with this, expression levels of H2-Ab1, which encodes an MHC-II molecule, and Ciita, a key transcriptional inducer of MHC-II, increased in the proximal colon during CDI and peaked 3 d.p.i. (Figure 2E). LFiD upregulated H2-Ab1 and Ciita throughout CDI. HFiD had no major effects during early infection (1–2 d.p.i.) but markedly reduced H2-Ab1 and Ciita during recovery (3–4 d.p.i.) (Figure 2E). Immunostaining of colon sections for MHC-II 4 d.p.i. confirmed higher expression in the epithelium of LFiD mice compared with SD and HFiD mice (Figure 2F). Flow cytometry of live CD45⁻ EpCAM⁺ colonic epithelial cells (Figure 2G) showed that LFiD increased both the frequency and geometric mean fluorescence intensity (gMFI) of MHC-II expression compared with the SD group 4 d.p.i., while HFiD reduced MHC-II expression (Figure 2H). HFiD-fed mice also exhibited lower epithelial MHC-II expression 2 d.p.i. compared with the LFiD group (Figure S1D). No significant differences in MHC-II expression were observed in CD45⁺ hematopoietic cells 4 d.p.i. (Figure S1E).

Given the importance of MHC-II in CD4⁺ T cell priming and differentiation, we next evaluated the frequency of CD4⁺ T cells in the colonic epithelium and LP. LFiD mice displayed a higher frequency of CD4⁺ IELs compared with SD and HFiD groups, though no significant differences were observed in the colonic LP 4 d.p.i. (Figure 2I). Additionally, mRNA levels of Ifng and Il17a were elevated, while Il10 levels were reduced in the proximal colon of LFiD mice compared with SD and HFiD mice (Figure 2J). These findings suggest that dietary fiber content modulates MHC-II in colonic epithelial cells, affecting CD4⁺ IEL frequency and colonic cytokine content during CDI.

MHC-II deficiency on intestinal epithelium mitigates the impact of LFiD on CDI

We investigated whether MHC-II deficiency affects CDI severity. H2-Ab1^−/− mice on LFiD, and to a lesser extent those on SD, showed milder CDI outcomes than WT mice, as indicated by reduced weight loss and clinical scores (Figure 3A). LFiD-fed H2-Ab1^−/− mice also had reduced C. difficile CFU burden in the feces (Figure 3B) and reduced intestinal permeability (Figure 3C). Both LFiD- and SD-fed H2-Ab1^−/− mice displayed less colon shortening (Figure S2A) and reduced absolute numbers of neutrophils (Figure 3D) 4 d.p.i. than WT mice, indicating milder inflammation. The numbers of dendritic cells and monocytes were unaffected (Figure S2B). No differences were observed between HFiD H2-Ab1^−/− and HFiD WT mice (Figures 3A–3D). Thus, MHC-II deficiency on the intestinal epithelium lessens the impact of LFiD on CDI.

Figure 3. MHC-II deficiency in intestinal epithelial cells mitigates CDI outcomes.

(A) Body weight and clinical score during CDI in WT or H2-Ab1^−/− mice on different diets (n = 5).

(B) C. difficile burden in feces 4 d.p.i. (n= 4).

(C) Intestinal permeability assessed by FITC-dextran 4 d.p.i. (n= 4).

(D) Number of CD11b⁺ Ly6G⁺ neutrophils in the colonic LP 4 d.p.i. (n = 4).

(E) Experimental design of bone marrow (BM) chimera: CD45.2 WT or H2-Ab1^−/− mice were reconstituted with CD45.1 WT BM (n = 5). Mice were maintained on SD.

(F) Body weight and clinical score of WT or H2-Ab1^−/− BM chimeric mice in CDI (n= 5).

(G) C. difficile burden 4 d.p.i. (n = 5).

(H) Intestinal permeability measured by FITC-dextran assay 4 d.p.i. (n = 5).

(I) Body weight and clinical score of H2-Ab1^fl/fl and H2-Ab1^ΔIEC mice during CDI (n= 5). Mice were fed SD.

(J) Fecal C. difficile burden 5 d.p.i. (n= 5).

(K) Colon length in H2-Ab1^fl/fl and H2-Ab1^ΔIEC mice uninfected 5 d.p.i. (n = 4–5).

(L) Bacterial translocation in liver, spleen, and mLN (n= 5).

(M) Representative H&E colon sections and cumulative histopathological score 5 d.p.i. (n = 5). Scale bars, 100 μm.

(N) Fecal lipocalin-2 levels 5 d.p.i. (n= 3).

(O) Number of CD11b⁺ Ly6G⁺ neutrophils and CD11b⁺ Ly6C⁺ monocytes in colonic LP 5 d.p.i. (n= 4–9).

(P) S100a8 mRNA levels in proximal colon 5 d.p.i. (n = 4–5).

(Q and R) Frequency of CD3⁺ CD4⁺ T cells (Q) and CD4⁺ Foxp3⁺ Treg or CD4⁺ T-bet⁺ Th1 (R) cells in the epithelium of H2-Ab1^fl/fl and H2-Ab1^ΔIEC mice either uninfected or 5 d.p.i. (n= 4–9).

(S) Quantification of CD3⁺ CD4⁺, CD4⁺ Foxp3⁺, and CD4⁺ T-bet⁺ T cells in colonic LP 5 d.p.i. (n = 4–9).

(T) mRNA expression of cytokines in the proximal colon 5 d.p.i. (n = 4–5).

Error bars, mean ± SEM. Statistical analysis: one-way ANOVA with Tukey’s post hoc test. *p< 0.05, **p< 0.01, ***p< 0.001, ****p< 0.0001; ns, not significant.

Since MHC-II is expressed on both stromal and hematopoietic cells, we conducted a bone marrow chimera experiment to pinpoint the compartment in which MHC-II deficiency impacts CDI. CD45.2 WT and CD45.2 H2-Ab1^−/− mice were irradiated and then reconstituted with bone marrow from CD45.1 WT mice (Figure 3E). 8 weeks later, successful reconstitution was confirmed by the predominance of CD45.1⁺ (donor) cells over CD45.2⁺ (recipient) cells in the spleen (Figure 3E). After CDI, H2-Ab1^−/− mice reconstituted with WT bone marrow showed greater resistance to infection (Figure 3F), a lower C. difficile CFU burden (Figure 3G), and reduced intestinal permeability (Figure 3H) 4 d.p.i. compared with WT mice that received WT bone marrow.

To validate these findings, we conditionally deleted H2-Ab1 in IECs by crossing H2-Ab1^fl/fl mice with Villin^Cre mice (H2-Ab1^ΔIEC). H2-Ab1^ΔIEC mice were more resistant to CDI than H2-Ab1^fl/fl mice, as shown by reduced weight loss and clinical score (Figure 3I) and lower C. difficile CFU burden in feces 5 d.p.i. (Figure 3J). At steady state, the colons were shorter in H2-Ab1^ΔIEC mice compared with H2-Ab1^fl/fl mice. However, H2-Ab1^fl/fl mice showed increased colon shortening 5 d.p.i. (Figure 3K). Resistance to CDI in H2-Ab1^ΔIEC mice was further confirmed by reduced bacterial translocation, improved histopathological score, lower fecal lipocalin-2 levels, fewer infiltrating neutrophils and monocytes, and decreased expression of S100a8, a marker of pro-inflammatory cell recruitment, compared with H2-Ab1^fl/fl mice 5 d.p.i. (Figures 3L–3P).

We further examined whether MHC-II deficiency in IECs affects the impact of dietary fibers on CDI. H2-Ab1^ΔIEC mice remained more resistant to CDI than H2-Ab1^fl/fl mice even on an LFiD, which exacerbated CDI in WT mice (Figure S2C; see also Figure 1). H2-Ab1^ΔIEC mice on LFiD showed reduced inflammation and enhanced epithelial barrier integrity compared with H2-Ab1^fl/fl mice on the same diet (Figures S2D–S2G). Conversely, an HFiD made H2-Ab1^fl/fl mice as resistant to CDI as H2-Ab1^ΔIEC mice. These findings suggest that the absence of MHC-II in the intestinal epithelium lessens the detrimental effects of an LFiD on CDI but does not enhance the protective effects of an HFiD.

