Butyrate suppresses mucosal inflammation in inflammatory bowel disease primarily through HDAC3 inhibition in monocytes and macrophages

Abstract

Butyrate‐producing gut bacteria and luminal butyrate levels are reduced in Inflammatory Bowel Diseases (IBDs). Butyrate has anti‐inflammatory properties through mechanisms not well‐characterized in IBDs. Here, we determined the butyrate anti‐inflammatory effect on primary IBD tissues and intestinal cell models to identify key target cells and pathway(s) involved. Cytokines, monocarboxylate transporter‐1 (MCT1), G‐protein‐coupled receptor‐109A (GPR109A), and histone deacetylase‐3 (HDAC3) levels were analyzed in IBD and healthy tissues using cytometric bead arrays, RNA‐seq analysis and immunofluorescence. Inflammatory markers and phagocytosis in butyrate‐treated colonic organoids, primary monocytes or THP‐1 macrophages, were assessed by qPCR, flow cytometry and amikacin protection assays, when relevant combined with GPR109A or HDAC3 antagonists. Butyrate suppressed TNF and IL‐6 secretion by > 50% in ex vivo‐cultured inflamed IBD biopsies. MCT1 expression was reduced in inflamed epithelium and cytokine‐exposed organoids, while IL‐18 was reduced 0.5‐fold in organoids, and both were restored by butyrate, without suppressing pro‐inflammatory gene expression. GPR109A and HDAC3 were elevated in IBD tissues and upregulated by butyrate in cultured mucosa. Butyrate also suppressed IL‐6, TNF‐α, CD40, and CD80 by > 50% and enhanced adherent‐invasive Escherichia coli (AIEC) phagocytosis by 62% in monocytes/macrophages. Histone acetylation (H3K9ac) increased > 5‐fold, mimicking the HDAC inhibitor SAHA. Contrary, specific GPR109A inhibition and gene G‐protein‐coupled receptor inhibition did not alter butyrate's effects. Butyrate restores MCT1 and IL‐18 gene expression in inflamed epithelial cells, showing limited anti‐inflammatory effects. Instead, butyrate targets HDAC3 in mononuclear cells, suppressing inflammation in IBD gut mucosa. The cell‐type‐specific effects of butyrate offer mechanistic insights that support its therapeutic relevance in IBDs.

Butyrate mitigates intestinal inflammation in IBD. It does so primarily by suppressing pro‐inflammatory cytokines and costimulatory molecules in activated monocytes and macrophages, concomitant with enhancing pathogen clearance. In human colonic organoids, butyrate promotes epithelial expression of key genes involved in barrier function and immune signaling, indicating its dual action in restoring gut homeostasis.

Abbreviations

AIEC

adherent‐invasive Escherichia coli

CD

Crohn's disease

CD40/CD80

cluster of differentiation 40 or 80

CM

cytokine mix

GPR109A

G‐protein‐coupled receptor 109A

H3K9ac

acetylation of lysine 9 on histone H3

HDAC3

histone deacetylase 3

IBDs

inflammatory bowel diseases

IFN‐γ

interferon‐γ (gamma)

IL

interleukin

LP

lamina propria

LPS

lipopolysaccharide

MCT1

monocarboxylate transporter 1

SCFA

short‐chain fatty acid

TNF

tumor necrosis factor

UC

ulcerative colitis

Introduction

Inflammatory bowel diseases (IBD), such as, Crohn's disease (CD) and ulcerative colitis (UC), show an increasing incidence and prevalence worldwide over the past decades [1]. IBD is characterized by chronic inflammation in the gastrointestinal tract, with alternating periods of activity and remission, driven by an exacerbated immune response and disruptions to the epithelial barrier, arising from a complex interplay of genetic, microbial, and environmental factors [2]. In particular, the infiltration of TNF‐α‐producing cells, such as inflammatory monocytes and M1 macrophages, into the lamina propria damages the intestinal epithelial barrier [3]. Therefore, reducing the inflammatory profile of these innate immune cells is crucial to sustaining remission in IBD.

A dysregulated gut microbiome, known as dysbiosis, is a relevant factor in IBD, marked by an increase in Enterobacteriaceae (e.g., E. coli) and a decrease in Firmicutes (e.g., F. prausnitzii), among others [4, 5]. The reduction in butyrate‐producing Firmicutes and butyrate luminal levels can disturb gut mucosal homeostasis. Butyrate, a short‐chain fatty acid (SCFA) produced by microbial fermentation of dietary fibers, supports gut homeostasis by enhancing epithelial integrity and modulating immune responses [6]. It strengthens the intestinal barrier by serving as an energy source for colonocytes and promoting tight junction assembly, mucus production, and antimicrobial peptide expression [7, 8, 9]. Butyrate also protects against cytokine‐induced increases in permeability [10] and exerts anti‐inflammatory effects by modulating the activity of innate immune cells [11, 12, 13].

Consequently, the enrichment of butyrate‐producing bacteria was explored as a treatment for IBD, using pre‐ and probiotics, synbiotics, and fecal microbiota transplantation (FMT), which may suppress clinical symptoms in UC patients [14, 15].

Cellular uptake of butyrate occurs through passive diffusion or active transport mediated by the monocarboxylate transporter 1 (MCT1) and sodium‐coupled monocarboxylate transporter 1 (SMCT1). The expression of SLC16A1 (encoding MCT1) and butyrate‐oxidation related genes is reduced in the inflamed gut mucosa of IBD patients [16, 17, 18]; however, it remains unclear which mucosal cell types are most affected. Additionally, butyrate activates G protein‐coupled receptors 109A (GPR109A, encoded by HCAR2), 43 (GPR43), and 41 (GPR41), inhibiting adenylate cyclase or activating phosphoinositide phospholipase C (by Gαi or Gαq proteins), respectively. Although mucosal GPR109A's role in inhibiting macrophage activation and promoting tolerogenic macrophages and dendritic cells (DCs) has been demonstrated in murine models [12, 19], its expression and function have not yet been thoroughly investigated in IBD patients. Furthermore, GPR109A signaling maintains gut epithelial integrity by increasing IL‐18 levels in colonic epithelium [12] and serum [20] in mouse colitis models. Among these butyrate transporters and receptors, MCT1 and GPR109A have a higher affinity for butyrate compared to other SCFAs (e.g., acetate and propionate) [21, 22].

Butyrate also modulates gene expression through epigenetic modifications, enhancing histone acetylation by inhibiting histone deacetylases (HDACs) [23] or activating histone acetyltransferases (HATs) [24]. Altered histone acetylation has been observed in IBD, with either increased histone 4 acetylation [25] or decreased histone 3 acetylation [26], alongside decreased HDAC levels, including epithelial HDAC3 mRNA levels in CD and UC mucosa [27, 28].

Thus, three main mechanisms—MCT1‐mediated butyrate uptake, GPR109A signaling, and HDAC‐mediated histone modification—may contribute to butyrate's anti‐inflammatory actions. Nevertheless, it remains unclear which mechanism is the primary target in IBD or in which cell type it is most relevant.

In this study, we elucidate the effects of butyrate on inflammatory mediators, as well as MCT1, GPR109A, and HDAC3 expression in ex vivo‐cultured gut mucosa of IBD patients and in vitro human colon epithelial and mononuclear IBD models.

Results

Butyrate reduces exacerbated levels of secreted cytokines in intestinal mucosa of IBD patients

Reducing pro‐inflammatory cytokine production in the intestinal mucosa is a key therapeutic objective in IBD. To assess the modulatory effect of butyrate treatment, we measured cytokine levels in supernatants from ex vivo‐cultured intestinal biopsies from patients with CD (n = 9), UC (n = 9), and HC (n = 10) (Fig. 1). As expected, biopsies from inflamed regions of CD and UC patients produce higher TNF levels during 24 h of culturing compared to flanking non‐inflamed mucosa (CD: 5.7‐fold, P = 0.019; UC: 2.6‐fold, P = 0.041) and mucosa from HC (CD: 3.7‐fold, P = 0.021; UC: 2.9‐fold, P = 0.027). Butyrate treatment (2 mm), used as an approximation of physiological mucosal exposure [29], significantly reduced TNF levels by 61% in inflamed UC biopsies (P = 0.039), with similar trends observed in inflamed tissue from CD patients and HC, showing 56% and 31% reductions, respectively (Fig. 1A).

Fig. 1.

Butyrate reduces type 1, type 2, and type 17 cytokine levels in ex vivo‐cultured mucosa media from healthy controls and inflammatory bowel disease patients. The effect of butyrate (2 mm) or dexamethasone treatment (100 nm) for 24 h on cytokine secretion: TNF (A), IL‐6 (B), IL‐17A (C), IFN‐γ (D), IL‐2 (E), and IL‐10 (F) was evaluated in cultured biopsies from healthy controls (HC, n = 10), non‐inflamed (NI), and inflamed (I) biopsies from patients with inflammatory bowel disease (IBD), including Crohn's disease (CD, n = 9), and ulcerative colitis (UC, n = 9). Cytokine analysis was performed on the culture medium after 24 h using a cytometric bead array (CBA). Statistical analysis: Mann–Whitney test (A–D, F) and unpaired t‐test (E) (black: between HC and IBD); Wilcoxon matched‐pairs signed‐rank test (A–D, F) and paired t‐test (E) (green: comparing NI and I tissue in patients with IBD; blue: between control and butyrate or dexamethasone treatment). *P < 0.05; **P < 0.01.

