Microbial Anti-Inflammatory Molecule (MAM) secreted by Faecalibacterium prausnitzii ameliorates colitis through autophagy and gut microbiota modulation

Abstract

Objective

Faecalibacterium prausnitzii is a major commensal bacterium that plays a crucial role in intestinal homeostasis. Its secretes an microbial anti-inflammatory molecule (MAM) that functions as a potential therapeutic agent in inflammatory bowel disease (IBD). However, the molecular mechanisms through which MAM exerts its beneficial effects remain incompletely understood. Here, we investigate whether MAM modulates autophagy and to evaluate its therapeutic potential in a murine model of colitis.

Methods

A genetically engineered Lactococcus lactis strain expressing MAM was administered to mice with dextran sulfate sodium (DSS)-induced or 2,4,6-trinitrobenzene sulfonic acid (TNBS)-induced colitis. The therapeutic effects of MAM were evaluated using clinical scoring, histopathological analysis, and inflammatory cytokines measurement. Intestinal barrier function was assessed based on tight junction protein expression. Autophagy-related signaling pathways were analyzed using western blotting and immunohistochemistry. To determine how autophagy affects MAM, DSS-treated mice were treated with the autophagy inhibitor hydroxychloroquine (HCQ) and bafilomycin A1 (BafA1). Gut microbiota composition was profiled 16 S rRNA gene amplicon sequencing. Fecal short-chain fatty acid (SCFA) levels were quantified by gas chromatography–mass spectrometry (GC-MS).

Results

MAM treatment significantly alleviated colitis in both DSS- and TNBS-induced models, as evidenced by reduced disease activity, improved colon length, attenuated histopathological damage, and decreased pro-inflammatory cytokine production. MAM enhanced intestinal barrier integrity by upregulating ZO-1 and occludin. Notably, MAM counteracted the inhibitory effect of HCQ/BafA1 on autophagy, enhancing autophagic flux despite autophagy suppression. Furthermore, MAM treatment significantly increased the fecal concentrations of acetate, propionate, and butyrate. 16 S sequencing revealed the enrichment of beneficial taxa, including Lactobacillus and Lachnospiraceae_NK4A136_group, The abundance of these taxa was positively correlated with SCFA levels and improvement of colitis symptoms.

Conclusion

We identified MAM as a multifunctional microbial effector with potent anti-inflammatory properties. MAM improves intestinal barrier function and attenuates experimental colitis through dual mechanisms involving the modulation of autophagy and gut microbiota composition. These findings highlight the translational potential of MAM and support further investigation into its use as a novel therapeutic strategy for IBD.

Introduction

Inflammatory Bowel Disease (IBD), which includes Crohn’s disease (CD) and ulcerative colitis (UC), comprises a group of chronic gastrointestinal disorders characterized by recurrent episodes of intestinal inflammation [1].IBD pathogenesis involves a complex interplay among genetic predisposition, environmental triggers, immune response dysregulation, and gut microbiota dysbiosis [2, 3]. Among these factors, alterations in the gut microbial composition—particularly the depletion of beneficial commensals—have emerged as key contributors to disease initiation and progression [4–6].

Faecalibacterium prausnitzii, one of the most abundant commensal bacteria in the healthy human gut, is widely recognized for its beneficial characteristics, including its anti-inflammatory properties [7]. Numerous studies have reported a significant reduction in F. prausnitzii abundance in patients with IBD, which is negatively correlated with disease severity [8]. The administration of F. prausnitzii or its secreted products has been shown to ameliorate colitis, reduce the levels of pro-inflammatory cytokines, and enhance epithelial barrier function in preclinical models [9].

In 2014, E Quévrain et al. [10] identified microbial anti-inflammatory molecule (MAM), a 15 kDa protein secreted by F. prausnitzii. MAM exerts potent anti-inflammatory effects through the inhibition of the NF-κB signaling pathway, significantly reducing intestinal inflammation [11]. This finding shed light on the immunomodulatory mechanisms by which F. prausnitzii exerts its protective effects in the gut. However, the underlying molecular mechanisms remain incompletely understood.