Lack of MHC-II in intestinal epithelium reduces Th1 IELs during CDI

Given the reduced inflammation in H2-Ab1^ΔIEC mice during CDI, we asked whether the absence of MHC-II in IECs impacts T cell responses. H2-Ab1^ΔIEC mice had fewer CD4⁺ IELs compared with H2-Ab1^fl/fl mice after CDI but no difference at steady state (Figure 3Q). No differences were observed in other IEL populations, including CD8⁺ and γδT cells at steady state or 5 d.p.i. (Figure S3A). Within CD4⁺ IELs, H2-Ab1^ΔIEC mice showed increased Foxp3⁺ CD4⁺ Treg and decreased T-bet⁺ CD4⁺ Th1 IELs compared with H2-Ab1^fl/fl mice 5 d.p.i. (Figure 3R). No changes were observed for RORγt⁺ Th17 and GATA3⁺ Th2 (Figure S3B). The numbers of total CD4⁺ T cells, Treg, Th1, Th17, and CD8⁺ T cells in the colonic LP were similar between H2-Ab1^fl/fl and H2-Ab1^ΔIEC mice at steady state and 5 d.p.i. (Figures 3S and S3C). Next, we assessed cytokine production in the proximal colon 5 d.p.i., normalized to H2-Ab1^fl/fl uninfected controls (Figure 3T). H2-Ab1^ΔIEC mice exhibited lower levels of Ifng and Tnf mRNA compared with H2-Ab1^fl/fl mice and higher levels of Il10 (Figure 3T). These findings suggest that the lack of epithelial MHC-II expression in H2-Ab1-^ΔIEC mice leads to a reduction in CD4⁺ Th1 IELs during CDI.

The impact of LFiD on CDI partly depends on T cells

IFN-γ has been shown to induce MHC-II expression in IECs.^38,40,41 Thus, we asked whether T cells contribute to the effects of LFiD on CDI outcomes and the induction of epithelial MHC-II. We examined the impact of SD, LFiD, or HFiD on CDI in Rag1^−/− mice. In contrast to WT mice (see Figure 1), LFiD did not impact CDI recovery relative to SD and HFiD in Rag1^−/− mice (Figure S3D). Similarly, no differences were observed in colon length (Figure S3E), fecal lipocalin-2 levels (Figure S3F), or S100a8 mRNA expression 4 d.p.i. (Figure S3G). Moreover, levels for Ifng, Tnf, and Il10 mRNA did not show significant differences between groups 4 d.p.i. (Figure S3H). These findings suggested that the impact of an LFiD on CDI is at least partly due to adaptive immune cells, most likely T cells, as they influence the severity of the infection and inflammation. However, we did note increased bacterial translocation in LFiD-fed Rag1^−/− mice compared with SD and HFiD groups 4 d.p.i. (Figure S3I). Additionally, Rag1^−/− mice on LFiD still exhibited a higher frequency of MHC-II⁺ CD45⁻ EpCAM⁺ cells compared with those on SD and HFiD (Figure S3J), suggesting that some effects of LFiD may involve other mechanisms.

Acetate controls MHC-II in colonic epithelium and influences CDI outcomes

Since dietary fibers are a source of SCFAs,⁴² we investigated whether SCFAs influence MHC-II expression in the intestinal epithelium during CDI. Mice were fed either LFiD or HFiD for 2 weeks. In a cohort of HFiD mice, microbiota and SCFA production were disrupted by a 7-day antibiotic regimen (vancomycin, neomycin, ampicillin, and metronidazole [VNAM]). The antibiotic-treated HFiD mice were then divided into two groups: one group received a mix of SCFA in their drinking water throughout the experiment (HFiD + VNAM + SCFA), while the other remained untreated (HFiD + VNAM). All groups received CDI antibiotics on day −6, clindamycin on day −1, and were infected with 10⁸ CFU of C. difficile on day 0 (Figure 4A). As previously observed, WT mice on HFiD showed greater resistance to CDI compared with those on LFiD, as evidenced by body weight changes and clinical scores (Figure 4B). However, when HFiD mice were treated with VNAM, this protective effect was lost, leading to increased weight loss, clinical score, colon shortening, tissue inflammation, and bacterial translocation to the liver, spleen, and mLNs (Figures 4B–4E). Conversely, supplementing the HFiD + VNAM mice with SCFAs restored their resistance to CDI, reduced tissue inflammation, and prevented epithelial barrier leakage (Figures 4B–4E). Additionally, HFiD mice treated with VNAM showed increased MHC-II expression by IECs 4 d.p.i., an effect that was reversed by SCFA supplementation (Figure 4F). This suggests that SCFAs play a role in downregulating epithelial MHC-II expression. The changes in MHC-II expression were directly correlated with alterations in the frequency of CD4⁺ IELs (Figure 4G). Given that we previously found that the impact of diet on CDI is dependent on MHC-II (see Figure S2), we conducted similar experiments in H2-Ab1^−/− mice. Regardless of whether these mice were treated with HFiD, HFiD + VNAM, or HFiD + VNAM + SCFA, they remained equally resistant to CDI (Figures S4A–S4C), confirming that the effect of SCFAs on CDI is MHC-II-dependent.

Figure 4. SCFA suppresses epithelial MHC-II expression in the colon during CDI.

(A) Experimental design: mice were fed LFiD or HFiD from day −28. On day −14, HFiD mice received either a 7-day antibiotic regimen (VNAM) or no treatment. SCFA was provided in the drinking water to a group of HFiD + VNAM mice from day −6 until the end of the experiment. All mice received CDI antibiotics and were infected with C. difficile.

(B) Body weight and clinical score of mice on different fiber diets with/without VNAM and SCFA treatment (n = 4).

(C) Colon length of mice on different fiber diets, with/without VNAM and SCFA, 4 d.p.i. (n = 4).

(D) H&E colon sections and histopathological score 4 d.p.i. Scale bars, 100 μm.

(E) Bacterial translocation in liver, spleen, and mLN by 16S rDNA 4 d.p.i. (n= 4).

(F) MHC-II expression on colonic epithelial cells 4 d.p.i. (n= 4).

(G) Frequency of CD3⁺ CD4⁺ IELs 4 d.p.i. (n= 4).

(H) Body weight and clinical score of SCFA-treated mice (n = 4). Mice received acetate (Ac), butyrate (Bt), propionate (Pr), or water (control, Ct) from day −7, followed by antibiotics and C. difficile infection (day 0). All mice were maintained on SD.

(I and J) Colon length (I) and H&E colon sections with respective histopathological scores (J) of SCFA-treated mice 5 d.p.i. (n = 4). Scale bars, 100 μm.

(K and L) Frequency of MHC-II⁺ CD45⁻ EpCAM⁺ cells (K) and CD45⁺ CD3⁺ CD4⁺ IELs (L) in SCFA-treated mice 5 d.p.i. (n= 4).

(M) Ifng and Il10 mRNA levels in proximal colon 5 d.p.i., normalized to control (n = 4).

Error bars, mean ± SEM. Statistical analysis: one-way ANOVA with Tukey’s post hoc test. *p< 0.05, **p< 0.01, ***p< 0.001, ****p< 0.0001; ns, not significant.

We investigated the impact of individual SCFAs—acetate, propionate, and butyrate—on CDI outcomes and MHC-II expression in the intestinal epithelium. Mice were treated with these SCFAs in their drinking water starting 7 days before infection. All SCFA-treated mice showed significant protection during CDI compared with untreated controls, as demonstrated by reduced weight loss and clinical score, reduced colon shortening, and lower histopathological score (Figures 4H–4J), which was consistent with previous studies.^33,42 When analyzing MHC-II expression, we found that acetate and propionate were more effective than butyrate in reducing both the frequency and gMFI of MHC-II in CD45⁻ EpCAM⁺ IECs 5 d.p.i. (Figure 4K). Additionally, acetate-treated mice had a lower frequency of CD4⁺ IELs compared with controls (Figure 4L), along with reduced expression of Ifng and increased Il10 in the proximal colon 5 d.p.i. (Figure 4M). Il10 was also increased in mice treated with other SCFAs. In conclusion, among the SCFAs tested, acetate had the most pronounced effect on reducing MHC-II expression and CD4⁺ IEL expansion.

FFAR2 expression on the intestinal epithelium contributes to acetate-mediated effects on CDI but not to MHC-II suppression

Since acetate and propionate had a more pronounced effect on MHC-II expression, we investigated their role further using mice lacking the free fatty acid receptor 2 (FFAR2), which has a high affinity for acetate and propionate.⁴² We found that Ffar2^−/− mice treated with either an LFiD or an HFiD did not show significant differences in CDI outcomes compared with Ffar2^−/− mice on an SD (Figure 5A). There were no remarkable differences in colon shortening (Figure 5B), fecal lipocalin-2 levels (Figure 5C), bacterial translocation (Figure 5D), or histopathological score 4 d.p.i. (Figure 5E) between these groups. Additionally, there were no significant changes in the expression of H2-Ab1 and Ciita mRNA in the intestinal epithelium (Figure 5F) or in the levels of Ifng and Il10 mRNA in the proximal colon 4 d.p.i. (Figure 5G). These results suggest that the absence of FFAR2 eliminates the diet-related effects on CDI observed in WT mice.

Figure 5. FFAR2-deficiency exacerbates gut inflammation during CDI.

(A) Body weight and clinical score of Ffar2^−/− mice on different fiber diets during CDI (n = 5).

(B–E) Colon length (B), fecal lipocalin-2 levels (C), bacterial translocation to the liver, spleen, and mLN (D), and representative H&E colon sections with histopathological scores (E) 4 d.p.i. of Ffar2^−/− mice on SD, LFiD, and HFiD. Scale bars, 100 μm.