Likewise, biopsies from inflamed tissue from CD and UC patients secreted more IL‐6, compared to flanking non‐inflamed tissue (5.8‐fold, P = 0.014 and 2.8‐fold, P = 0.004, respectively) and HC (10.4‐fold, P = 0.004 and 5.9‐fold, P = 0.002, respectively). Butyrate treatment reduced IL‐6 secretion by 51% in inflamed mucosa from CD patients (P = 0.049), with a similar trend observed in the flanking non‐inflamed tissue (52% reduction, P = 0.054) (Fig. 1B). Moreover, IL‐17A secretion was higher in UC inflamed mucosa (3.4‐fold, P = 0.014, compared to HC), but was not suppressed by butyrate. Interestingly, butyrate did suppress IL17A secretion in non‐inflamed mucosa of UC patients (P = 0.021), although values are in the lower part of the detection limit (Fig. 1C). No prominent effects of butyrate treatment were observed on IFN‐γ and IL‐2 levels, which were not significantly enhanced under basal conditions, except for higher IFN‐γ levels in inflamed versus non‐inflamed mucosa of CD patients (3.2‐fold, P = 0.036) (Fig. 1D,E).

Interestingly, inflamed mucosa from CD patients produced more IL‐10—an anti‐inflammatory cytokine—compared to the flanking non‐inflamed mucosa (4.3‐fold, P = 0.017), which may reflect an inflammation‐resolution process. However, butyrate suppressed IL‐10 levels in inflamed mucosa from CD patients (81% reduction, P = 0.007), as well as in both inflamed (81%, P = 0.013) and non‐inflamed (86%, P = 0.017) mucosa from UC patients (Fig. 1F).

Notably, treatment of IBD and HC biopsies with dexamethasone, a synthetic glucocorticoid, led to reductions in TNF and IL‐6 levels similar to those observed with butyrate (Fig. 1A,B). However, a notable difference was observed for IL‐10, as butyrate reduced IL‐10 secretion, whereas dexamethasone had no significant effect (Fig. 1F). These findings suggest that although both compounds suppress pro‐inflammatory cytokines through partially overlapping anti‐inflammatory pathways, their divergent effects on IL‐10 point to distinct regulatory mechanisms as well.

Taken together, these results show an immune‐regulatory effect of butyrate on human intestinal tissue, especially by decreasing inflammatory cytokine production in the mucosa of IBD patients.

Differential expression of MCT1, GPR109A, and HDAC3 in mucosa of IBD patients

To identify potential cellular targets underlying the anti‐inflammatory effects of butyrate in the intestinal mucosa, we analyzed the expression, localization, and cell‐type‐specific distribution of key butyrate‐regulated molecules, MCT1 (a butyrate transporter), GPR109A (a butyrate‐responsive receptor), and HDAC3 (butyrate‐inhibited enzyme) in non‐inflamed and inflamed intestinal tissues from IBD patients.

First, bulk RNA‐sequencing data of colonic biopsies from HC, CD, and UC patients [30] showed a reduced expression of SLC16A1 (encoding MCT1) in inflamed tissue of CD and UC patients compared to the non‐inflamed tissue and HC tissue (P < 0.001). Remarkably, the non‐inflamed mucosa from CD or UC patients expressed higher levels of SLC16A1 compared to HC (P < 0.001) (Fig. 2A, left panel). HCAR2 and HDAC3 levels were also increased in the non‐inflamed and inflamed colonic mucosa of IBD patients compared to HC (P < 0.05). Furthermore, HCAR2 levels were enhanced in inflamed versus non‐inflamed tissue of IBD patients (P < 0.0001), and similarly, HDAC3 levels were elevated in inflamed compared to non‐inflamed UC tissue (P < 0.01) (Fig. 2A, middle and right panels).

Fig. 2.

MCT1, GPR109A, and HDAC3 expression in intestinal mucosa of healthy controls and inflammatory bowel disease patients. (A) Gene expression of SLC16A1, HCAR2, and HDAC3 from bulk RNA‐sequencing data [30] in colon biopsies of healthy controls (HC n = 48), non‐inflamed (NI) and inflamed (I) mucosa from patients with Crohn's disease (CD; NI = 162, I = 73), and ulcerative colitis (UC; NI = 145 and I = 121). The Y‐axis in the graphs represents normalized gene counts, obtained through trimmed mean of M‐values normalization method of raw counts from high‐throughput sequencing. These normalized values are used to depict relative gene expression levels, enabling comparison across samples. Data are presented as medians and interquartile ranges. Statistical analysis: Kruskal–Wallis test. *P < 0.05; **P < 0.01, ***P < 0.001; ****P < 0.0001. (B) Representative immunofluorescence images and quantification of MCT1, GPR109A and HDAC3 distribution in the epithelium and lamina propria (LP) (image resolution: 0.50 μm per pixel). Positive immunoreactivity is indicated by white arrows for colonic epithelium and white arrowheads for lamina propria (LP) mononuclear cells, evaluated in mucosa areas from HC (n = 3), and NI and I mucosa from CD and UC patients (n = 3 per group) from the Chilean research center. Images for MCT1 and GPR109A were zoomed from a 20× magnification, while images for HDAC3, were zoomed from a 40× magnification (scale bar: 50 μm).

An immunofluorescent microscopy of intestinal tissues revealed that MCT1 was predominantly detected at a basolateral localization in surface epithelial cells of the colon mucosa in HC and non‐inflamed tissues of CD and UC patients (Fig. 2B, arrows in upper panels). Moreover, MCT1‐specific staining was observed in immune cells within the lamina propria (LP) of non‐inflamed tissues (Fig. 2B, arrowheads in upper panels). Consistent with SLC16A1 expression, MCT1 protein levels were reduced in epithelial and immune cells in inflamed tissue of IBD patients (Fig. 2B upper panels). GPR109A protein was detected in epithelium and predominantly enriched in immune cells within the lamina propria of inflamed tissue from UC patients (Fig. 2B middle panels; arrowheads pointing to GPR109A staining in immune cells). HDAC3 protein was detected in various mucosal cells of HC and IBD samples, with levels appearing increased in IBD tissues (Fig. 2B lower panels). Quantification of MCT1, GPR109A, and HDAC3 levels in the epithelium and LP compartments of the colon mucosa, performed in a small cohort of HC and IBD patients (n = 3 per group), revealed inter‐individual variability but showed a trend towards reduced MCT1 and increased GPR109A expression in the epithelium and LP of inflamed UC tissues, as well as increased HDAC3 levels in inflamed regions (Fig. 2B, bottom section).

In agreement with these findings, publicly available single‐cell RNA‐seq data from colon mucosa derived from HC, CD, and UC patients [31, 32] (Fig. 3) reveal that SLC16A1 was predominantly expressed in the epithelial compartment, including various cell types such as enterocyte subtypes, goblet cells, stem cells, and immature enterocytes. Its expression was reduced in inflamed tissue compared to non‐inflamed regions in both CD and UC datasets, with a more pronounced decrease observed in UC samples. HCAR2 expression was primarily detected in immune cells, with the highest levels found in dendritic cells and, particularly, monocytes, from inflamed CD tissue. HDAC3 was expressed in both epithelial and immune cell populations. Within the epithelium, elevated expression was observed in cycling cells and stem cells from inflamed regions of both CD and UC, compared to non‐inflamed samples and healthy controls. Among immune populations, HDAC3 was predominantly expressed in monocytes, dendritic cells, and cycling B and T cells. Remarkably, HDAC3 expression was enhanced in monocytes from inflamed CD tissues relative to non‐inflamed regions. In contrast, UC samples showed modestly higher HDAC3 expression in non‐inflamed mucosa compared to both inflamed tissue and healthy controls.

Fig. 3.

SLC16A1, HCAR2, and HDAC3 expression in colon epithelial and immune cells. Dot plots of single‐cell RNA‐seq (scRNA‐seq) data show genes in columns and cells in rows. The dot size represents the percentage of cells expressing the gene, and the color indicates the scaled mean expression. Healthy controls (HC), Crohn's disease (CD), ulcerative colitis (UC), and non‐inflamed (NI) or inflamed (I) mucosa are indicated as specified in the datasets. ScRNA‐seq datasets for SLC16A1, HCAR2, and HDAC3 are shown in (A, C) for colon epithelial cells and (B, D) for colon immune cells. Data were obtained from the studies of (A, B) Kong et al. (colon tissue from CD patients and HC) [31] and (C, D) Smillie et al. (colon tissue from UC patients and HC) [32].

Given these disease‐ and cell‐type‐specific expression patterns, we next examined whether butyrate modulates the expression of SLC16A1, HCAR2, and HDAC3 in a manner consistent with its proposed anti‐inflammatory role. To this end, we examined the expression in a small cohort of ex vivo‐cultured colonic biopsies treated with butyrate. Similar to the RNA‐seq data, except for SLC16A1 (Fig. 4A), HCAR2 levels were increased in inflamed mucosa of IBD patients compared to HC (8.5‐fold, P = 0.012) and to flanking non‐inflamed mucosa (3.7‐fold, P = 0.026). Butyrate treatment induced HCAR2 in inflamed mucosa (3‐fold, P = 0.020) (Fig. 4B). Similarly, butyrate increased HDAC3 in inflamed mucosa (1.5‐fold, P = 0.017) and showed a trend towards induction in non‐inflamed tissue (1.4‐fold, P = 0.063) (Fig. 4C).

Fig. 4.