Autophagy, a conserved cellular process for the degradation of damaged organelles and pathogens, has been implicated in intestinal homeostasis and host-microbe interaction regulation [12–15]. Emerging evidence suggests crosstalk between microbial metabolites and host autophagy. For instance, short-chain fatty acids (SCFAs) produced by commensal bacteria can induce autophagy in intestinal epithelial cells, thereby promoting barrier integrity and suppressing inflammation [16, 17]. Given that F. prausnitzii is a major butyrate producer and that MAM modulates signaling pathways (e.g., NF-κB, and mTOR) linked to autophagy regulation, we hypothesized that MAM may ameliorate colitis through autophagy-dependent mechanisms and the concomitant modulation of the gut microbiota.

To test this hypothesis, we used a recombinant Lactococcus lactis system to deliver MAM in murine models of colitis. We attempted to delineate the role of autophagy in MAM-mediated protection and explore its impact on microbial community structure. Our findings may provide mechanistic insights into the therapeutic potential of F. prausnitzii-derived molecules in IBD.

Materials and methods

Materials and reagents

Dextran sulfate sodium (DSS, molecular weight 36–50 kDa, MP Biomedicals), hydroxychloroquine (HCQ, MedChemExpress, USA; Cat. No. HY-W031727), bafilomycin A1 (BafA1, Selleck; Cat. No. S1413), 2,4,6-trinitrobenzenesulfonic acid (TNBS, Sigma; Cat. No. P2297), dimethyl sulfoxide (DMSO), polyethylene glycol 300 (PEG300), and Tween-80 (all from MedChemExpress).

The following antibodies were used in the study: rabbit anti-His-Tag (E-AB-40506, Elabscience) rabbit anti-AMPKα (D5A2, 5831 S CST), rabbit anti-p-AMPKα (Thr172, 2535 S CST), rabbit anti-ULK1 (8054 S CST), rabbit anti-p-ULK1 (Ser555, 5869 S CST), rabbit anti-t-ULK1 (8572 CST), rabbit anti-mTOR (2972 S CST), rabbit anti-p-mTOR (Ser2448, 2971 S CST), rabbit anti-LC3A/B (4108 S CST), rabbit anti-LC3B (2775 S CST), rabbit anti-p62 (5114 S CST), rabbit anti-Beclin-1 (12994 S CST), rabbit anti-ATG16L1 (8089T CST), rabbit anti-Atg5 (12994 S CST), rabbit anti-MUC2 (A4767 Abclonal), rabbit anti-Claudin1 (ab307692 Abcam), rabbit anti-ZO-1 (ab96587 Abcam), rabbit anti-GAPDH (2118 CST), and rabbit β-actin (4970 CST).

Plasmid construction

In accordance with the method reported by Quévrain et al. [11], recombinant plasmids containing the MAM gene (MAM) and empty control plasmids (EMPTY) were constructed and subsequently transformed into L. lactis MG1363 strain for subsequent animal experiments.

Bacterial culture

F. prausnitzii (strain A2-165) (American Type Culture Collection (ATCC, Manassas, VA, USA)) was cultured overnight at 37 °C in YCFA medium under anaerobic conditions (90% N₂, 5% CO₂, and 5% H₂). Following this, the cells were harvested by centrifugation at 1500 × g for 15 min at 4 °C and resuspended in PBS (5 × 10⁹ CFU/mL). The supernatant was filtered through a 0.22 μm membrane and stored for further use.

L. lactis MG1363 strains carrying MAM or EMPTY were cultured overnight at 30 °C in MRS broth, harvested by centrifugation, resuspended in PBS (5 × 10⁹ CFU/mL), and administered to mice via oral gavage.