(F) H2-Ab1 and Ciita mRNA levels in colonic epithelial cells from Ffar2^−/− mice 4 d.p.i., normalized to SD group (n= 4).

(G) mRNA levels of Ifng and Il10 in the proximal colon of Ffar2^−/− mice 4 d.p.i. (n = 4).

(H) Body weight, clinical score, and colon length of Ffar2^fl/fl and Ffar2^ΔIEC mice maintained on SD during CDI (n = 5).

(I–J) Representative H&E colon sections with histopathological scores (I), and bacterial translocation (J) 5 d.p.i. Scale bars, 100 μm.

(K) MHC-II expression in CD45⁻ EpCAM⁺ epithelial cells 5 d.p.i. (n = 4).

(L) Frequency of colonic CD3⁺ CD4⁺ IELs 5 d.p.i. (n= 4).

(M) Ifng and Il10 mRNA expression in the proximal colon 5 d.p.i. (n = 4).

(N) Schematic of experimental design and representative immunofluorescence images of small intestine organoids for MHC-II and EpCAM expression, with or without IFN-γ and acetate (n= 3). Scale bars, 50 μm. DAPI stains nuclei.

(O) FACS plot and MHC-II gMFI of CD45⁻ EpCAM⁺ cells from organoid cultures stimulated or not with IFN-γ and acetate (n = 4).

Error bars, mean ± SEM. Statistical analysis: t test or one-way ANOVA with Tukey’s post hoc test. *p< 0.05, **p < 0.01; ns, not significant.

To explore whether acetate and FFAR2 directly affect epithelial cells and influence CDI outcomes and MHC-II expression, we examined mice with a conditional deletion of Ffar2 specifically in IECs (Ffar2^ΔIEC). These Ffar2^ΔIEC mice showed compromised resistance to CDI, as evidenced by increased weight loss and higher clinical scores compared with Ffar2^fl/fl mice (Figure 5H). They also had a higher histopathological score (Figure 5I) and increased bacterial translocation to the spleen and mLN (Figure 5J), although there were no differences in colon length (Figure 5H). Despite the worse disease outcome in Ffar2^ΔIEC mice, which was consistent with the protective role of acetate-FFAR2 in colon epithelial cells during colitis,⁴³ we did not find differences in MHC-II expression levels in the colon epithelium 5 d.p.i. compared with Ffar2^fl/fl mice (Figure 5K). Similarly, there were no differences in the frequency of CD4⁺ IELs (Figure 5L) or in Ifng mRNA level, although Il10 mRNA was reduced (Figure 5M).

To confirm that SCFAs modulate MHC-II expression in IECs through a cell-extrinsic mechanism, we conducted in vitro experiments. We treated the human Caco-2 epithelial cell line with IFN-γ for 24 h, with or without acetate, propionate, or butyrate. Although IFN-γ increased MHC-II expression in these cells, as indicated by human leukocyte antigen DR isotype (HLA-DR) levels, the addition of SCFAs did not reduce it (Figure S5A). We further tested this by culturing small intestine crypts from WT mice in a 3D system for 6 days (Figure 5N). In these intestinal organoids, IFN-γ treatment increased MHC-II (Figures 5N, 5O, S5B, and S5C). However, acetate did not prevent IFN-γ-induced MHC-II. Additionally, there were no differences in MHC-II expression between organoids derived from colonic crypts of WT and Ffar2^−/− mice when treated with IFN-γ and acetate (Figure S5D). These results suggest that the SCFA-mediated regulation of epithelial MHC-II during CDI is a cell-extrinsic effect.

Acetate-FFAR2 signaling suppresses epithelial MHC-II expression through induction of IL-22 in hematopoietic cells

We next explored the role of acetate-FFAR2 signaling in the immune compartment during CDI. Ffar2^fl/fl Vav^iCre (Ffar2^ΔVAV) mice, which lack FFAR2 in immune cells, showed greater weight loss and higher clinical score (Figure 6A). They also had more colon shortening (Figure S6A), increased expression of S100a8 mRNA (Figure S6B), and compromised epithelial barrier integrity (Figures 6B and S6C) compared with Ffar2^fl/fl mice after infection. In these Ffar2^ΔVAV mice, we observed heightened MHC-II expression in IECs (Figures 6C and S6D), increased frequency of CD4⁺ IELs (Figure 6C), and elevated Ifng and reduced Il10 mRNA levels in the colon (Figure 6D) 5 d.p.i. Thus, acetate-FFAR2 signaling suppresses epithelial MHC-II expression through hematopoietic cells.

Figure 6. Fiber and acetate-induced IL-22 production by ILC3s suppresses epithelial MHC-II expression during CDI.

(A) Body weight and clinical score of Ffar2^fl/fl and Ffar2^ΔVAV mice during CDI (n = 5). Mice were fed SD.

(B–E) Intestinal permeability by FITC-dextran (B), MHC-II gMFI in CD45⁻ EpCAM⁺ cells and frequency of CD3⁺ CD4⁺ IELs (C), and cytokines mRNA levels 5 d.p.i. (D and E) (n = 4).

(F–I) Body weight, clinical score, and colon length (F); relative fecal C. difficile abundance (G); MHC-II gMFI in CD45⁻ EpCAM⁺ cells and frequency of CD3⁺ CD4⁺ IELs (H); and Ifng and Il10 mRNA in proximal colon (I) of WT mice on SD treated with anti-IgG isotype or anti-IL-22 neutralizing antibody 1 and 3 d.p.i. (n= 4–5).

(J) Frequency of IL-22-producing ILC3s and CD4⁺ T cells in colonic LP at steady state (n = 4).

(K) Experimental design: CD45.1 mice were irradiated (11 Gy) and received a mixture of 3 × 10⁶ CD45.1/2 (WT) and 3 × 10⁶ CD45.2 (Ffar2^−/−) BM cells. Mice were allowed to reconstitute for 8 weeks, then treated with CDI antibiotics and infected with C. difficile. Mice were maintained on SD.

(L) Representative FACS plots showing BM reconstitution in the spleen and frequency of colonic ILC3s in mixed BM chimera mice 4 d.p.i. (n= 4).

(M) Frequency of CD45.1/2⁺ or CD45.2⁺ IL-22-producing ILC3s 4 d.p.i. after ex vivo isolation and incubation under different conditions: unstimulated (US), 10 ng/mL IL-23, or 50 ng/mL PMA + 500 ng/mL ionomycin (n = 4).

(N) FACS plots and frequency of MHC-II-expressing epithelial cells in WT, Rag2^−/−, and Rag2yc^−/− mice fed either LFiD or HFiD 4 d.p.i. (n = 4).

(O–R) Body weight, clinical score, and colon length (O); Il22 mRNA in proximal colon (P); MHC-II gMFI in CD45⁻ EpCAM⁺ cells and frequency of CD3⁺ CD4⁺ IELs (Q); and Ifng and Il10 mRNA (R) in proximal colon of AhR^fl/fl and AhR^ΔRORγt mice on SD (n = 5).

Error bars, mean ± SEM. Statistical analysis: t test or one-way ANOVA with Tukey’s post hoc test. *p< 0.05, **p < 0.01, ***p < 0.001; ns, not significant.

Given that IL-22 was recently reported to suppress epithelial MHC-II expression, partly by regulating endoplasmic reticulum stress,⁴⁴ and that acetate-FFAR2 signaling enhances IL-22 production by ILC3s,^12,45 we examined the impact of the acetate-FFAR2-IL-22 pathway on MHC-II expression during CDI. First, we confirmed that Ffar2^ΔVAV mice had reduced IL-22 levels in the proximal colon compared with Ffar2^fl/fl mice 5 d.p.i. (Figure 6E). We also assessed whether diet influences IL-22 levels. WT mice on HFiD had a higher frequency of IL-22-producing CD45⁺ cells in colonic LP at steady state compared with those on SD or LFiD (Figure S6E). Additionally, Il22 mRNA expression was upregulated in mice on HFiD compared with those on SD and LFiD 4 d.p.i. (Figure S6F). Mice treated with acetate and propionate also showed increased Il22 mRNA expression 5 d.p.i. (Figure S6G). By contrast, Il22 mRNA expression did not differ between Ffar2^−/− mice on different diets (Figure S6H). Together, these data indicated that fibers and SCFAs induce IL-22 through an FFAR2-dependent mechanism involving hematopoietic cells.