Regulation of SLC16A1, HCAR2, and HDAC3 expression by butyrate in ex vivo‐cultured biopsies. mRNA levels of SLC16A1 (A), HCAR2 (B) and HDAC3 (C) were measured in ex vivo‐cultured biopsies from colonic mucosa of healthy controls (HC) (n = 3) and non‐inflamed (NI, n = 6) and inflamed (I, n = 8) inflammatory bowel disease (IBD) tissues, determined by TaqMan RT‐qPCR. Statistical analysis: Wilcoxon matched‐pairs signed‐rank test for NI vs. I tissue and control vs. butyrate treatment (B, C); Mann–Whitney test for HC vs. IBD (B). *P < 0.05.

These findings show that MCT1, GPR109A, and HDAC3 are differentially expressed in the mucosa of healthy individuals and IBD patients, with MCT1 predominantly found in epithelial cells, GPR109A in mononuclear cells, and HDAC3 in both cell types. Furthermore, butyrate modulates HCAR2 and HDAC3 gene expression in ex vivo‐cultured biopsies.

Butyrate regulates MCT1 and IL‐18 mRNA expression in human colonic organoids

As the intestinal epithelium is the first tissue layer exposed to luminal butyrate, and MCT1 mediates its cellular uptake, we used human crypt‐derived colonic organoids to investigate how butyrate regulates MCT1 expression and to assess its potential anti‐inflammatory effects in a physiologically relevant epithelial model of healthy and IBD‐like conditions.

Four to six independent human colonic organoid lines were cultured in differentiation medium (DM) favoring enrichment of mature epithelial cell types. As expected, the marker of stem cells (LGR5) was suppressed, while the marker for enterocytes (intestinal alkaline phosphatase, ALPI) was increased (Fig. 5A). Intestinal organoids were then exposed to increasing butyrate concentrations (0.2–20 mm) to reflect both luminal and mucosal exposure levels, in the presence or absence of a cytokine mix to simulate the inflammatory environment of IBD (5 ng·mL⁻¹ CM: TNF‐α, IL‐1β, and IFN‐γ) for 6 h. Treated organoids did not exhibit clear morphological changes (Fig. 5B), remained viable, and did not show significant signs of cell death by necrosis, as reflected by LDH release (Fig. 5C). Cytokine exposure reduced SLC16A1 expression to ~ 0.4‐fold compared to control (P = 0.0004), which was counteracted by butyrate which increased SLC16A1 expression to 1.3‐ and 1.9‐fold at 2 and 20 mm, respectively (Fig. 5D; P = 0.008 and P = 0.016). In line with the observed mRNA expression, butyrate (2 mm) also tended to increase MCT1 protein expression in non‐inflammatory and inflammatory conditions. However, the reduction in SLC16A1 gene expression observed after 6 h of cytokine mix exposure was not reflected at the protein level at the same time point, likely due to the delayed kinetics of protein synthesis (Fig. 5E).

Fig. 5.

Butyrate induces MCT1 and IL‐18 mRNA expression in human colonic organoids without affecting inflammatory mediators and tight junction gene expression. (A) mRNA levels of colonic organoid cell markers cultured in expansion medium (EM) or differentiation medium (DM) (n = 3). (B) Bright‐field images of colonic organoids cultured in DM and treated with 2 mm butyrate in the presence or absence of an inflammatory cytokine mix (CM) (5 ng·mL⁻¹ of TNF‐α, IL‐1β, and IFN‐γ), for 6 h (objective 10×, scale bar: 250 μm). (C) Lactate dehydrogenase (LDH) release from organoids was measured in conditioned medium following treatments (n = 5 or n = 6). (D) mRNA levels of SLC16A1 (n = 6) and (E) representative image and quantification of MCT1 protein expression (n = 3, objective 40×, scale bar: 100 μm, quantification of image resolution: 0.30 μm per pixel) were determined by TaqMan RT‐qPCR and immunofluorescence in PFA‐fixed/paraffin‐embedded organoids, respectively. (F) mRNA levels of NOS2, TNFA, IL1B, CXCL8, IRF1, TFF1, DUOX2, LCN2, IL18, OCLN, CLDN3, and TJP1 (n = 4 or n = 6) were determined by RT‐qPCR. Statistical analysis: Paired t‐test (A, D), F (except NOS2) and Wilcoxon matched‐pairs signed‐rank (F, NOS2). *P < 0.05; **P < 0.01; ***P < 0.001.

Gene expression of IBD‐associated pro‐inflammatory markers (NOS2, TNFA, IL1B, CXCL8, IRF1, TFF1, DUOX2, and LCN2) and pro‐healing genes (IL18 and tight junction components OCLN, CLDN3, and TJP1) was evaluated in colonic organoids. Despite strong organoid line‐specific variability, cytokine treatment robustly increased NOS2, IL1B, CXCL8, and IRF1 mRNA expression by more than 10‐fold. Butyrate only suppressed NOS2 and IRF1 expression in cytokine‐treated organoids and did so only at the highest concentration of this SCFA (20 mm) (Fig. 5F). Cytokine exposure significantly decreased mRNA levels of IL18 and tight junction components (OCLN, CLDN3, TJP1) to approximately 0.5–0.6‐fold of control levels. Interestingly, cotreatment with butyrate (2 and 20 mm) reversed the cytokine‐induced suppression of IL18, increasing its expression to levels exceeding 1.3‐fold relative to control. Moreover, butyrate also enhanced IL18 expression under non‐inflammatory conditions (20 mm butyrate; P = 0.0232; Fig. 5F). IL6 mRNA was undetectable in colonic organoids under both non‐inflammatory and inflammatory conditions.

These findings indicate that in primary intestinal epithelial cells, butyrate induces the expression of its transporter SLC16A1 and IL18 under both basal and inflammatory conditions, while exerting limited effects on cytokine‐induced pro‐inflammatory and epithelial barrier genes, except for reducing NOS2 and IRF1 expression at high concentrations.

Butyrate decreases pro‐inflammatory phenotype of inflammatory macrophages and monocytes

Given that monocytes and macrophages are key contributors to the inflammatory response in IBD, we next explored the potential anti‐inflammatory effects of butyrate on LPS + IFN‐γ‐treated human THP‐1‐derived macrophages as a model of M1‐like macrophages that mimic key features of intestinal inflammation. The combination of LPS (a bacterial component) and IFN‐γ (an immune‐derived cytokine) increased surface expression of the costimulatory molecules CD40 and CD80 by 7.5‐ and 2.5‐fold, respectively. Butyrate treatment suppressed this induction in a dose‐dependent manner, with significant reductions observed at concentrations between 0.2 and 2 mm (30–70% for CD40 and 16–56% for CD80), and near‐complete suppression at 20 mm (82% and 69% for CD40 and CD80, respectively) (Fig. 6A,B). Interestingly, butyrate reduced the phagocytic activity of LPS + IFN‐γ‐treated macrophages towards a commensal E. coli strain by 33% (HS; P = 0.031), while enhancing phagocytosis of an IBD‐associated adherent‐invasive E. coli (AIEC) strain by 62% (CD2‐a; P = 0.034; Fig. 6C). These results show that butyrate reduces the pro‐inflammatory phenotype of human macrophages, while more selectively clearing IBD‐associated pathogenic gut bacteria. To investigate genes related to phagocytic macrophage activity, we evaluated the transcript levels of CD48 and MARCKS. We observed that butyrate reduced the levels of CD48 under both basal and inflammatory contexts to 0.2‐fold (P = 0.008 and P = 0.0006), with non‐significant differences for MARCKS (Fig. 6D). This pattern is consistent with reduced commensal uptake, but does not align with the increased phagocytic behavior associated with CD2‐a. These findings suggest that butyrate may downregulate phagocytosis‐related genes, particularly affecting commensal bacteria, while AIEC uptake likely involves alternative receptors or pathways not dependent on CD48.

Fig. 6.

Butyrate reverts inflammatory mediators and enhances adherent‐invasive Escherichia coli (AIEC) phagocytosis in human THP‐1 macrophages. THP‐1 macrophages (n = 3) were treated with butyrate in the presence or absence of lipopolysaccharide (LPS) + IFN‐γ (0.1 + 20 ng·mL⁻¹) for 24 h. (A) CD40 and (B) CD80 geometric mean fluorescence intensity (gMFI) in THP‐1‐macrophages (n = 3 or n = 5) were determined by flow cytometry. (C) Intracellular E. coli HS (commensal) and CD2‐a (AIEC) strains, normalized to control, were phagocytosed by THP‐1 macrophages treated with 2 mm butyrate (HS: n = 6; CD2‐a: n = 7). (D) mRNA levels of CD48 and MARCKS (n = 3) were determined by RT‐qPCR. Statistical analysis: One‐way ANOVA with Bonferroni post‐test (A, B) and Wilcoxon matched‐pairs signed‐rank test (C, D). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

As shown in Fig. 3, both monocytes and macrophages exhibit high HDAC3 and HCAR2 (GPR109A) expression in samples from healthy controls (HC) and IBD patients. Additionally, butyrate modulates the expression of these genes in ex vivo mucosal tissue (Fig. 4). To further explore whether this regulation is concentration‐dependent, we treated primary human monocytes and THP‐1‐derived macrophages with low (0.2 mm), medium (2 mm), and high (20 mm) concentrations of butyrate in the presence or absence of LPS + IFN‐γ, modeling both non‐inflammatory and inflammatory conditions.