Animal model

Six-to-eight-week-old male C57BL/6J mice were purchased from the Guangdong Medical Laboratory Animal Center (license number: SYXK 2022-0002). The animals were housed under specific pathogen-free (SPF) conditions with a 12-hour light/dark cycle, controlled temperature (24 ± 1 °C), and 50%-70% humidity. All treatments were delivered either by oral gavage or intraperitoneal (i.p.) injection, depending on the specific agent and experimental design at 100 µL per 10 g body weight. The procedures were approved by the Animal Ethics Committee of the Second Affiliated Hospital of South China University of Technology (Guangzhou First People’s Hospital).

The healthy control (Control) group received regular drinking water and no interventions, providing a baseline for comparison with colitis-induced groups. Dextran sulfate sodium model. Acute colitis was induced by administering 3% (w/v) DSS in sterile drinking water for 7 consecutive days.The DSS/TNBS group received daily oral gavage/intraperitoneal injection of the corresponding vehicle matched to that used in the treatment groups. To evaluate the therapeutic effects of various treatments, mice were pretreated with the corresponding agents beginning 7 days before DSS administration, The experimental duration was 14 days. (1) To evaluate the effects of F. prausnitzii and its supernatant, the mice were divided into four groups (n = 6): (i) Control, (ii) DSS + phosphate-buffered saline (PBS), (iii) DSS + F. prausnitzii, and (iv) DSS + F. prausnitzii supernatant. (2) To assess the effects of MAM, four groups (n = 6 per group) were established: (i) Control group, (ii) DSS + PBS group, (iii) DSS + MAM group, and (iv) DSS + EMPTY group, using the same dosage and administration route. (3) To investigate the role of autophagy in MAM–mediated protection, five groups (n = 5 per group) were formed: (i) Control; (ii) DSS + PBS; (iii) DSS + HCQ; (iv) DSS + HCQ + MAM; (v) DSS + HCQ + EMPTY. HCQ was dissolved in a vehicle solution composed of 5% DMSO, 40% PEG300, 5% Tween-80, and 50% saline (0.9% NaCl). The solution was prepared fresh daily and administered orally at a dose of 40 mg/kg/day [18, 19]. (4) To assess autophagic flux in vivo, mice were administered a single dose of BafA1 (0.5 mg/kg, i.p.) or the vehicle 4 h before sacrifice. Eight groups (n = 6 per group) were formed: (i) Control; (ii) DSS + PBS; (iii DSS + MAM; (iv) DSS + EMPTY; (v) Control + BafA1; (vi) DSS + BafA1; (vii) DSS + BafA1 + MAM; (viii) DSS + BafA1 + EMPTY. TNBS model. (1) To explore the role of autophagy in the therapeutic effect of MAM, the mice were anesthetized with isoflurane. A 1.5 × 1.5 cm area on the back of the mice was shaved and gently depilated using a commercial hair removal cream. Five groups (n = 6 per group) were formed: (i) Control; (ii) TNBS + PBS; (iii) TNBS + hydroxychloroquine (HCQ); (iv) TNBS + HCQ + MAM; (v) TNBS + HCQ + EMPTY. For cutaneous sensitization, a solution of 1% TNBS (w: v) was prepared by mixing 5% TNBS in water with an equal volume of a 1:4 (v: v) mixture of olive oil and acetone. The Control group was administered 150 µL of vehicle solution (olive oil: acetone = 1:4, without TNBS). After 7 days, the mice were anesthetized with isoflurane and intrarectally with 100 µL of 2.5% TNBS in 50% ethanol. The mice were euthanized on day 4 after the TNBS challenge. Throughout the experimental period, the mice were orally administered MAM, EMPTY, or HCQ daily, as indicated. The mice were euthanized on day 11 for sample collection.

Disease activity index (DAI) scores

The mice were weighed daily, and their weight loss was scored according to the percentage decrease in weight (Table 1). Diarrhea severity was assessed based on stool consistency and water content, with scores ranging from 0 to 4. Rectal bleeding was assessed based on occult or visible blood detection in feces and scored from 0 to 4. The overall DAI was calculated as the average of the three individual scores.