To define whether diet and acetate impact on CDI outcomes and epithelial MHC-II expression were due to IL-22 signaling, we treated C. difficile-infected mice with either an anti-IL-22 neutralizing antibody or an anti-IgG isotype control 1 and 3 d.p.i. Blocking IL-22 worsened CDI outcomes, as evidenced by increased weight loss, clinical score, and colon shortening (Figure 6F) and a greater C. difficile burden 5 d.p.i. (Figure 6G). Additionally, mice treated with the anti-IL-22 antibody showed higher epithelial MHC-II expression and a greater frequency of CD4⁺ IELs (Figure 6H). They also had elevated Ifng and reduced Il10 mRNA levels (Figure 6I). IL-22 blockade also impaired the protective effect of HFiD on CDI, as evidenced by increased disease severity, delayed recovery, and impaired pathogen clearance (Figures S6I–S6K). These findings align with those observed in acetate-treated mice.¹² HFiD also failed to suppress H2-Ab1 and Ifng mRNA expression in anti-IL-22-treated mice (Figure S6L). Together, these results emphasize that IL-22 produced by FFAR2⁺ immune cells in response to fibers and acetate is essential for suppressing epithelial MHC-II expression during CDI and ameliorating the disease outcomes.

IL-22 downregulating epithelial MHC-II derives from ILC3s

IL-22 is produced by ILC3s and CD4⁺ T cell subtypes, such as Th17.⁴⁶ To investigate which IL-22-producing lymphocytes are impacted by the dietary fiber content, we repeated the diet treatment for 2 weeks in mice at steady state. Mice on an LFiD showed reduced IL-22 production by ILC3s (Figure 6J) and had fewer ILC3s in the colonic LP compared with those on a SD (Figure S7A). By contrast, mice on an HFiD had higher IL-22 levels (Figure 6J) and more ILC3s in the colonic LP (Figure S7A) compared with both the SD and LFiD groups. No differences in IL-22 production by CD4⁺ T cells were observed across diets (Figures 6J and S7A). Additionally, HFiD increased Il22 mRNA expression in the proximal colon of Rag1^−/− mice 4 d.p.i. (Figure S7B), confirming that HFiD enhances IL-22 production in the CD45⁺ compartment even without T cells (Figures S7C and S7D). HFiD-fed Rag1^−/− mice also showed a higher frequency of ILC3s in the colonic LP compared with SD and LFiD Rag1^−/− mice (Figure S7E).

To confirm the role of FFAR2 in IL-22 production by ILC3s, we performed a mixed bone marrow chimera experiment. Irradiated CD45.1 WT recipient mice were reconstituted with a 1:1 mixture of WT (CD45.1/2) and Ffar2^−/− (CD45.2) bone marrow cells. 8 weeks later, mice were infected with C. difficile to assess ILC3 numbers and IL-22 production (Figure 6K). Both CD45.2⁺ and CD45.1/2⁺ cells were equally represented in the spleen after 8 weeks. However, in the colonic LP, WT ILC3s were more prevalent than Ffar2^−/− cells (Figure 6L). WT ILC3s were also more efficient than Ffar2^−/− cells in producing IL-22 even when stimulated ex vivo with IL-23 or phorbol 12-myristate 13-acetate (PMA)/ionomycin (Figure 6M). These results support the role of FFAR2 in enhancing ILC3-driven IL-22 production to regulate epithelial MHC-II expression during CDI.

To definitively test whether IL-22 produced by FFAR2⁺ ILC3s in response to HFiD and acetate suppresses colonic epithelial MHC-II expression during CDI, we assessed CDI outcomes in WT, Rag2^−/− (lacking T and B cells), and Rag2γc^−/− (lacking all T cells, B cells, and ILCs) mice fed either LFiD or HFiD. HFiD did not protect Rag2γc^−/− mice as effectively as it did for WT mice (Figures S7F and S7G). By contrast, Rag2^−/− mice showed no differences in CDI outcomes across diets, exhibiting results similar to HFiD-fed WT mice but with better recovery compared with LFiD-fed WT mice (Figures S7F and S7G). Furthermore, suppression of epithelial MHC-II in HFiD-fed mice 4 d.p.i. was observed in Rag2^−/− but not Rag2γc^−/− mice (Figures 6N and S7H).

We also examined Ahr^fl/fl Rorc^Cre (Ahr^ΔRORγt) mice, which selectively lack ILC3s, and found they exhibited more severe disease during CDI compared with Ahr^fl/fl mice (Figure 6O). These mice had lower Il22 mRNA levels in the proximal colon 5 d.p.i. (Figure 6P), increased MHC-II expression in CD45⁻ EpCAM⁺ IECs, and a higher frequency of CD4⁺ IELs (Figure 6Q). Additionally, Ahr^ΔRORγt mice had elevated Ifng and reduced Il10 mRNA levels in the proximal colon 5 d.p.i. (Figure 6R). HFiD also failed to protect Ahr^ΔRORγt mice against CDI (Figures S7I–S7K) and did not suppress H2-Ab1 and Ifng expression in the proximal colon 4 d.p.i. (Figure S7L). Overall, these findings conclusively demonstrate that ILC3s are the primary source of IL-22 induced by FFAR2-acetate signaling, which suppresses epithelial MHC-II expression during CDI.

Epithelial expression of HLA-DR and accumulation of CD4⁺ IELs correlate with tissue damage in human CDI

We investigated whether human patients with active CDI exhibit increased epithelial MHC-II (HLA-DR) expression. We analyzed colon sections from patients with active CDI with those from uninfected individuals displaying normal histological features. CDI was associated with significant tissue damage, including barrier disruption, bleeding, inflammatory cell infiltration, goblet cell loss, and edema (Figures 7A and 7B). The frequency of HLA-DR⁺ EpCAM⁺ cells was higher in the inflamed colonic tissue of CDI patients compared with controls (Figures 7C and 7D). We also found a positive correlation between the level of epithelial HLA-DR expression and the severity of tissue inflammation (Figure 7E). This suggests that increased epithelial MHC-II expression is associated with more severe tissue damage during CDI.

Figure 7. Histopathological and immunological analysis of human colon biopsies in uninfected controls and CDI patients.

(A) Representative H&E-stained colon sections from uninfected controls displaying normal histological features (N) and CDI patients (n= 10). Scale bars, 250 μm.

(B) Histopathological scoring of controls (N) and CDI patients (n= 10), evaluated across 10 parameters, each scored from 0 to 3.

(C) HLA-DR and EpCAM staining in colon sections from controls and CDI patients (n = 10). Scale bars, 100 μm. DAPI stains nuclei.

(D) Quantification of HLA-DR⁺ EpCAM⁺ cells in controls vs. CDI patients (n= 10).

(E) Correlation between histopathological score and HLA-DR⁺ EpCAM⁺ cells in CDI (n= 10).

(F) CD4 and EpCAM staining of colon sections from controls and CDI patients (n= 10). Scale bars, 100 μm. DAPI stains nuclei.

(G) Quantification of CD4⁺ cells in controls vs. CDI patients (n= 10).

(H) Relative CD4⁺ cell numbers in EpCAM⁺ cells area (n= 10).

(I and J) Correlation of CD4⁺ cell numbers (I) and CD4⁺ cells in EpCAM⁺ area (J) with histopathological score and number of HLA-DR⁺ EpCAM⁺ cells (n= 10).

Error bars, mean ± SEM. Statistical analysis: t test for comparisons; Pearson’s test for correlations. *p < 0.05, **p < 0.01, ***p< 0.001, ****p < 0.0001.

Given that MHC-II supports CD4⁺ T cell responses, we next assessed whether CDI patients have an increased number of CD4⁺ T cells in their colonic biopsies. Indeed, CDI patients showed a higher total number of CD4⁺ cells in their colonic tissue compared with healthy individuals (Figures 7F and 7G). Additionally, there was a significant increase in CD4⁺ cells within the epithelial layer (Figure 7H). The frequency of CD4⁺ cells, both in the tissue and specifically in the epithelium, was positively correlated with the histopathological score (Figure 7I). Similarly, the frequency of CD4⁺ cells was positively correlated with the expression of HLA-DR by EpCAM⁺ epithelial cells (Figure 7J). Overall, these findings indicate that active CDI in humans is characterized by elevated epithelial HLA-DR expression, which is associated with increased accumulation of CD4⁺ IELs and heightened tissue inflammation and damage.

DISCUSSION

This study highlights how dietary fibers, their microbiota-derived degradation product acetate, and its receptor FFAR2 influence epithelial MHC-II expression and the abundance of pathogenic CD4⁺ IELs during CDI, ultimately affecting infection outcomes. The epithelium of the small intestine naturally expresses MHC-II under normal conditions, playing a crucial role in inducing Tregs and maintaining tolerance to microbiota antigens.^47–49 By contrast, the colonic epithelium only expresses MHC-II during inflammation.^50,51 This study reveals that LFiD selectively increases MHC-II expression in the colonic epithelium compared with SD during CDI, hindering recovery from infection. By contrast, HFiD reduces MHC-II expression and enhances recovery outcomes in CDI. Previous studies that reported MHC-II deficiency did not impact CDI focused only on fecal bacterial load,²¹ potentially overlooking its broader impact on clinical outcomes, such as inflammation and tissue repair.