We observed that HCAR2 expression was highly donor‐dependent but strongly induced by LPS + IFN‐γ up to 85‐fold in monocytes (Fig. 7A) and 19‐fold in THP‐1 macrophages (Fig. 7B). Under non‐inflammatory conditions, butyrate also induced HCAR2 in a concentration‐dependent manner in THP‐1 macrophages (Fig. 7B).

Fig. 7.

Butyrate induces GPR109A and HDAC3 mRNA expression and reduces IL‐6 and TNF‐α mRNA levels in human inflammatory monocytes in a GPR109A‐independent manner. mRNA levels of HCAR2 and HDAC3 were determined by RT‐qPCR in human peripheral blood monocytes (A, C) and THP‐1 macrophages (B, D) (n = 4 or n = 6) treated with butyrate in the presence or absence of lipopolysaccharide (LPS) + IFN‐γ (0.1 + 20 ng·mL⁻¹) for 6 h. Human peripheral blood monocytes were treated with butyrate (2 mm) in the presence or absence of LPS + IFN‐γ (0.1 + 20 ng·mL⁻¹) for 6 h, with or without pretreatment of 100 μm mepenzolate bromide (MB) or Pertussis toxin (PT), for 1 h. mRNA levels of (E) IL6 and (F) TNFA (n = 3 or n = 4) were determined by RT‐qPCR. Statistical analysis: One‐way ANOVA with Bonferroni post‐test (B–D); Wilcoxon matched‐pairs signed‐rank test (A); and paired t‐test (E, F). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

In contrast, HDAC3 mRNA expression was significantly upregulated by butyrate at both 2 and 20 mm concentrations, under both inflammatory and non‐inflammatory conditions. This increase ranged from 2.5‐ to 5.9‐fold in monocytes (Fig. 7C) and from 1.7‐ to 2.7‐fold in THP‐1 macrophages (Fig. 7D). Notably, HDAC3 expression was not affected by LPS + IFN‐γ alone (Fig. 7C,D).

In summary, our data indicate that under inflammatory conditions, butyrate upregulates HCAR2 and HDAC3 expression in a manner that plateaus beyond 2 mm, suggesting a nonlinear, threshold‐like response.

Pharmacological antagonists of GPR109A and HDAC3 were used to investigate their potential role in butyrate's anti‐inflammatory effect in these innate immune cells. Notably, gene expression of the cytokines2 IL‐6 and TNF‐α was increased in LPS + IFN‐γ‐treated primary human monocytes, reaching > 1700‐fold and > 50‐fold increases, respectively (Fig. 7E,F). Cotreatment with butyrate reduced IL‐6 expression by 94% (P = 0.043) and TNF‐α expression by 69% (P = 0.0096) relative to LPS + IFN‐γ treatment alone. Next, these experiments were performed in the presence of the GPR109A inhibitor mepenzolate bromide (MB) or the G (αi/o) PCR inhibitor pertussis toxin (PT). Neither MB nor PT affected the butyrate‐mediated suppression of IL6 and TNFA in LPS + IFN‐γ‐treated human monocytes (Fig. 7E,F), demonstrating that its effect is independent of GPR109A and potentially other butyrate‐sensing GPCRs such as GPR41 and GPR43.

As an alternative mechanism, butyrate may inhibit HDAC3 activity, thereby suppressing the induction of inflammatory cytokines. To investigate this, we evaluated the acetylation of histone H3, a substrate of HDAC3, in THP‐1 macrophages and monocytes treated with or without LPS + IFN‐γ, and cotreated with butyrate or SAHA, a class I HDAC inhibitor. We found that the acetylation of lysine 9 on histone H3 (H3K9ac) was similarly increased by both butyrate and SAHA in THP‐1 macrophages, with fold changes ranging from 5 to 9.8 (Fig. 8A). Additionally, H3K9ac was induced by butyrate and SAHA in primary monocytes, with fold‐changes ranging from 5.7‐ to 8.5‐fold, both in the absence (P = 0.0005 for butyrate and SAHA) and in the presence of inflammatory conditions (P = 0.0004 and P = 0.0017 for butyrate and SAHA, respectively) (Fig. 8B). Consistent with HDAC3 inhibition, IL‐6 secretion induced by LPS + IFN‐γ was elevated by approximately 150‐fold compared to control and was reduced by both butyrate (P = 0.0009) and SAHA (P = 0.0006) to 15‐ and 28‐fold, respectively, representing reductions of 90% and 81% relative to LPS + IFN‐γ alone. A similar trend was also observed for the effects of butyrate and SAHA on TNF secretion (Fig. 8C,D).

Fig. 8.

Butyrate induces histone acetylation in THP‐1 macrophages and monocytes and reduces IL‐6 content in human inflammatory monocytes similar as the HDAC3 antagonist SAHA. Acetylation on histone H3 lysine 9 (H3K9ac) was determined intracellularly by flow cytometry in (A) THP‐1 macrophages (n = 3) and in (B) human peripheral blood monocytes (n = 5) treated with butyrate or SAHA in the presence or absence of lipopolysaccharide (LPS) + IFN‐γ for 6 h. Secreted IL‐6 (C) and TNF (D) levels by human monocytes (n = 5) were analyzed using cytometric bead array (CBA). Statistical analysis: Wilcoxon matched‐pairs signed‐rank test and paired t‐test *P < 0.05; **P < 0.01; ***P < 0.001.

In summary, our results from IBD ex vivo and in vitro models demonstrate that butyrate supports intestinal immune and epithelial regulatory mechanisms under basal conditions and mitigates inflammation‐associated perturbations, consistent with a role in maintaining intestinal homeostasis. While butyrate exerted anti‐inflammatory effects in IBD mucosa primarily by modulating immune cell responses, its impact on intestinal epithelial cells was more limited. It significantly reduced IL‐6 and TNF secretion, two key pro‐inflammatory cytokines, in patient‐derived biopsies and modulated monocyte and macrophage activation and phagocytic activity. In contrast, in epithelial organoids, butyrate primarily enhanced mRNA expression of its transporter SLC16A1 and induced IL18, with minimal impact on inflammatory gene expression. Mechanistically, butyrate acted independently of GPR109A and involved HDAC3 inhibition, resulting in increased histone acetylation and suppression of cytokine expression.

Discussion

In this study, we demonstrate that butyrate plays a crucial role in suppressing excessive cytokine secretion in ex vivo gut mucosal tissue from IBD patients, while reducing pro‐inflammatory markers in macrophages and monocytes in vitro. This anti‐inflammatory effect is linked to the ability of butyrate to induce histone acetylation (H3K9ac), indicative of HDAC3 inhibition. Furthermore, we show that butyrate enhances mRNA expression of the butyrate transporter MCT1 and IL‐18 in non‐inflamed and inflamed intestinal epithelial cells, which may increase its availability, supporting its role as an energy source and in epithelial repair.

Butyrate plays a key role in maintaining intestinal homeostasis. In IBD, where butyrate levels and its microbial producers are reduced, we hypothesized that restoring this metabolite could help re‐establish mucosal balance. To investigate butyrate's immunomodulatory effects under both non‐inflamed and inflamed conditions, we analyzed patient‐derived mucosal biopsies and cellular in vitro IBD models using epithelial cells and monocytes/macrophages exposed to defined inflammatory stimuli (cytokine mix or LPS + IFN‐γ). Using physiologically relevant concentrations of butyrate, we observed specific responses in maintaining and restoring a non‐inflammatory state after it was disrupted by inflammatory stimuli. Multiple health‐promoting effects of this microbial metabolite have been reported in in vitro and animal models [11, 33, 34]. Here, we analyzed the direct effect of butyrate on inflamed gut mucosal tissue of IBD patients and identified macrophages and monocytes as relevant players in its anti‐inflammatory actions.

Since butyrate suppressed inflammation in ex vivo‐cultured biopsies from IBD patients, we explored whether it regulates pro‐inflammatory mediators (e.g., IL‐6 and TNF‐α) in primed‐mononuclear cells exposed to inflammatory conditions, mimicking resident and/or infiltrating macrophages in IBD mucosa (exhibiting a classically activated M1 phenotype) [3]. To replicate these conditions in vitro, we activated peripheral mononuclear cells with the TLR4 agonist LPS and IFN‐γ, two central orchestrators of the innate and adaptive immune response. We first determined that, through pharmacological inhibition, GPR109A does not participate in the suppressing effect of butyrate on IL‐6 and TNF‐α mRNA expression. However, both butyrate and SAHA (a HDAC inhibitor) induce H3K9ac and reduce IL‐6 secretion, consistent with findings in bone marrow‐derived macrophages (BMDM) [11]. Moreover, our results show that butyrate suppresses costimulatory molecule CD40/CD80 expression, confirming earlier studies [35]. These effects of butyrate were accompanied by enhanced phagocytic activity towards IBD‐associated pathogenic E. coli (AIEC) strain. Compared to the commensal E. coli strain HS, the CD2‐a AIEC strain harbors a greater number of virulence genes, including papC, chuA, ratA, and pduC, which have been associated with increased epithelial cell invasion and enhanced intracellular survival within macrophages [36, 37]. Furthermore, the presence or expression of these virulence genes in AIEC strains can directly modulate bacterial susceptibility to phagocytosis [38, 39]. We observed that butyrate stimulation reduced CD48 transcript levels, a change that may be related to the phagocytic activity towards the commensal strain HS, but not with the uptake of AIEC CD2‐a. These findings suggest that butyrate may differentially affect phagocytosis depending on the bacterial strain. Notably, AIEC phagocytosis has been reported to require activation of Proline‐rich tyrosine kinase 2 (PYK2) [40], indicating that an alternative CD48‐independent mechanism may mediate AIEC uptake. The specific receptors and signaling pathways involved in the butyrate‐dependent phagocytosis activation, and how they differ between commensal and AIEC recognition, remain to be elucidated. Furthermore, bacterial viability and the inflammatory context are likely important factors influencing butyrate's immunomodulatory effects and should be explored in future studies.