Table 1.

Disease activity index (DAI) scoring table [20]

Weight Loss(%)	Stool Consistency	Occult Blood/Visible Blood in Stool	Score
0	Normal (formed stool)	Normal	0
≥ 1 and < 5	Intermediate	/	1
≥ 5 and < 10	Loose (increased water content, non- adherence to anus)	Occult blood test positive	2
≥ 10 and < 15	Intermediate	/	3
≥ 15	Watery (adherence to anus)	Visible blood	4

Histological evaluation of colitis and alcian blue-periodic acid schiff

At the end of the experiment, fresh colon tissues were harvested from the mice in each group, rinsed, and fixed in 4% paraformaldehyde solution for 24 h. The tissues were then dehydrated, embedded, and sectioned into 4 μm slices. Hematoxylin-eosin (HE) staining was performed to evaluate general histopathological changes. Additionally, AB-PAS staining was conducted to assess mucin production and goblet cell morphology. For each mouse, three non-consecutive cross-sections of the distal colon (spaced ≥ 100 μm apart) were prepared. From each section, five randomly selected, non-overlapping high-power fields were captured and analyzed. Histopathological scoring was performed by a blinded observer using a previously established scoring system (Table 2).

Table 2.

Histopathological scoring [21]

Mucosal Epithelial Damage	Inflammatory Cell Infiltration	Score
No epithelial damage	Rare inflammatory cell infiltration in the lamina propria	0
Single epithelial damage	Increased inflammatory cells and neutrophil infiltration in the lamina propria	1
Mucosal fusion and ulceration	Inflammatory cell infiltration also found in the submucosa	2
Extensive damage involving the entire layer of the mucosal epithelium	Full-thickness inflammatory cell infiltration in the submucosa	3

Note: Each component was scored independently. Total histopathological score = Epithelial Damage Score + Inflammatory Infiltration Score

Intestinal permeability assay

Intestinal barrier function was assessed by measuring the in vivo gut permeability to fluorescein isothiocyanate–dextran (FITC-dextran, 3,000–5,000 Da; MCE, HY-66019). The mice were fasted for 12 h with free access to water. FITC-dextran (50 mg/mL) was prepared in sterile PBS and administered by oral gavage at a dose of 100 µL per 10 g body weight. Four hours after gavage, blood samples were collected from the orbital sinus or via cardiac puncture. Sera samples were isolated by centrifugation and transferred to black 96-well plates. Fluorescence intensity was measured using a microplate reader at an excitation wavelength of 490 nm and an emission wavelength of 530 nm. Standard curves were generated using known concentrations of FITC-dextran in serum to quantify intestinal permeability.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was extracted from colon tissues using TRIzol reagent (Takara, Japan) in accordance with the manufacturer’s instructions. Complementary DNA (cDNA) was synthesized from total RNA using the PrimeScript™ RT Master Mix (Takara, Japan). RT-qPCR was performed to assess the mRNA expression levels of target genes using gene-specific primers (listed in Table 3). Relative mRNA expression was quantified using the comparative cycle threshold method (2^-ΔΔCt), with Actb (β-actin) as the internal reference gene.

Table 3.

Primer sequences used for real-time PCR

Primers		Sequence
ACTB	Forward	CATTGCTGACAGGATGCAGAAGG
	Reverse	TGCTGGAAGGTGGACAGTGAGG
TFF3	Forward	TCCAAGCCAATGTATGGTGCCG
	Reverse	CAGGGCACATTTGGGATACTGG
Occludin1	Forward	TGGCAAGCGATCATACCCAGAG
	Reverse	CTGCCTGAAGTCATCCACACTC
MUC2	Forward	CCCAGAAGGGACTGTGTATG
	Reverse	TGCAGACACACTGCTCACA
ZO-1	Forward	GTTGGTACGGTGCCCTGAAAGA
	Reverse	GCTGACAGGTAGGACAGACGAT
IL18	Forward	GACAGCCTGTGTTCGAGGATATG
	Reverse	TGTTCTTACAGGAGAGGGTAGAC
IL6	Forward	TACCACTTCACAAGTCGGAGGC
	Reverse	CTGCAAGTGCATCATCGTTGTTC
IL1β	Forward	TGGACCTTCCAGGATGAGGACA
	Reverse	GTTCATCTCGGAGCCTGTAGTG
TNF-α	Forward	TTAGAAAGGGGATTATGGCTCA
	Reverse	ACTCTCCCTTTGCAGAACTCAG