Our study shows that LFiD increases epithelial MHC-II expression due to a depletion of SCFAs, particularly acetate. Further experiments with FFAR2 conditional knockout mice reveal that acetate attenuates epithelial MHC-II expression through external signaling. Specifically, acetate binds to FFAR2 on ILC3s, prompting them to secrete IL-22, which in turn suppresses MHC-II expression on epithelial cells. These findings are consistent with a recent study showing that IL-22 inhibits MHC-II expression on epithelial cells by regulating the endoplasmic reticulum stress.⁴⁴ Since IFN-γ induces MHC-II expression on the intestinal epithelium,^38,40,41 whereas IL-22 suppresses it, MHC-II levels on IECs appear to be regulated by a dynamic balance between these activating and inhibitory signals. This perspective may help explain previous studies demonstrating that the microbiota and a high-fat diet play crucial roles in regulating intestinal epithelial MHC-II expression.^49,52 In addition to stimulating ILC3 production of IL-22, we find that SCFAs also directly protect IECs from CDI independently of MHC-II. Moreover, SCFAs promote IL-10 production by Tregs consistent with previous research showing that SCFA-producing Clostridia enhance FoxP3⁺ Treg generation and IL-10 production through multiple mechanisms.^53–55

Recent reports have highlighted the role of intestinal epithelial MHC-II in inducing CD4⁺ T cells.^56–60 One of the main discoveries of our study is that enhanced MHC-II expression induced by LFiD leads to the expansion of CD4⁺ IELs that are biased toward producing IFN-γ. This exacerbates inflammation, delaying or preventing resolution of CDI. Interestingly, while early IFN-γ production during CDI in mice is beneficial,¹¹ chronic CDI in patients is associated with elevated IFN-γ levels,^61,62 corroborating that sustained IFN-γ by CD4⁺ IELs may impair CDI recovery. This event is likely to occur in recurrent CDI, where T cells play a more prominent role compared with acute infection.²⁰ Thus, HFiD and SCFA could be especially beneficial in this context. Our analysis of biopsy specimens from individuals with CDI corroborates a positive correlation between epithelial MHC-II expression, the accumulation of CD4⁺ IELs, and tissue damage. Overall, the consistency between clinical histopathology and mouse experimental data presented here may inform preventive dietary strategies and targeted therapies. These approaches could leverage the complex interactions between diet, the epithelial barrier, immune responses, and human susceptibility to C. difficile.

RESOURCE AVAILABILITY

Lead contact

Additional information requests for resources and reagents can be directed to Marco Colonna (mcolonna@wustl.edu).

Materials availability

All material generated in this paper will be made available by the lead contact upon request.

Data and code availability

This paper does not report original code.

Upon request, the lead contact can provide any additional information required to reanalyze the data reported in this paper.

STAR★METHODS

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

C57BL/6J, B6.SJL-Ptprc^a Pepc^b/BoyJ (CD45.1), B6.129S2-H2^dlAb1-Ea/J (H2-Ab1^−/−), B6.129X1-H2-Ab1^b-tm1Koni/J (H2-Ab1 or MHC-II-flox), B6.Cg-Tg(Vil1-cre)1000Gum/J (Villin-Cre), C57BL/6J-Rag1^em10Lutzy/J (Rag1^−/−), and B6.Cg-Rag2^tm1.1Cgn/J (Rag2^−/−) mice were purchased from Jackson Laboratory. C57BL/6NTac.Cg-Rag2^tm1Fwa Il2rg^tm1Wjl (Rag2γc^−/−) mice were purchased from Taconic Biosciences. Ffar2^−/− and Ffar2-flox mice were provided by Prof. Brian Layden at the University of Illinois at Chicago, USA. AhR^fl/fl RORyt-Cre mice were previously described by our group.⁶⁴ All mouse strains were backcrossed to a C57BL/6 background and housed under specific pathogen-free conditions at the animal facility of Washington University in St. Louis. Both male and female mice were used in all experimental groups. Mice were kept in filtered cages with corn cob bedding, provided with standard chow diet and drinking water ad libitum, and housed up to 5 per cage. Cages were cleaned weekly in a laminar airflow cabinet, and mice were maintained on a 12-hour light:12-hour dark cycle. For genotype comparisons, conditional knockout mice were compared to flox/flox Cre⁻ littermate controls. Whole-body knockout mice were bred separately and co-housed with wild-type mice, which were kept in the same facility room, after weaning until the end of the experiment to minimize environmental and microbiome variations. All animal procedures were approved by the Washington University Animal Studies Committee.

Fiber diet regimens and SCFA treatment

For 21 days prior to and throughout the CDI protocol, mice were fed one of three diets: a standard chow diet (SD; Purina PicoLab Rodent Diet 20), a low fiber (LFiD) diet containing 5% cellulose, or a high fiber (HFiD) diet containing 5% cellulose plus 10% soluble inulin. The LFiD and HFiD diets were obtained from Research Diets, Inc. (D10012M and D19071901). In SCFA experiments, mice were maintained on the standard chow diet and given drinking water supplemented with 150 mM of acetate, propionate, and/or butyrate, or left untreated. Supplemented water was replaced every 3–4 days to avoid contamination. The pH of the SCFA solutions was adjusted to 7.2, following previously described methods.³⁵

C. difficile infection

The Clostridioides difficile VPI 10463 strain was cultured on BHI blood agar at 37°C under anaerobic conditions using BD GasPak^™ EZ Anaerobe Container System Sachets with Indicator. Age- and gender-matched mice were pre-treated with an antibiotic mixture (0.4 mg/mL kanamycin, 0.035 mg/mL gentamicin, 0.035 mg/mL colistin, 0.215 mg/mL metronidazole, 0.045 mg/mL vancomycin; Sigma Aldrich) in drinking water for 4 days. After 24 hours without antibiotics, mice received a single intraperitoneal dose of clindamycin (10 mg/kg; Sigma Aldrich) and were infected with 10⁸ CFU of C. difficile spores via oral gavage the following day. Spores were prepared by streaking C. difficile on anaerobic blood agar for 6–8 days to induce sporulation, as previously described.^19,65,66 Mice were weighed daily and assessed for clinical severity using a score from 0 (normal) to 15 (dead) by a blinded evaluator, based on activity level, posture, coat condition, diarrhea, and ocular/nasal discharge, with each parameter rated from 0 to 3.³⁶ Mice were analyzed at days 4 and 5 post-infection, as these time points capture both acute inflammation and the early recovery phase of CDI. Significant differences in disease severity between dietary groups were observed during this period.

Patients biopsies

Colon tissue sections were obtained from 20 human subjects, comprising 10 uninfected individuals (normal) and 10 patients diagnosed with Clostridioides difficile infection (CDI). C. difficile colitis is a relatively common disease. To identify CDI-positive cases for this study, the pathology database at Washington University School of Medicine in St. Louis was searched for C. difficile colitis from 2010 to 2023. Slides from 10 randomly selected colectomy specimens of CDI-positive patients were retrieved and reviewed to confirm the diagnosis. For the control group, colon sections were obtained from histologically normal tissue of 10 uninfected individuals, ensuring no evidence of active gastrointestinal disease in the sampled areas. The study included an equal number of males and females in each group, with participants’ ages ranging from 37 to 85 years and an average age of 57. Detailed participant information, including age and gender, is provided in Table S1. Ancestry, race, ethnicity, and socioeconomic status were not recorded to protect participant privacy, and these factors are not expected to influence the results of this study. The study was approved by the Washington University in St. Louis Institutional Review Board (RB ID number: 202308010), and informed consent was obtained from all participants in compliance with ethical guidelines.

Tissue specimens were fixed in 10% formalin for 24 hours, embedded in paraffin, and sectioned at 4 μm thickness. For CD4 and HLA-DR staining, sections were deparaffinized, rehydrated, and subjected to antigen retrieval using Trilogy solution (Sigma). Comparisons between histologically normal uninfected controls and CDI biopsies were conducted to evaluate the histopathological and molecular alterations associated with CDI. This pilot study utilized a sample size of 20 participants based on similar studies in digestive diseases. Participants were allocated to experimental groups based on their disease status. Power analysis was not performed due to a lack of prior data for effect size estimation.

METHOD DETAILS

Fecal C. difficile burden

Stool samples collected post-infection were weighed and homogenized by vortexing in 1 mL of sterile PBS. After allowing the mixture to settle for 10 minutes, the supernatants were serially diluted up to 10⁻⁶. The dilutions were then plated on cycloserine-cefoxitin-fructose agar (CCFA) supplemented with horse blood and incubated anaerobically at 37°C for 2 days. Abundance of C. difficile in the fecal samples was also assessed by qPCR quantification of the tcdb gene (toxin B) using specific primers. Bacterial load was determined from a standard curve generated with serial dilutions of Escherichia coli genomic DNA targeting the 16S rDNA V4 region.

Fecal lipocalin-2 quantification

Fecal samples were weighed and 50 mg were reconstituted in 1 mL of sterile PBS with 0.1% Tween 20, vortexed for 20 seconds, and centrifuged at 8000 × g for 10 minutes. Lipocalin-2 levels were quantified using a Mouse Lipocalin-2/NGAL ELISA kit (R&D Systems) with a standard curve. Supernatants were diluted 4- to 100-fold, and absorbance was measured at 450 nm.