Notably, human monocytes differentiated into macrophages in the presence of butyrate exhibit enhanced antimicrobial activity (mediated by HDAC3 inhibition) [34], highlighting its regulatory role in canonical phagocytic functions.

Similarly, treatment with the HDAC inhibitors (valproic acid and SAHA) ameliorated DSS‐ and TNBS‐induced colitis, promoting a local hyperacetylation of H3 in lamina propria mononuclear cells (LPMC) and decreased IL‐6 and IFN‐γ production [41]. Interestingly, valproic acid reversed hypoacetylation in H3 (H3K27ac) and decreased IL‐6 levels in active IBD mucosa [26]. These findings demonstrate that HDAC inhibitors are valuable treatment options for IBD due to their anti‐inflammatory properties. In fact, a clinical trial is currently underway to assess the safety of using vorinostat (HDAC inhibitor SAHA) for treatment of moderate‐to‐severe CD, in combination with Ustekinumab (anti‐IL12/IL23) maintenance therapy (ClinicalTrials.gov ID NCT03167437).

As anticipated, when examining the entire gut mucosa, TNF, IL‐6, IFN‐γ, and IL‐17A were increased in the supernatants of ex vivo‐cultured inflamed mucosa from active UC and/or CD patients, without revealing a distinct cytokine profile for either condition. Interestingly, 24‐h butyrate treatment suppressed TNF, IL‐6, and IL‐17A production by ex vivo‐cultured intestinal biopsies from both inflamed and non‐inflamed mucosa of IBD patients. These cytokines are central to the initiation and progression of inflammation in IBD [42], mainly driven by NFκB activation, which is inhibited by butyrate [29, 35, 43].

Remarkably, IL‐10 secretion was elevated in inflamed mucosa of CD patients compared to their non‐inflamed counterpart, possibly contributing to mucosal regeneration [44]. These elevated IL‐10 levels were counteracted by butyrate, consistent with previous findings [35], while butyrate‐inhibited IL‐10 production in LPS‐primed human monocytes [13] and human monocyte‐derived macrophages (hMDMs) [45]. Moreover, in the presence of LPS, butyrate‐mediated NLRP3 activation and IL‐1β release in hMDMs [45], suggesting that butyrate's anti‐inflammatory effects depend on the inflammatory context.

Interestingly, the anti‐inflammatory effects of dexamethasone on TNF and IL‐6 parallel those of butyrate. Dexamethasone, like clinically used corticosteroids such as prednisone and budesonide, acts through the glucocorticoid receptor to inhibit pro‐inflammatory signaling pathways, including NF‐κB [46]. In contrast to butyrate, dexamethasone did not alter IL‐10 secretion by IBD mucosa, suggesting that both compounds modulate immune responses; they do so through partially overlapping yet distinct pathways.

Butyrate did not downregulate pro‐inflammatory mediators in primary intestinal organoids, as was observed in monocytes and macrophages. Specifically, low butyrate concentrations (2 mm) were more effective in reducing inflammation in gut mucosa, monocytes, and macrophages, whereas a higher concentration (20 mm) was required to reduce NOS2 and IRF1 levels in organoids. This suggests that butyrate may be preferably used as an energy source in epithelial cells rather than modulating inflammatory pathways. This involves positive regulation of its own transporter (SLC16A1/MCT1) as described earlier [47]. Therefore, the anti‐inflammatory effect on whole mucosa is likely due to HDAC3 inhibition‐mediated responses in mononuclear cells and independent of GPR109A.

Nevertheless, we detected GPR109A mostly in inflamed UC mucosa, being induced by inflammation in macrophages, as reported earlier [19]. GPR109A activation induces anti‐inflammatory functions in mouse macrophages and DCs [12]. These findings suggest that, despite HDAC inhibition being the primary regulatory mechanism in human mononuclear cells, inflammation‐induced GPR109A overexpression may enhance butyrate's effects in immune cells. Accordingly, the GPR109A receptor also senses niacin (nicotinic acid, a form of vitamin B3), an essential human nutrient that promotes mucosal healing and reduces inflammation [12].

Inflamed mucosa of patients with IBD exhibits very low content of MCT1, as previously reported [16, 17], and observed in RNA‐seq analyses [30]. MCT1 mRNA and protein levels were similarly reduced in colonic organoids exposed to cytokines but were dose‐dependently restored by butyrate cotreatment. Reduced MCT1 expression by inflammatory cytokines may hamper efficient butyrate uptake as an energy source for the inflamed epithelium, thus compromising its critical barrier function. Indeed, chronic and acute inflammation downregulate MCT1 mRNA and protein expression, as shown in other epithelial cell models [16, 17]. Interestingly, our results reveal for the first time that butyrate restores SLC16A1 expression in human colonic organoids under inflammatory conditions, highlighting its potential relevance for IBD. Additionally, butyrate restored IL18 expression, which was suppressed under inflammatory conditions. This finding aligns with previous reports demonstrating IL18 induction by the butyrate‐producing bacterium Faecalibacterium prausnitzii [48] and earlier observations showing butyrate‐mediated activation of the IL18 promoter in intestinal epithelial cells (IECs) [49]. IL‐18 plays a pivotal role in maintaining intestinal tolerance [50], supporting epithelial barrier integrity and host defense against infections [51], and facilitating responses to intestinal injury [52]. However, IL‐18 has a dual function: While it exerts protective, anti‐inflammatory effects during the early phase of DSS‐induced colitis, it can adopt a pro‐inflammatory role as the disease progresses [53].

In this context, our model—using 6 h of cytokine exposure—likely reflects an acute or early inflammatory phase. Nevertheless, a more comprehensive evaluation of butyrate's regulatory effect on IL‐18 in mucosal inflammation would benefit from the inclusion of additional immune cell types and microbial components. Overall, our findings suggest that butyrate may contribute to epithelial protection and repair, in part through IL‐18 induction. However, butyrate did not reverse the cytokine‐mediated suppression of tight junction gene expression (OCLN, CLDN3, and TJP1). Furthermore, the subcellular localization of tight junction proteins is crucial for understanding how inflammation affects barrier integrity, as seen in IBD and other diseases, and how butyrate has a potential protective role, as previously demonstrated in human cell lines [7, 8]. In this scenario, human colonic monolayers or 3D organoids (exposed to longer incubation periods or higher inflammatory stimuli concentration than in our study), show that butyrate is ineffective or has even detrimental effects in counteracting inflammation‐induced changes in metabolic‐associated genes or barrier integrity [16, 54]. Therefore, future research must expand on time and concentration parameters to better understand the most physiologically relevant conditions for IBD models.

The use of IBD‐derived organoids may provide a more representative system, as they retain disease‐relevant transcriptional signatures, such as dysregulation of genes involved in antimicrobial defense and epithelial function (e.g., LYZ, CLCA1, AQP8) [55]. However, several studies have reported a gradual loss of the inflammatory phenotype during extended in vitro culture, likely due to the absence of immune and microbial signals [56, 57]. To overcome this limitation, inflammation can be experimentally induced in both IBD‐ and non‐IBD‐derived organoids using pro‐inflammatory stimuli, including the cytokine cocktail used in our study (TNF‐α, IL‐1β, and IFN‐γ) and others [56, 57]. Notably, similar SLC16A1 expression [16] and epithelial responses to butyrate have been observed in both IBD and non‐IBD colonic organoids, across different culture formats [16, 54]. Thus, although disease‐specific differences, such as those between Crohn's disease and ulcerative colitis, are important, inflammation‐induced organoids derived from non‐IBD controls provide a robust and informative model to study epithelial responses to butyrate.

Within the limitations of our study, particularly the limited sample sizes and non‐overlapping datasets, we were not able to formally assess the relationships between MCT1, GPR109A, and HDAC3 with cytokine levels in non‐inflamed and inflamed tissue, nor how butyrate modulates these relationships at the individual level. Additionally, although evaluating the impact of concurrent medications in IBD patients would be valuable, our sample size was insufficient to allow statistically robust comparisons. Although there are inherent limitations, the integration of patient samples and cell‐specific models enhances the translational relevance of our findings and provides novel insight into butyrate's immunomodulatory potential in IBD.

In conclusion, butyrate exerts a potent inhibitory effect on exacerbated cytokine production in IBD mucosa, while displaying anti‐inflammatory effects in monocytes and macrophages through histone acetylation (H3K9ac) and HDAC3 inhibition. Moreover, butyrate enhances MCT1 and IL‐18 mRNA expression in epithelial cells, collectively supporting its potential benefit for IBD therapy.