Western blot analysis

Colon protein samples were prepared using RIPA buffer (Beyotime, China) supplemented with a protease inhibitor cocktail (Solarbio, China). Proteins were separated by 10% SDS-PAGE gel and transferred to a PVDF membrane (Invitrogen). After blocking with 10% non-fat milk for 1 h at room temperature, membranes were incubated overnight at 4 °C with primary antibodies. After washing, each membrane was incubated with HRP-conjugated secondary antibodies for one hour at room temperature. Protein bands were visualized using an ECL detection kit (Fdbio Science, China), using β-actin or GAPDH as the internal protein. Densitometric analysis was performed using ImageJ software.

Immunohistochemistry (IHC)

Paraffin-embedded tissue Sect. (3 μm) were dewaxed and subjected to antigen retrieval. Endogenous peroxidase activity was blocked by treating the sections with 3% hydrogen peroxide. The sections were blocked with 3% bovine serum albumin (BSA) before treatment primary antibodies overnight at 4 °C. The following day, the sections were treated with secondary antibodies conjugated to mouse or rabbit IgG. Thereafter, immunostaining was performed using diaminobenzidine (DAB) chromogen. Hematoxylin was used for nuclei counterstaining. Positive signals were quantified as the percentage of positively stained area per HPF using the Fiji: ImageJ software. At least three HPFs were analyzed per section, and the average value per mouse was used for statistical comparison.

Immunofluorescence (IF)

The tissue sections were treated with primary antibodies overnight at 4 °C, and then with fluorescent secondary antibodies for 1 h at room temperature. After washing with PBS, the sections were mounted with 4’,6-diamidino-2-phenylindole (DAPI) for nuclei staining. Fluorescent images were acquired using a fluorescence microscope (Leica, Germany) and analyzed using the Fiji: ImageJ software.

Gut microbiota analysis

Cecal contents of the mice were collected at sacrifice and immediately flash-frozen in liquid nitrogen. Genomic DNA was extracted from samples using the QIAamp Fast DNA Stool Mini Kit (Qiagen, Cat. No. 51504) according to the manufacturer’s instructions. The V3–V4 hypervariable regions of the bacterial 16S rRNA gene were amplified by PCR using universal primers (341F: 5’-CCTACGGGNGGCWGCAG-3’ and 806R: 5’-GGACTACHVGGGTATCTAAT-3’). Amplicons were purified, indexed, and sequenced on an Illumina NovaSeq 6000 platform (GENE DENOVO, Guangzhou, China). Bioinformatic analysis was performed using Omicsmart, an interactive real-time online platform for data analysis http://www.omicsmart.com).

Statistical analysis

Unpaired Student^,s t-test, Wilcox test, or Mann-Whitney test was used for two-group comparisons. One-way ANOVA, Kruskal-Wallis’s test, or Tukey^,s post hoc test was used for comparisons among three or more groups. For time-course data (e.g., body weight, disease activity index), two-way ANOVA with repeated measures was used with group and time as factors. The results were considered statistically significant at P < 0.05 (*P < 0.05, **P < 0.01, ***P < 0.005, ***P < 0.0001). GraphPad Prism 9.4.0 and SPSS 26 software were ued in all analyses.