H&E and Alcian Blue/PAS histologic analysis

Colons were harvested, washed with cold PBS, and opened longitudinally. The tissues were fixed in 4% neutral phosphate-buffered formalin for 24 hours and then transferred to 70% ethanol. After fixation, the tissues were paraffin-embedded, sectioned, and stained with hematoxylin and eosin (H&E) or Alcian Blue/PAS. Slides were digitized using the NanoZoomer-SQ Digital Slide Scanner (C13140–01; Hamamatsu Photonics K.K.). Histopathological analysis of H&E-stained sections was performed using a scoring system from 0 to 30, based on the sum of 10 parameters rated from 0 (normal) to 3 (severe).³⁶ The frequency of goblet cells in Alcian Blue/PAS-stained sections was quantified using a scale from 0 (absence) to 3 (enrichment).³⁶ All measurements were conducted by an investigator blinded to the experimental groups.

Immunohistochemistry (IHC)

TUNEL (Terminal deoxynucleotidyl transferase dUTP Nick End Labeling) and Ki-67 staining were performed on paraffin-embedded tissue sections to detect apoptotic and proliferative cells, respectively. Tissue sections were first deparaffinized in xylene and rehydrated through a graded ethanol series. Following rehydration, the sections were rinsed in phosphate-buffered saline (PBS) and permeabilized with 20 μg/ml Proteinase K for 20 minutes at room temperature. For TUNEL staining, positive controls were generated by treating sections with 1 μg/ml DNase I in PBS for 10 minutes. The TUNEL reaction mixture, prepared according to the manufacturer’s instructions (Roche, Basel, Switzerland), was applied to the sections, followed by incubation in a humidified chamber at 37°C for 60 minutes. After washing with PBS, sections were incubated with a peroxidase-conjugated antibody, and the signal was visualized using diaminobenzidine (DAB) substrate. Similarly, Ki-67 staining was conducted by applying a Ki-67 specific primary antibody, followed by incubation with a secondary antibody conjugated to peroxidase, and visualization with DAB. Sections were counterstained with hematoxylin, dehydrated, mounted, and analyzed under a light microscope. Positive TUNEL and Ki-67 staining appeared as brown nuclei, whereas hematoxylin-stained nuclei appeared blue. The extent of staining was scored on a scale from 0 (normal) to 3 (severe).

Immunofluorescence (IF)

Tissue sections were deparaffinized using xylenes and 2-propanol, followed by antigen retrieval in Trilogy solution (MilliporeSigma) under boiling conditions for 20 minutes. Blocking and permeabilization were conducted in PBS containing 2% BSA and 0.1% Triton X-100 for 1 hour at room temperature. The sections were then incubated overnight at 4°C with primary antibodies, including MHC-II (1:200, Biolegend), EpCAM (1:200, Abcam), and CD4 (1:1000, in-house produced). Following PBS washes, fluorophore-conjugated secondary antibodies (Alexa Fluor 488, 594, 647; Thermo Fisher Scientific) were applied for 1 hour at room temperature. Nuclear staining was achieved using DAPI (1:20,000) for 5 minutes at room temperature. The sections were subsequently mounted with ProLong^™ Glass Antifade Mountant and covered with 1.5 mm cover slips. Fluorescence images were captured using a Zeiss LSM880 Confocal Laser Scanning Microscope with Airyscan, equipped with 405 nm, 488 nm, 543 nm, and 633 nm lasers. Image quantification was performed using cellSens (Olympus) and Fiji, with additional processing in Adobe Photoshop CC and Imaris v8.3 (Bitplane).

Whole-Slide Imaging and Data Analysis

Fluorescently labeled slides were digitized using a 3DHisetch Pannoramic P250 Flash III whole slide scanner (Epredia, Kalamazoo, MI, USA) through Zeiss Plan-Apochromat 20×/0.80 lens. The digital images were then analyzed with custom-written algorithms for a digital pathology software (Visiopharm, 2023.9×63, Broomfield, CO, USA) to determine the number of cells.

FITC-Dextran intestinal permeability

Mice received an oral gavage of FITC-Dextran (4 kDa; Sigma) at 250 mg/kg in 200 μL of sterile PBS. Following the gavage, mice were fasted for 4 hours. Next, mice were anesthetized, and blood samples were collected via cardiac puncture. Serum FITC-Dextran fluorescence intensity was measured using a BioTek Synergy H1 Plate Reader (Vermont, USA) with excitation at 485 nm and emission at 528 nm. A standard curve was generated with serial dilutions of 100 μg/mL FITC-Dextran in PBS.

Bacterial 16S rDNA quantification

Total DNA was extracted from the liver, spleen, and mesenteric lymph nodes of C. difficile-infected mice using the DNeasy Blood & Tissue Kit (Qiagen), following the manufacturer’s instructions. Bacterial DNA levels were quantified by qPCR with iTaq Universal SYBR Green Supermix (Bio-Rad) using primers targeting the 16S rDNA V4 region (Table S2). Bacterial load for both assessments was calculated using a standard curve based on serial dilutions of Escherichia coli genomic DNA and expressed as 16S rDNA gene copies per gram of tissue.

IEL and LP cells isolation

Colon was harvested from mice, longitudinally cut, and washed twice with Hanks’ Balanced Salt Solution (HBSS, ThermoFisher) to remove luminal contents. Intraepithelial lymphocytes were isolated by incubating the tissue with HBSS/HEPES containing 10% bovine calf serum (BCS), 5 mM EDTA, and 1.5 mM DTT at room temperature for two 20-minute washes. For LP isolation, the tissue was then incubated at 37°C with shaking for 40 minutes in complete RPMI medium supplemented with 10% BCS and 1 mg/mL collagenase IV (Sigma). Immune cells were further purified using a 40%/70% Percoll gradient (Cytiva), then washed and filtered through a 35 μm cell strainer.

Intestinal epithelial cells isolation

Colon tissue was dissected, flushed, longitudinally opened, and washed with cold PBS. The tissue was minced into 1–2 cm pieces and incubated on ice in PBS containing 30 mM EDTA and 1.5 mM DTT for 20 minutes. Tissue was then transferred to PBS with 30 mM EDTA and incubated at 37°C for 10 minutes. After manual shaking, the tissue was removed, and the cell solution was centrifuged at 1000 × g for 5 minutes at 4°C. The pellet was resuspended in 10% FBS in PBS and centrifuged again. For single-cell dissociation, the pellet was resuspended in PBS containing 1 mg/mL collagen/dispase and 0.2 mg/mL DNase I, and incubated at 37°C for 10 minutes, with shaking every 2 minutes. The solution was filtered through 70-μm and 40-μm strainers, collected, and centrifuged at 1000 × g for 5 minutes. The final pellet was resuspended in FACS Buffer.

Flow cytometry

Surface staining was performed after blocking Fc receptors with in-house produced anti-CD16/CD32. Conjugated antibodies (eBioscience, BD, BioLegend) were diluted in FACS buffer and applied to the samples on ice for 20 minutes in the dark. Dead cells were excluded using the Zombie Aqua^™ Fixable Viability Kit (BioLegend). For intracellular staining, cells were fixed and permeabilized using either the BD Cytofix/Cytoperm^™ Fixation/Permeabilization Kit for cytokine staining or the eBioscience^™ Foxp3/Transcription Factor Staining Buffer Set for transcription factor staining, following the manufacturer’s protocols. Flow cytometry analysis was conducted on a FACS-Canto^™ or FACS-Symphony^™ using BD FACS-Diva^™ Software (BD Biosciences), and data were analyzed with FlowJo v.9.5.2 (Tree Star).

Quantitative gene expression

Total RNA was extracted from snap-frozen tissue or cells using the RNeasy Mini Kit (Qiagen, Inc.) according to the manufacturer’s protocol. RNA was reverse transcribed to cDNA using the iScript^™ cDNA Synthesis Kit (Bio-Rad) and specific target gene oligonucleotides (Table S2). Quantitative PCR (qPCR) was performed with iTaq Universal SYBR Green Supermix (Bio-Rad). Gene expression levels were quantified using the 2^ΔΔCt method, with Gapdh or B2m as the reference gene (Table S2).

Bone marrow chimera

CD45.2 WT and CD45.2 H2-Ab1^−/− recipient mice were irradiated with a single dose of 11 Gy and reconstituted with 5×10⁶ CD45.1 WT bone marrow cells, administered via retroorbital injection. After 8 weeks, spleen was collected and analyzed to assess the efficiency of hematopoietic reconstitution by quantifying the proportion of CD45.1⁺ and CD45.2⁺ cells. For the mixed BM chimera experiment, CD45.1 WT mice were irradiated (11 Gy) and intravenoulsy injected 1:1 mixture of 3×10⁶ CD45.1/2 WT and 3×10⁶ CD45.2 Ffar2^−/− bone marrow cells, followed by an 8-week reconstitution period.