Our findings, along with prior studies, support the anti‐inflammatory potential of butyrate in intestinal models, reinforcing its relevance for future research, including clinical trials. Clinical outcomes in butyrate‐treated IBD patients have been inconsistent [9], likely due to differences in formulation, delivery route, and host‐related factors. For instance, no consistent clinical or biochemical improvements were observed in UC patients treated with butyrate enemas [9, 58] or in pediatric IBD patients receiving oral butyrate in pH‐sensitive capsules [59]. In contrast, promising results have been reported with targeted formulations, such as lipophilic microcapsules, which maintained remission and modulated the microbiota [60, 61]. These findings underscore the importance of colon‐specific delivery strategies and promoting in situ butyrate production. Given the altered microbiota and pH environment in IBD, therapeutic approaches combining butyrate with prebiotics, probiotics, synbiotics, or fecal microbiota transplantation may be more effective.

Materials and methods

Patients

Adult patients with active IBD (CD and UC) and healthy controls (HC) who attended the Gastroenterology Service at Clínica Las Condes (CLC) and the Department of Gastroenterology and Hepatology at University Medical Center Groningen (UMCG) were consecutively enrolled (between 2017 and 2023) (Table 1). The diagnosis of IBD was established based on clinical, endoscopic, histologic, and radiological criteria [62]. Disease activity in CD and UC patients was determined using the Simple Endoscopic Score for Crohn's Disease (SES‐CD) and the Mayo Endoscopic Sub‐score [63], respectively, including patients with SES‐CD ≥ 7 and Mayo Endoscopic Sub‐score 2–3. Healthy controls were individuals undergoing colorectal cancer screening colonoscopies who were found to be free of cancer or polyps. Patients receiving antibiotic therapy or following a low‐FODMAPs diet, as well as those diagnosed with non‐classifiable IBD, indeterminate colitis, infectious ileocolitis, asthma, atopic dermatitis, autoimmune diseases, celiac disease, or hypertension were excluded. Demographic and clinical variables were collected from the institutional IBD Registry (including age, gender, disease duration, Montreal Classification, and current IBD treatment), using REDCap electronic data capture tools hosted at Clínica Las Condes.

Table 1.

Clinical and demographic characteristics of patients included in the study for CBA analysis. Montreal classification: CD: A, Age at diagnosis (years): A1, < 17; A2, 17–40; A3, > 40. L, Location: L1, Terminal ileal ± limited caecal disease; L2, Colonic; L3, Ileocolonic; L4, Isolated upper disease. B, Behavior: B1, Nonstricturing, nonpenetrating; B2, Stricturing; B3, Penetrating. UC: E, Extension: E1, Ulcerative proctitis; E2, Left‐sided UC (distal to splenic flexure); E3, Extensive (proximal to splenic flexure). S, Severity S0: Clinical remission; S1, Mild UC; S2, Moderate UC; S3, Severe UC.

	HC	CD	UC
Number, n	10	9	9
Gender, n (F/M)	6/4	6/3	6/3
Age, years (median, range)	54.5 (37–69)	32 (20–61)	32 (17–42)
Disease Duration, years (median, range)	–	3 (0–7)	5 (1–17)
Treatment, n
5‐ASA		2	6
Corticosteroids	–	2	2
Immunosuppressants		4	2
Biological therapy		1	0
Without treatment		1	1
Gut sections
Noninflamed
i/ac/tc/	0/0/0/	0/4/0/	0/7/0/
dc/sc/r	0/10/0	0/2/3	0/1/1
Inflamed
i/ac/tc/	–	1/4/0/	0/0/0/
dc/sc/r		1/2/1	1/6/2
Montreal classification in CD
A1/A2/A3		0/8/1
L1/L2/L3/L4	‐	1/5/3/0	‐
B1/B2/B3		9/0/0
Montreal classification in UC
E1/E2/E3	–	–	1/2/6
S0/S1/S2/S3			0/0/6/3

Adult patients undergoing surgery for colorectal cancer or intestine donors for transplant at the Department of Gastroenterology and Hepatology at University Medical Center Groningen, who agreed to participate, were included for the collection of healthy colon tissue from the flanking regions of surgical resection specimens. These tissues were used for crypt‐derived colonic organoid generation and culture.

Mucosa and blood samples

Biopsies from inflamed (I) and non‐inflamed (NI) mucosa of CD (n = 9) and UC (n = 9) patients with active disease, as well as biopsies from the sigmoid mucosa of HC (n = 10) enrolled in CLC, Chile, were ex vivo cultured to determine cytokine levels in conditioned media after butyrate treatment using Cytometric Bead Array (CBA). Cytokine concentrations in conditioned medium were directly compared between various treatment groups as reported by us and others [29, 35, 64], as the size and protein content of biopsies taken for projects in our centers are highly consistent.

All biopsies were cultured for the same duration (maximum 24 h during which viability is preserved [65]) and under identical conditions, thereby minimizing potential variability in viability across samples.

Additionally, GPR109A, MCT1, and HDAC3 expression were detected in paraformaldehyde‐fixed/paraffin‐embedded intestinal mucosa by immunofluorescence (IF) and in cryopreserved biopsies by RT‐qPCR from patients enrolled in CLC in Chile and UMCG in the Netherlands, respectively.

Peripheral blood from healthy controls (HC) from Chile and The Netherlands was used for monocyte isolation and ex vivo culture. After butyrate treatment, gene expression was assessed by RT‐qPCR, cytokine levels in conditioned media using Cytometric Bead Array (CBA), and histone acetylation (H3K9) by flow cytometry.

Ethics

Written consent was obtained from all participants prior to sample collection. This study was conducted according to the Declaration of Helsinki principles, approved by the Local IRB of Clínica Las Condes (Santiago, Chile, certified 16/10/2017) and the Medical Ethics Committee of the University Medical Center Groningen (UMCG, Groningen, The Netherlands, Parelsnoer RB No. 2008/338 and GEID RB No. 2016/424 protocols).

Treatment conditions

For butyrate effect experiments, 2 mm sodium butyrate (Sigma‐Aldrich, Saint Louis, MO, USA) was added to the culture medium, while 100 nm dexamethasone (Steraloids, Inc., Newport, RI, USA) was used as a positive control, as previously described in ex vivo models [29, 64].

In addition, a range of butyrate concentrations (0.2–20 mm) was used in colonic organoid, monocyte, and THP‐1 macrophage cultures, either alone or in co‐incubation with inflammatory stimuli, based on earlier studies [11, 33]. For the monocyte and THP‐1 macrophage experiments, inflammatory stimuli included 0.1 ng·mL⁻¹ lipopolysaccharide (LPS) (E. coli strain O111:B4; Sigma‐Aldrich, Saint Louis, MO, USA) and 20 ng·mL⁻¹ IFN‐γ (R&D Systems, Inc., Minneapolis, MN, USA), incubated for 6 or 24 h, following a modified protocol from a previous report [66]. In the organoid experiments, to simulate the inflammatory environment of IBD, a cytokine mix (CM) containing 5 ng·mL⁻¹ of each cytokine (hTNF‐α, hIFN‐γ, and hIL‐1β; R&D Systems, Inc.) was used for 6 h at concentrations consistent with those secreted by intestinal immune cells. These conditions induced key pro‐inflammatory markers while preserving organoid viability, enabling the assessment of regulatory mechanisms such as butyrate‐mediated modulation.

For GPR109A inhibition or G (αi/o) protein‐coupled receptors (GPCRs) inhibition, human monocytes were pretreated with 100 μm mepenzolate bromide or 10–50 ng·mL⁻¹ pertussis toxin (both from Merck KGaA, Darmstadt, Germany) for 1 h before co‐incubation with LPS + IFN‐γ and butyrate for 6 h. For HDAC inhibition, human monocytes and THP‐1 macrophages were treated with 10 μm SAHA‐Vorinostat (Axon Medchem, Groningen, The Netherlands) for 6 h.

Explant culture

Intestinal tissue biopsies (~ 1 mm³) were placed in 24‐well plates (one biopsy per well) and cultured at 37 °C and 5% CO₂ in serum‐free Advanced Dulbecco's modified Eagle's medium/F12 (DMEM/F12; Gibco‐Life Technologies, NY, USA), supplemented with 50 μg·mL⁻¹ gentamicin (Sigma‐Aldrich, Saint Louis, MO, USA), 200 μg·mL⁻¹ normocin™ (InvivoGen, San Diego, CA, USA), and 0.02 μg·mL⁻¹ fungizone (Gibco, Thermo Fisher Scientific, Waltham, MA, USA). Butyrate and dexamethasone treatments were performed as described in the treatment conditions. After 24 h, supernatants were collected and stored at −80 °C until further cytokine measurement.