Results

Treatment with F. prausnitzii and its supernatant improves experimental ulcerative colitis in mice

Previous studies have demonstrated that F. prausnitzii A2-165 and its supernatant effectively ameliorate the symptoms of TNBS-induced colitis in mice, including weight loss attenuation and colon inflammation reduction [22]. Here, we established a mouse model of acute colitis by treating the mice with 3% DSS and treated intragastrically administering PBS (DSS + PBS), F. prausnitzii (DSS + A2-165), or its supernatant (DSS + A2-165 Sup) Fig. 1A. By post-induction day 4, DSS-treated mice exhibited progressive weight loss and diarrhea. By day 5, a statistically significant difference in body weight was observed between mice from the DSS and control groups (Fig. 1B), accompanied by hematochezia. Clinical symptoms, including lethargy, marked weight loss, severe diarrhea, and bloody stools, worsened by days 6 and 7, leading to significantly elevated DAI values (Fig. 1C) and pronounced colon shortening (Fig. 1D-E). Both F. prausnitzii and its supernatant alleviated weight loss, decreased DAI scores, and mitigated colon shortening. HE staining revealed that DSS-induced colitis was associated with epithelial damage, inflammatory cell infiltration, and goblet cell depletion. These pathological changes were considerably reversed in mice treated with either F. prausnitzii or its supernatant (Fig. 1F-G). Quantitative PCR analysis revealed the indicated upregulation of intestinal barrier-related genes (including Occludin1, MUC2, and ZO-1) in both treatment groups (Fig. 1H), alongside the downregulation of pro-inflammatory cytokines IL-18, IL-6, and IL-1β (Fig. 1I-K). IF analysis corroborated these mRNA findings (Fig. 1L), supporting the enhancement of intestinal barrier function by both F. prausnitzii and its supernatant (Fig. 1M). Taken together, these data suggest that both F. prausnitzii and its supernatant confer protection against DSS-induced colitis in mice by reinforcing the intestinal barrier and suppressing pro-inflammatory cytokine production.

Fig. 1.

F. prausnitzii and its supernatant alleviate dextran sulfate sodium (DSS)-induced colitis in mice. (A) Experimental design; (B) Body weight changes; #: Control vs. DSS + PBS, *: DSS + A2-165 /DSS + A2-165 Sup - vs. DSS + PBS; (C) Disease Activity Index (DAI); #: Control vs. DSS + PBS, *: DSS + A2-165 / DSS + A2-165 Sup - vs. DSS + PBS; (D) Colon length measurement; (E) Gross appearance of the colon; (F) Histological scores; (G) Representative images of hematoxylin-eosin (HE)-stained colon sections; (H-K) Relative mRNA expression levels of Occludin1, MUC2, ZO-1 IL-18, IL-6, and IL-1β in the colon; (L) Immunofluorescence staining for MUC2 in colon sections. Blue: DAPI; Green: MUC2. Scale bar: 200 μm; (M) Relative Fluorescence Intensity of MUC2 in the colon normalized to that in the control group. (n = 3 per group). Scale bars: 200 μm (G, L); Data are presented as mean ± SEM (n = 5–6 per group). *, P < 0.05, **, P < 0.01, ***, P < 0.001, ****, P < 0.0001, as determined by two-way repeated measures ANOVA followed by Tukey’s multiple comparisons test (Fig. B and C) and one-way ANOVA followed by Tukey’s multiple comparisons test (D, F, H, I, J, K, M)