Caco-2 cell line culture

Caco-2 epithelial cells were maintained in complete RPMI medium (RPMI 1640 with 10% fetal bovine serum, 1% penicillin-streptomycin, and 1% L-glutamine) at 37°C in a 5% CO₂ incubator. Upon reaching confluence, cells were stimulated for 24 hours with 10 ng/mL interferon-gamma (IFNγ; Proteintech) and varying concentrations of short-chain fatty acids (SCFAs): acetate (25 and 50 μM), propionate (5 and 10 μM), and butyrate (1 and 10 μM). After stimulation, cells were harvested for downstream analyses to evaluate the effects of IFNγ and SCFAs on HLA-DR expression.

Intestinal organoid 3D culture

Small intestine or colon organoid cultures were prepared as previously described.⁶⁷ Intestinal crypts were plated in 20 μL of Cultrex Basement Membrane Extract Type 2 (Bio-Techne / R&D Systems) on 24-well plates and cultured in IntestiCult Intestinal Organoid Growth Medium (STEMCELL Technologies) for 6 days, with a medium change on day 3. On day 5, 10 ng/mL of murine interferon-gamma (IFNγ; Proteintech) and 25 μM of acetate were added to the organoid cultures for 12 hours. Following analysis on day 6, organoids were collected using Cell Recovery Solution (Corning), incubated on ice for 30 minutes, and then centrifuged. To dissociate the organoids into single cells, the pellet was incubated at 37°C for 4 minutes in TrypLE solution, followed by vigorous pipetting. The resulting cell suspension was stained with anti-EpCAM and MHC-II antibodies, with dead cells excluded using DAPI.