Organoid culture

Human intestinal organoid cultures were established from crypt‐derived resection tissues isolated from patients undergoing surgery for colorectal cancer or intestinal transplantation (n = 6) as previously described [67], with minor modifications. Briefly, 5 mm³ tissue sections were washed with cold Hanks' balanced salt solution (HBSS, without calcium and magnesium) pH 7.4 (1×) (Gibco‐Life Technologies, Landsmeer, The Netherlands). Afterward, the tissue was incubated with TrypLE Express (Gibco, Landsmeer, The Netherlands) and 10 nmol·L⁻¹ Y27632 (Sigma‐Aldrich, Zwijndrecht, The Netherlands) for 30 min at 37 °C to release crypts and epithelial cells. The cells were then resuspended and washed twice with DMEM/F12 medium (Thermo Fisher, MA, USA), supplemented with 1× antibiotic/antimycotic, 1× GlutaMAX (both from Thermo Fisher), 10 mmol·L⁻¹ HEPES (Sigma‐Aldrich), and 50 μg·mL⁻¹ gentamicin (Gibco‐Life Technologies), before being pelleted and embedded in Matrigel (Corning, Bedford, MA, USA). The crypts were seeded in 10 μL drops (three drops per well in a 24‐well plate) and cultured in an expansion medium (EM): DMEM/F12 medium supplemented with 50% homemade Wnt‐3a conditioned medium (L‐Wnt3A cells were a kind gift of Dr R.G.J. Vries, Hubrecht Institute, Utrecht, The Netherlands), 100 ng·mL⁻¹ R‐spondin‐1 (R&D Systems, Abingdon, UK), 100 ng·mL⁻¹ Noggin (R&D Systems), 1× B27 (Gibco), 1.25 mmol·L⁻¹ N‐acetylcysteine (Sigma‐Aldrich), 50 ng·mL⁻¹ EGF (Gibco), 10 mmol·L⁻¹ nicotinamide (Sigma‐Aldrich), 10 μmol·L⁻¹ SB202190 (Sigma‐Aldrich), 10 μmol·L⁻¹ Y‐27632 (Sigma‐Aldrich) and 500 nmol·L⁻¹ A83 (Tocris Bioscience, Minneapolis, USA). One week after plating, 100 μg·mL⁻¹ primocin (Life Technologies Landsmeer, The Netherlands) was added to the culture medium. In some experiments, colonic organoids (n = 3) were cultured in a modified EM (EM+) without SB202190 and supplemented with 100 ng·mL⁻¹ IGF‐1 and 50 ng·mL⁻¹ FGF‐2 (both from STEM CELL, Köln, Germany), as previously described [68]. To induce epithelial differentiation, intestinal organoids were cultured for 4 days in a differentiation medium (DM), which had the same composition as EM/EM+, but lacked Wnt3‐conditioned medium, nicotinamide, SB202190, Y‐27632, IGF‐1 and FGF‐2.

Peripheral blood monocytes culture

Human CD14+ monocytes were isolated from peripheral blood mononuclear cells (PBMCs) using human anti‐CD14 microbeads and MS columns (MACS, Miltenyi Biotec, Bergisch Gladbach, Germany). Briefly, PBMCs were obtained from 20 to 30 mL whole blood samples collected in EDTA‐containing tubes (BD Vacutainer) using Lymphoprep density gradient (STEM CELL, Köln, Germany) and washed in buffer composed of 1 mm EDTA (Sigma‐Aldrich, Zwijndrecht, The Netherlands), 10% RPMI 1640 medium, and PBS pH 7.4 (1×) (both from Gibco‐Life Technologies, Landsmeer, The Netherlands). The PBMCs were then incubated with anti‐CD14 microbeads (15 min, 4 °C) to positively select monocytes via magnetic separation using MS columns attached to the MACS separator (MACS, Miltenyi Biotec).

To explore the effects of butyrate on gene and protein expression, isolated human monocytes were cultured in RPMI 1640 medium supplemented with 10% heat‐inactivated fetal bovine serum (FBS), 100 U·mL⁻¹ penicillin, and 100 μg·mL⁻¹ streptomycin (all from Gibco‐Life Technologies, NY, USA) at 37 °C in 5% CO₂‐humidified air.

Cell line culture

The human THP‐1 monocytic cell line (RRID:CVCL_0006) (ATCC^® TIB‐202™) was cultured in RPMI 1640 medium supplemented with 10% heat‐inactivated fetal bovine serum (FBS), 100 U·mL⁻¹ penicillin, and 100 μg·mL⁻¹ streptomycin (all from Gibco‐Life Technologies, NY, USA), and 5 μm β‐mercaptoethanol (Sigma‐Aldrich, Saint Louis, MO, USA), and maintained at 37 °C in 5% CO₂‐humidified air. For macrophage differentiation, we followed a 72‐h protocol as previously described [69], and an inflammatory (M1) profile was induced using LPS + IFN‐γ as outlined in the cell culture conditions. The experiments were performed with mycoplasma‐free cells.

Bacterial strains and culture conditions

Two E. coli strains were grown for a phagocytosis assay (amikacin protection assay) in THP‐1 macrophages. The E. coli CD2‐a strain was previously isolated from ileal biopsies of a patient with Crohn's disease, presenting adherent‐invasive characteristics [37]. The E. coli HS strain is a non‐pathogenic commensal bacterium [70] and was used as a control for comparison with the pathogenic CD2‐a strain. Both strains were maintained at 37 °C in RPMI medium without antibiotics overnight before the experiments.

Amikacin protection assay

To evaluate the phagocytic function of THP‐1 macrophages, an amikacin protection assay was performed as we previously described [69]. Briefly, THP‐1 macrophages treated with 2 mm butyrate for 24 h in the presence or absence of 0.1 ng·mL⁻¹ LPS + 20 ng·mL⁻¹ IFN‐γ were infected with E. coli HS (commensal) and CD2‐a (AIEC) strains (multiplicity of infection, MOI = 10) for 30 min. After infection, macrophages were incubated with 100 μg·mL⁻¹ amikacin (Sigma‐Aldrich, St Louis, MO, USA) for 1 h to eliminate extracellular bacteria and lysed for 5 min in 1% Triton X‐100 (Bio‐Rad, Hercules, CA, USA) in PBS (1×, pH 7.4, Gibco‐Life Technologies, Waltham, MA, USA). Lysates were serially diluted and plated on lysogeny broth agar plates (Merck, Burlington, MA, USA). After overnight incubation at 37 °C, colonies were quantified.

RNA isolation, DNase treatment, cDNA synthesis, and real‐time quantitative reverse transcription PCR (RT‐qPCR)

Total RNA from intestinal biopsies from patients with IBD and healthy individuals was extracted using the RNeasy Plus Mini kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Total RNA from colonic organoids and peripheral blood monocytes was extracted with TRIzol^® (Invitrogen/Life Technologies, Carlsbad, CA, USA) following the manufacturer's protocol. RNA concentration was determined by spectrophotometric analysis using either a Synergy 2 (BioTek Instruments, Inc., Winooski, VT, USA) or a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

For RNA samples (0.5 or 1 μg) from intestinal biopsies, monocytes, and THP‐1 macrophages, DNA was removed using the Turbo DNA‐free™ kit (Thermo Fisher Scientific, Waltham, MA, USA). cDNA was synthesized using oligo‐dT (Promega, Madison, WI, USA) and AffinityScript reverse transcriptase (Agilent Technologies Inc., CA, USA), following the manufacturer's instructions. RT‐qPCR was performed with 10 ng cDNA using TaqMan™ Universal Master Mix II no UNG (Thermo Fisher Scientific), and amplification was performed on a Stratagene Mx3000P Multiplex QPCR System.

For RNA from colonic organoids, DNase I treatment was performed, and cDNA was synthesized using 10× RT Buffer, random primers, 10 mm dNTPs, RNase OUT, and M‐MLV reverse transcriptase (all from Invitrogen, Waltham, MA, USA), following the manufacturer's instructions. RT‐qPCR was performed with 10 ng cDNA using 2× reaction buffer QPCR Master Mix plus‐dTTP (Eurogentec, Maastricht, The Netherlands), and amplification was performed with a StepOne Plus Real‐Time PCR system.

Gene expression was analyzed using TaqMan Gene Expression Assays for SLC16A1 (Hs01560299_m1), HCAR2 (Hs02341584_s1), HDAC3 (Hs00187320_m1), IRF1 (Hs00971965_m1), TFF1 (Hs00907239_m1), DUOX2 (Hs00204187_m1), and LCN2 (Hs01008571_m1), MARCKS (Hs00158993_m1), and CD48 (Hs00381156_m1) (Applied Biosystems, Waltham, MA, USA) or TaqMan primers and probes from Eurogentec (Maastricht, The Netherlands; see Table 2). The primers and probes were designed using primer express 3.0.1 software (Applied Biosystems), aiming, where possible, to anneal to separate exonic coding regions. Their specificity for the target genes was verified using the ncbi blast tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Relative transcript quantification was determined by the 2^−ΔCt method, with 18S rRNA used as a housekeeping gene for RT‐qPCR normalization.

Table 2.

Primers and probe sequences.

Gene	Forward	Reverse	Probe
18S	CGGCTACCACATCCAAGGA	CCAATTACAGGGCCTCGAAA	CGCGCAAATTACCCACTCCCGA
TNFA	CCCTGGTATGAGCCCATCTATC	AAAGTAGACCTGCCCAGACTCG	ATCAATCGGCCCGACTATCTCGACTTT
IL6	TGGCTGAAAAAGATGGATGCT	CAAACTCCAAAAGACCAGTGATGA	CCAGGCAAGTCTCCTCATTGAATCCAGATT
NOS2	GGCTCAAATCTCGGCAGAATC	GGCCATCCTCACAGGAGAGTT	TCCGACATCCAGCCGTGCCAC
CXCL8	CTTGGCAGCCTTCCTGATTT	TTGGAGTATGTCTTTATGCACTGACA	CAAAACTGCACCTTCACACAGAGCTGCA
IL1B	ACAGATGAAGTGCTCCTTCCA	GTCGGAGATTCGTAGCTGGAT	CTCTGCCCTCTGGATGGCGG
IL18	CCAAGGAAATCGGCCTCTATT	CTTCACAGAGATAGTTACAGCCATACCT	TTCTGACTGTAGAGATAATGCACCCCGGAC
OCLN	GATGAGCAGCCCCCCAAT	GGTGAAGGCACGTCCTGTGT	TGCAGACACATTTTTAACCCACTCCTCGA
CLDN3	CAGGCGTGCTGTTCCTTCTC	GTAGAAGTCCCGGATAATGGTGTT	CCCTGCTCACCCTCGTGCCG
TJP1	CAGTGCCTAAAGCTATTCCTGTGA	GCACGCCCCCATTGC	TGGCCACAGCCCGAGGCATATTT
LGR5	TGTTTCAGGCTCAAGATGAACGT	AGCAGGTGTTCACAGGGTTTG	CCCTTCATTCAGTGCAGTGTTCACCTTCC
ALPI	TCTCCTTTGGTGGCTACACCTT	ACAGGATGGACGTGTAGGCTTT	TCCATCTTCGGGTTGGCCCCC
MUC2	CCTGCAGAGCTATTCAGAATTCC	ATCTTCTGCATGTTCCCAAACTC	CTCTGACGGCGTGCTCTTCAGTCCC
VIL1	TGACCCTGAGACCCCCATC	TCAGCAGTGATCTGGCTCCA	TTGTGGTGAAGCAGGGACACGAGC