MAM from F. prausnitzii supernatant improves experimental acute colitis in mice

E Quévrain et al. [11] identified MAM, a 15 kDa protein produced by F. prausnitzii, as a potent anti-inflammatory mediator. Here, we cloned the MAM-encoding plasmid (MAM) into L. lactis MG1363 and prepared an empty vector (EMPTY) using an identical method [23]. We used an anti-His tag antibody to detect proteins in both the culture supernatant and bacterial cells of L. lactis MG1363 carrying either an empty vector or the MAM-expressing plasmid (Fig. 2L). One week prior to the induction of acute colitis with 3% DSS, the mice were intragastric administered L. lactis MG1363 harboring either MAM or EMPTY. Four days after DSS induction, mice in both the DSS + PBS and DSS + EMPTY groups displayed weight loss and diarrhea, whereas those in the DSS + MAM group gained weight, with statistically significant differences with corresponding parameters in mice from the other two groups (Fig. 2B). During the entire disease course, the DSS + MAM group experienced significantly less weight loss. Consistently, MAM administration led to marked improvements in clinical symptoms, including reduced diarrhea and hematochezia, resulting in lower DAI scores (Fig. 2C). Moreover, MAM treatment mitigated colon shortening (Fig. 2D-E) and ameliorated histopathological damage, such as epithelial disruption and inflammatory cell infiltration (Fig. 2F-G). Intestinal permeability testing showed that mice from the DSS group had a damaged intestinal barrier. The intestinal permeability decreased in mice treated with MAM (Fig. 2H). The mRNA expression levels of the pro-inflammatory cytokines IL-6, IL-1β, and TNF-α were reduced in the MAM group (Fig. 2I). In parallel, the expression of immune barrier proteins (Claudin1 and MUC2) was upregulated following MAM treatment (Fig. 2J-K).

Fig. 2.

MAM alleviates DSS-induced colitis in mice. (A) Experimental design; (B) Body weight change. *: DSS + MAM- vs. DSS + PBS/DSS + EMPTY; #: Control vs. DSS + PBS; (C) Disease Activity Index; (D) Representative colon images; (E) Colon length; (F) Histological score; (G) HE staining of colon sections. Scale bars: 200 μm; (H) FITC-dextran concentration in the serum; (I) mRNA levels of Il-6, Il-1β, and TNF-α in colon tissue (n = 4–6 per group); (J) Relative fluorescence intensity of Claudin1 and MUC2; (K) Immunofluorescence staining for Claudin1 and MUC2. Blue: DAPI; Green: target protein (n = 3 per group), Scale bars: 200 μm; (L) Western blot with anti-His antibody for His-tagged MAM protein verification in L. lactis MG1363.Data are presented as mean ± SEM. *, P < 0.05, **, P < 0.01, ***, P < 0.001, ****, P < 0.0001, as determined by two-way repeated measures ANOVA followed by Tukey’s multiple comparisons test (B, C) and one-way ANOVA followed by Tukey’s multiple comparisons test (E, F, H, I)

MAM intervention alters gut microbiota composition and increases the abundance of beneficial bacterial genera

Gut microbiota dysbiosis is a well-established hallmark of ulcerative colitis. To evaluate the effect of MAM intervention on the gut microbiota, we performed 16 S rRNA gene amplicon sequencing of cecal contents. Venn diagram analysis revealed 1,094 shared operational taxonomic units (OTUs) among the all three groups, with 112 OTUs to the DSS group and 162 specifics to the MAM group. The OTU count significantly increased significantly in response to MAM treatment (Fig. 3A). Non-metric multidimensional scaling analysis at the OTU level demonstrated clear separation among the three groups (stress = 0.068; R² = 0.2463, p = 0.024), suggesting distinct community structures (Fig. 3C). Although the Shannon diversity index showed an increasing trend in the MAM group, this difference was not statistical significance (Fig. 3B).

Fig. 3.

Effects of MAM on the gut microbiota and colitis in DSS-treated mice. (A) Venn diagram showing the shared and unique operational taxonomic units (OTUs). (B) α-diversity analysis. Shannon diversity index of the gut microbiota in each group. (n = 6 per group); (C) β-diversity analysis. Nonmetric multidimensional scaling (NMDS) and analysis of dissimilarities between bacterial communities in sampling groups [ANOSIM] R = 0.2463; p-value = 0.024 [NMDS] Stress = 0.068. Ordination method: NMDS, distance method: Bray- Curtis index, taxonomic level: OTU, statistical method: analysis of groups similarities (ANOSIM); (D) Relative abundance of phyla in each group; (E) Relative abundance of genera in each group; (F) Cladogram generated using the results of Linear Discriminant Analysis Effect Size analysis; (G) Relative abundance of selected bacteria at the family and genus levels in each group. (n = 4–6 per group); (H) Spearman correlation analysis between the relative abundance of Lachnospiraceae_NK4A136_group and DAI score, weight loss, histological score, and colon length. (n = 5 per group). All microbial abundances were calculated based on 16 S rRNA gene amplicon sequencing data after normalization and are expressed as relative proportions (%)