IL-22 in vivo neutralization

C. difficile–infected mice received an i.p. dose of 150 μg anti-mouse IL-22 neutralizing antibody (clone 8E11; Genentech) or isotype control IgG2a (BioXcell) in 100 μl sterile PBS on days 1 and 3 p.i. Clinical evaluation continued until day 5 p.i.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical analyses were performed using GraphPad Prism version 8.0 (San Diego, CA, USA). Survival curves were generated using Kaplan-Meier estimates and compared using the log-rank test. Normality of data distribution was assessed using the Shapiro-Wilk test, and homogeneity of variances was evaluated with appropriate tests. For normally distributed samples, comparisons between two groups were made using a two-tailed unpaired Student’s t-test. For comparisons involving more than two groups, one-way ANOVA with Tukey’s post hoc test was applied. For non-normally distributed data, Mann-Whitney U tests were used for two-group comparisons, and the Kruskal-Wallis test with Dunn’s post hoc test was used for multiple group comparisons. Pearson’s correlation was used to evaluate relationships between variables. Data are presented as mean ± standard error of the mean (SEM), with statistical significance defined as *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns = not significant. Experimental groups were assigned randomly, and histological analyses were performed in a blinded manner. For murine organoid assays, 2–5 wells per mouse per ex vivo treatment were analyzed. Sample sizes, which were not determined by formal power analysis, are specified in the figure legends.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
B220 BUV737 Clone RA3-6B2	BD Biosciences	Cat# 612838, RRID:AB_2870160
CD11b PE Clone M1/70	Thermo Fisher Scientific	Cat# 12-0112-85, RRID:AB_465549
CD11c APC Clone N418	BioLegend	Cat# 117310, RRID:AB_313779
CD11c APC-Cy7 Clone N418	BioLegend	Cat# 117324, RRID:AB_830649
CD11c BV421 Clone N418	BioLegend	Cat# 117330, RRID:AB_11219593
CD11c PE-Cy7 Clone N418	BioLegend	Cat# 117318, RRID:AB_493568
CD19 PerCP/Cy5.5 Clone 6D5	BioLegend	Cat# 115534, RRID:AB_2072925
CD3 Alexa Fluor 700 Clone 17A2	BD Biosciences	Cat# 561388, RRID:AB_10642588
CD3e PE-CF594 Clone 145-2C11	BD Biosciences	Cat# 562286, RRID:AB_11153307
CD3ε PE-Cy7 Clone 145-2C11	BioLegend	Cat# 100320, RRID:AB_312685
CD3ε FITC Clone 145-2C11	BioLegend	Cat# 100306, RRID:AB_312671
CD3e PE Clone 145-2C11	Thermo Fisher Scientific	Cat# 12-0031-85, RRID:AB_465498
CD3ε PerCP/Cy5.5 Clone 145-2C11	BioLegend	Cat# 100328, RRID:AB_893318
CD4 APC Clone GK1.5	Thermo Fisher Scientific	Cat# 17-0041-83, RRID:AB_469321
CD4 BUV395 Clone GK1.5	BD Biosciences	Cat# 563790, RRID:AB_2738426
CD4 FITC Clone GK1.5	BioLegend	Cat# 100406, RRID:AB_312691
CD4 PE Clone GK1.5	BioLegend	Cat3 100408, RRID:AB_312693
CD4 PE-Cy7 Clone GK1.5	BioLegend	Cat# 100422, RRID:AB_312707
CD45 APC-Cy7 Clone 30-F11	BioLegend	Cat# 103116, RRID:AB_312981
CD45 BUV563 Clone 30-F11	BioLegend	Cat# 612924, RRID:AB_2870209
CD45 PE-Cy7 Clone 30-F11	BioLegend	Cat# 103114, RRID:AB_312979
CD45.1 Alexa Fluor 700 Clone A20	BioLegend	Cat# 110724, RRID:AB_493733
CD45.1 PE-Cy7 Clone A20	BioLegend	Cat# 110730, RRID:AB_1134168
CD45.1 biotin Clone A20	BioLegend	Cat# 110703, RRID:AB_313492
CD45.2 APC-Cy7 Clone 104	BioLegend	Cat# 109824, RRID:AB_830789
CD45.2 PE Clone 104	BioLegend	Cat# 109807, RRID:AB_313444
CD8a FITC Clone 53-6.7	BioLegend	Cat# 100706, RRID:AB_312745
CD8a PE-Cy7 Clone 53-6.7	BioLegend	Cat# 100722, RRID:AB_312761
CD8b FITC Clone YTS156.7.7	BioLegend	Cat# 126606, RRID:AB_961295
CD8b APC-Cy7 Clone YTS156.7.7	BioLegend	Cat# 126619, RRID:AB_2563950
CD90.2 (Thy1.2) APC Clone 30-H12	BioLegend	Cat# 105312, RRID: AB_313183
CD90.2 (Thy1.2) BV 785 Clone 30-H12	BioLegend	Cat# 105331, RRID:AB_2562900
CD326 (EpCAM) APC Clone G8.8	BioLegend	Cat# 118214, RRID:AB_1134102
F4/80 APC Clone BM8	BioLegend	Cat# 123116, RRID:AB_893481
F4/80 Percp-Cy5.5 Clone Bm8	BioLegend	Cat# 123128, RRID:AB_893484
Foxp3 PE Clone 150D	BioLegend	Cat# 320007, RRID:AB_492981
GATA3 AF488 Clone L50-823	BD Biosciences	Cat# 560077, RRID:AB_1645303
Human HLA-DR PE Clone L243 (G46-6)	BD Biosciences	Cat# 560943, RRID:AB_2033987
IL-22 PE Clone 1H8PWSR	Thermo Fisher Scientific	Cat# 12-7221-82, RRID:AB_10597428
KLRG1 Percp-Cy5.5 Clone 2F1/KLRG1	BioLegend	Cat# 138418, RRID:AB_2563015
Ly6C BV421 Clone HK1.4	BioLegend	Cat# 128031, RRID:AB_2562177
Ly6G FITC Clone 1A8	Biolegend	Cat# 127606; RRID:AB_1236488
MHC-II (I-A/I-E) PE-Cy7 Clone M5/114.15.2	BioLegend	Cat# 107630, RRID:AB_2069376
MHC-II (I-A/I-E) Percp-Cy5.5 Clone M5/114.15.2	BioLegend	Cat# 107625, RRID:AB_2191072
NK1.1 BV421 Clone PK136	BD Biosciences	Cat# 562921, RRID:AB_2728688
NK1.1 BV650 Clone PK136	Biolegend	Cat# 108735, RRID:AB_11147949
NK1.1 PerCP/Cy5.5 Clone PK136	Biolegend	Cat# 108728, RRID:AB_2132705
RORgt APC Clone B2D	Thermo Fisher Scientific	Cat# 17-6981-82, RRID:AB_2573254
T-bet BV421 Clone 04-46	BD Biosciences	Cat# 563318, RRID:AB_2687543
TCR γ/δ PE Clone GL3	BioLegend	Cat# 118108, RRID:AB_313832
Anti-human HLA-DP/DR/DQ Clone PdV5.2	Santa Cruz Biotechnology	Cat# sc-130013, RRID:AB_2012533
MHC-II (I-A/I-E) Alexa Fluor 647 Clone M5/114.15.2	BioLegend	Cat# 107618, RRID:AB_493525
Anti-Human CD4 Antibody, Clone OKT4	American Type Culture Collection (ATCC)	Cat# crl-8002, RRID:AB_2073240
Rabbit polyclonal anti-EpCAM	Abcam	Cat# ab71916; RRID:AB_1603782
Goat anti-Mouse IgG2b (γ2b), Alexa Fluor 488	Thermo Fisher Scientific	Cat# A-21141; RRID: AB_141626
Donkey anti-Rabbit IgG (H+L), Alexa Fluor Plus 594	Thermo Fisher Scientific	Cat# A-32754; RRID:AB_2762827
Goat anti-Mouse IgG1 (γ1), Alexa fluor 647	Thermo Fisher Scientific	Cat# A-21240; RRID:AB_2535809
Rabbit monoclonal anti-Ki-67	Thermo Fisher Scientific	Cat# MA5-14520, RRID:AB_10979488
Anti-mouse IL-22 antibody Clone 8E11	Genentech	Cat# 8E11.9, RRID:AB_2651129
InVivoMAb rat IgG2a isotype control	Bio X Cell	Cat# BE0089, RRID:AB_1107769
Experimental models: Cell lines
Caco-2 colorectal adenocarcinoma cells	American Type Culture Collection (ATCC)	Cat# HTB37
Experimental models: bacteria strains
Clostridioides difficile strain VPI 10463	American Type Culture Collection (ATCC)	Cat# 43255-FZ
Escherichia coli	American Type Culture Collection (ATCC)	Cat# 10798
Experimental models: Mouse strains
C57BL/6J	Jackson Laboratories	Strain# 000664
B6.SJL-Ptprca Pepc b/BoyJ (CD45.1)	Jackson Laboratories	Strain# 002014
B6.129S2-H2dlAb1-Ea/J (H2-Ab1−/−)	Jackson Laboratories	Strain# 003584
B6.129X1-H2-Ab1b_tm1Koni/J (H2-Ab1flox)	Jackson Laboratories	Strain# 013181
C57BL/6J-Rag1em10Lutzy/J (Rag1−/−)	Jackson Laboratories	Strain# 034159
B6.Cg-Tg(Vil1-cre)1000Gum/J (Villin-Cre)	Jackson Laboratories	Strain# 021504
B6.FVB-Tg(Rorc-cre)1Litt/J (RORgt-Cre)	Jackson Laboratories	Strain# 022791
B6.Cg-Rag2tm1.1Cgn/J (Rag2−/−)	Jackson Laboratories	Strain# 008449
C57BL/6NTac.Cg-Rag2tm1Fwa Il2rgtm1Wjl (Rag2 γc−/−)	Taconic Biosciences	Strain# 4111-M
Ffar2−/−	Prof. Brian Layden	N/A
Ffar2 flox	Prof. Brian Layden	N/A
AhR flox	This paper	N/A
Chemicals, peptides, recombinant proteins and others
0.05% Trypsin-EDTA	Gibco	Cat# 25300-054
10X red blood cells (RBC) lysis buffer	BioLegend	Cat# 420302
123count eBeads	ThermoFisher	Cat# 01-1234
1M HEPES Buffer	Corning	Cat# 25-060-Cl
2-propanol	MilliporeSigma	Cat# 190764
4’,6-Diamidino-2-Phenylindole Dihydrochloride (DAPI)	MilliporeSigma	Cat# D9542
Acetic acid	Sigma-Aldrich	Cat# A6283-1L
Agar	MilliporeSigma	Cat# A7921
Ampicillin sodium salt	Sigma-Aldrich	Cat# A9518
BD GasPak™ EZ Anaerobe Container System Sachets with Indicator	BD	Cat# 260001
Beta-Mercaptoethanol	Sigma	Cat# M-7522
Bovine Serum Albumin (BSA)	Rockland	Cat# BSA-1000
Brain Heart Infusion (BHI) agar	Sigma-Aldrich	Cat# 70138-500G
Brain Heart Infusion Broth	Sigma-Aldrich	Cat# 53286-500G
Butyric acid	Sigma-Aldrich	Cat# B103500-1L
Cell recovery solution	Corning	Cat# 354253
Clindamycin hydrochloride	Sigma-Aldrich	Cat# C5269
Colistin sulfate salt	Sigma-Aldrich	Cat# C4461
Collagenase-IV	Sigma-Aldrich	Cat# C4-BIOC
Collagenase/Dispase	Roche	Cat# 10269638001
Cultrex Basement Membrane Extract, Type 2	Bio-Techne / R&D Systems	Cat# 3532-010-02
Cycloserine-Cefoxitin-Fructose-Agar (CCFA)	AnaeroGRO	Cat# AG501
DL-Dithiothreitol solution (DTT)	Sigma-Aldrich	Cat# 646563
DNase I	Roche	Cat# 11284932001
EDTA 0.5M	Corning	Cat# 46-034-Cl
Fluorescein isothiocyanate-dextran (FITC-dextran)	MilliporeSigma	Cat# 46944
Formalin solution, neutral buffered, 10%	MilliporeSigma	Cat# HT501128
Gentamicin sulfate	Sigma-Aldrich	Cat# G1914
Glutamax (100x)	Gibco	Cat# 35050-061
Hoechst 33258	Invitrogen	Cat# H3569
Horse Blood	Colorado Serum Company	Cat# 30004
Inulin-enriched diet	Research Diets, Inc.	Cat# D19071901
Ionomycin calcium salt	Sigma-Aldrich	Cat# I0634
Kanamycin disulfate salt from Streptomyces kanamyceticus	Sigma-Aldrich	Cat# K1876
Kanamycin Sulfate (100x)	Gibco	Cat# 15160-054
Low fiber diet	Research Diets, Inc.	Cat# D10012M
MEM Nonessential amino acids (100x)	Corning	Cat# 25-025-Cl
Metronidazole	Sigma-Aldrich	Cat# M3761
Mouse IL-23	R&D	Cat# 1887-ML
Neomycin trisulfate salt hydrate	Sigma-Aldrich	Cat# N1876
Nuclease-free water	Invitrogen	Cat# AM9937
Paraformaldehyde (32%)	Electron Microscopy Sciences	Cat# 15714
PBS (1x)	Leinco	Cat# P364
PenStrep	Gibco	Cat# 15140-122
Percoll	Cytiva	Cat# 17089101
Phorbol 12-myristate 13-acetate (PMA)	Sigma-Aldrich	Cat# P8139
ProLong Glass Antifade Mountant	Life Technologies	Cat# P36930
Propionic acid	Sigma-Aldrich	Cat# 402907-500ML
RPMI 1640	Sigma	Cat# R8758-1L
SAV BUV661	BD	Cat# 612979
Sodium Pyruvate 100mM	Corning	Cat# 25-000-Cl
Trilogy	MilliporeSigma	Cat# 920P
Triton X-100	Sigma	Cat# T8787
TrypLE Express	Gibco	12605-010
Vancomycin hydrochloride from Streptomyces orientalis	Sigma-Aldrich	Cat# V2002
Xylenes	MilliporeSigma	Cat# 247642
Yeast Extract	Sigma-Aldrich	Cat# 92144-500G-F
Critical commercial assays
RNeasy Plus Mini Kit	QIAGEN	Cat# 74134
IntestiCult Intestinal Organoid Growth Medium	STEMCELL Technologies	Cat # 06005
LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit, for 405 nm excitation	Life Technologies	Cat# L34957
Mouse Lipocalin-2 (NGAL) DuoSet ELISA Kit	R&D	Cat# DY1857-05
Foxp3 transcription factor staining Buffer set	eBioscience	Cat# 00-5523-00
BD Cytofix/Cytoperm Plus	BD	Cat# 555028
QIAwave DNA Blood & Tissue Kit	Qiagen	Cat# 69556
PureLink™ Microbiome DNA Purification	ThermoFisher	Cat# A29790
Oligonucleotides
qPCR primers, see Table S1	This paper	N/A
Software and algorithms
GraphPad Prism	GraphPad	https://www.graphpad.com/features
Zen (Black edition)	Carl Zeiss Microscopy GmbH	https://www.micro-shop.zeiss.com/en/us/softwarefinder/software-categories/zen-black/
FlowJo (v10)	TreeStar	https://www.fiowjo.com/solutions/fiowjo/
Photoshop	Adobe	https://www.adobe.com/products/catalog.html
Illustrator	Adobe	https://www.adobe.com/products/catalog.html
ImageJ/Fiji	Schindelin et al.63	https://imagej.net/software/fiji/ RRID: SCR_002285

Highlights.

• An LFiD exacerbates C. difficile infection by enhancing MHC-II in colonic epithelium

• Epithelial MHC-II expands pathogenic CD4⁺ intraepithelial lymphocytes in CDI

• Dietary fibers and SCFA downregulate epithelial MHC-II, alleviating CDI

• Fiber-derived acetate promotes IL-22 by ILC3s, which inhibits epithelial MHC-II