Bulk RNA‐seq analysis

The raw RNA‐seq data underwent quality assessment using fastqc (v0.11.7) with default parameters. Adapters identified by fastqc were removed using cutadapt (v1.1) with default settings. sickle (v1.200) was then employed to trim low‐quality ends of reads, retaining sequences with a minimum length of 25 nucleotides and a quality score threshold of 20. The cleaned reads were aligned to the human genome (GRCh37, human_glk_v37) using hisat (v0.1.6), allowing for a maximum of two mismatches. Aligned reads were sorted using samtools (v0.1.19), and mapping statistics were generated using samtools flagstat and picard tools (v2.9.0). Samples with a low percentage of aligned reads (< 90%) were excluded. Gene‐level expression was quantified using htseq (v0.9.1) with annotation from ensembl version 75. Data normalization was performed using the trimmed mean of M‐values (TMM) method to account for differences in library sizes.

Single‐cell RNA‐seq analysis

Datasets were obtained from the Broad Institute and processed using python (version 10.8.4). Analysis was performed with the Scanpy and anndata packages for single‐cell data processing and visualization, and pandas for data manipulation.

Immunofluorescence microscopy

Glass microscope slides (Waldemar Knittel, Braunschweig, Germany) of paraffin‐embedded tissue or colonic organoids (4 μm) were deparaffinized and rehydrated using a graded ethanol series, followed by antigen retrieval in citrate buffer (pH 6.0).

Tissue and organoid samples were blocked with PBS 1× containing 1% BSA for 30 min at room temperature and incubated overnight at 4 °C with the following primary antibodies: rabbit anti‐human HCAR2 (GPR109A) (1 : 50, Cat No HPA028660, polyclonal), rabbit anti‐human SLC16A1 (MCT1) (1 : 250 for tissue or 1 : 100 for organoids, Cat No HPA003324, polyclonal) (both from Sigma‐Aldrich, Zwijndrecht, The Netherlands), or rabbit anti‐human HDAC3 (1 : 200, Cat No 10255‐1‐AP polyclonal, Proteintech, Rosemont, IL, USA). Subsequently, the samples were incubated with the secondary antibodies: Alexa‐Fluor 488 goat anti‐rabbit IgG (H + L) (1 : 400, A11008) or Alexa‐Fluor 594 anti‐mouse IgG (H + L) (1 : 400, A11005) (Invitrogen, Waltham, MA, USA). Hoechst or DAPI (1 : 1000, Roche, Mannheim, Germany) was used for nuclear staining. Finally, samples were mounted with Dako fluorescent mounting medium (Dako, Agilent Technologies Inc., Santa Clara, CA, USA) and visualized under a Nikon C2+ confocal microscope (MCT1 and GPR109A in tissue samples at 20× objective) (Nikon Instruments Inc, Melville, NY, USA) or a Leica SP8 confocal microscope (MCT1 in organoid samples and HDAC3 in tissue at 40× objective) (Leica Microsystems, Nussloch, Germany). Image processing was carried out using fiji/image j or lasx software. For immunofluorescence quantification, images were analyzed using qupath (version 0.5.1). At least three images were acquired per tissue or organoid condition. In colon tissue images, specific regions of interest (ROIs) corresponding to the lamina propria or epithelial compartments were manually selected using the annotation tools. For each ROI, the median fluorescence intensity was quantified using a pixel size of 0.5 μm. In organoid images, individual organoids were manually selected using the wand tool, and the median fluorescence intensity was measured for each organoid using a pixel size of 0.3 μm. The average intensity of images from the same tissue or organoid condition was used for graphing and comparative analysis.

Flow cytometry

THP‐1‐derived macrophages (1 × 10⁶ per well) were cultured in 2 mL of RPMI 1640 complete medium (Gibco‐Life Technologies, NY, USA) and butyrate in the absence or presence of inflammation for 24 h for M1 profile differentiation. After incubation, macrophages were detached using Versene 1× (Gibco‐Life Technologies, Waltham, MA, USA), and 1 × 10⁵ cells were surface‐stained with fluorochrome‐specific antibodies for human PE CD40 (Clone 5C3) and PE/Cy5 CD80 (Clone 2D10) (both from BioLegend, San Diego, CA, USA). Prior to staining, cells were treated with FcR Blocking Reagent (MACS, Miltenyi Biotec, Bergisch Gladbach, Germany) to avoid unspecific FcR staining. Cell viability was assessed using 7‐AAD (BD Biosciences Pharmingen, San Diego, CA, USA), according to the manufacturer's instructions. Data from at least 1 × 10⁴ cells were acquired from BD Accuri™ C6 Cytometer. For H3K9ac detection, human monocytes or THP‐1 macrophages treated with 2 mm butyrate or 10 μm SAHA for 6 h were fixed with 4% paraformaldehyde for 10 min and permeabilized with 90% methanol on ice for 5 min. After incubation with FcR Blocking Reagent, cells were stained with Alexa 488 Alexa Fluor^® 488 anti‐Histone H3 Acetylated Lysine 9 (K9ac) Antibody (Clone 2G1F9, BioLegend), and data were acquired using the NovoCyte Quanteon Cytometer (Agilent Technologies Inc., CA, USA).

Cytometric bead array (CBA)

Cytokine content in the conditioned media of ex vivo‐cultured explants and monocytes was detected using the BD™ CBA Human TH1/TH2/TH17 Kit (BD Biosciences Pharmingen, San Diego, CA, USA) following the manufacturer's instructions. Briefly, the supernatants of mucosa culture (exposed to butyrate or dexamethasone) and monocytes (exposed to butyrate or SAHA) were incubated with a capture bead mixture conjugated with anticytokine antibodies (TNF, IFN‐γ, IL‐2, IL‐6, IL‐10, and IL‐17A). It was then incubated with a PE‐conjugated detection antibody for 3 h at room temperature in the dark. After washing with 1× wash buffer and centrifuging at 200  g for 5 min at room temperature, data were acquired using the FACS Canto™ II flow cytometer and analyzed using the cellquest Software (both from BD Biosciences Pharmingen, San Diego, CA, USA) for conditioned medium of ex vivo‐cultured biopsies. For conditioned medium of monocytes, data were obtained using the NovoCyte Quanteon Cytometer (Agilent Technologies Inc., CA, USA) and analyzed by extrapolation on the calibration curves.

Lactate dehydrogenase (LDH) assay

Lactate dehydrogenase release was measured in the supernatant of organoid cultures using the CytoTox 96^® Non‐Radioactive Cytotoxicity Assay (Promega, Madison, WI, USA), following the manufacturer's instructions. Absorbance was measured at 490 nm using an EPOCH2 plate reader/spectrophotometer (BioTek Instruments, Inc., Winooski, VT, USA).

Statistical analysis

First, the D'Agostino & Pearson omnibus normality test was performed to evaluate the distribution of the data (non‐parametric or parametric). Based on the results, grouped analyses (one‐way ANOVA or Kruskal–Wallis test) or column analyses (paired or unpaired t‐test, Wilcoxon matched‐pairs signed‐rank test or Mann–Whitney test) were applied. Grouped analyses were used for flow cytometry results, while column analyses were used for CBA, TaqMan RT‐qPCR, and amikacin protection assay data. Bonferroni post‐test was used for ANOVA analysis. All analyses were performed using graphpad prism 10.2.3 software (GraphPad Software Inc., San Diego, CA, USA), with significance set at P ≤ 0.05.

Conflict of interest

The authors declare no conflict of interest.

Author contributions

DP‐V, MDFL, GD, KNF, and MAH designed the research. DP‐V conducted experiments and analyzed the data; MDFL, GL, KD‐C, TB, HM, M‐CC, YC, ML, and AMDCDP helped with experiments and data analysis. DS, RQ, and GD recruited patients, created the patients' database, and/or collected samples. RKW provided computing resources. MDFL, KNF, and MAH supervised the study. DP‐V, KNF, and MAH wrote the manuscript. M‐JG contributed with data analysis and interpretation. MDFL, GL, KD‐C, TB, DS, RQ, M‐JG, and GD helped with manuscript revision and data discussion. MAH and KNF approved and checked the content of the final published manuscript. All authors contributed to the article and approved the submitted version.

Data availability statement

Data supporting the gene expression analysis from bulk RNA‐sequencing is available in the Genome‐phenome Archive data repository database under accession code EGAS00001002702 [30]. Additional data underlying this article will be shared on reasonable request to the corresponding authors.