At the phylum level, MAM administration increased the abundance of Firmicutes (+ 2.59%) and Bacteroidota (+ 2.75%) and reduced the abundance of Verrucomicrobiota (− 4.79%) and Patescibacteria (− 1.2653%) relative to the DSS model group (Fig. 3D). At the genus level, MAM treatment reduced the relative abundance of Akkermansia (4.782%) and increased the abundance of Lactobacillus (0.5081%), Lachnospiraceae_NK4A136_group (4.1946%), and Bacteroides (4.2855%) (Fig. 3E). The Linear Discriminant Analysis (LDA) Effect Size (LEfSe) algorithm with an LDA score threshold of > 3.5 showed significant differences in the relative abundance of key taxonomic groups among cohorts, highlighting distinct microbial signatures associated with disease and therapeutic intervention (Fig. 3F).

Subsequently, we analyzed the top 10 differentially abundant bacteria at both the family and genus levels. As shown in Fig. 3G, Bacteroidaceae, Enterobacteriaceae, Lachnospiraceae, Bacteroides, and Lachnospiraceae_NK4A136_group were significantly enriched in the MAM-treated group compared to that in the DSS model group, with statistically significant differences. Among these, the Lachnospiraceae_NK4A136_group has been reported to produce SCFAs via the fermentation of dietary polysaccharides, exhibiting potential anti-colitis activity. Correlation analysis was performed between the abundance of Lachnospiraceae_NK4A136_group and other parameters, A positive correlation was observed with colon length and body weight and a negative correlation with DAI and histopathological scores (Fig. 3H).

To obtain functional insights into shifts in the microbial community at the taxonomic level, we performed metagenomic function prediction using Tax4Fun2 based on the 16 S rRNA gene sequencing data. As shown in the heatmap (Fig. 4A), significant alterations in microbial metabolic pathways were evident across experimental groups. Notably, pathways associated with SCFA biosynthesis—including butanoate metabolism and propanoate metabolism—were markedly enriched in the DSS + MAM group compared to that in the DSS group (ANOVA, p < 0.001). Similarly, key bile acid metabolism pathways, such as primary bile acid biosynthesis and secondary bile acid biosynthesis, were also upregulated in MAM-treated mice. In contrast, pathways linked to inflammation and pathogen invasion, including Staphylococcus aureus infection, bacterial invasion of epithelial cells, and NF-κB signaling, were significantly downregulated in the MAM group. These results indicate that MAM treatment is associated with significant alterations in both the taxonomic composition and predicted functional profile of the gut microbiota.

Fig. 4.

Functional analysis of gut microbiota and short-chain fatty acid (SCFA) metabolism in mice with DSS-induced colitis. (A) Heatmap showing the relative expression levels of key metabolic and signaling pathways in colon tissues (n = 6), (*) indicates significant differences between DSS and DSS + MAM groups; (B) Correlation heatmap between SCFA levels and intestinal inflammation parameters (n = 3); (C) Quantification of SCFAs by targeted metabolomics (n = 5). Statistical significance determined by one-way ANOVA followed by Tukey’s post hoc test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Based on these findings, we hypothesized that the protective effect of MAM on colitis might be mediated through SCFAs. Targeted metabolomic analysis using GC-MS confirmed a substantial increase in the levels of acetic acid, butyric acid, propionic acid, and total SCFAs in the MAM-treated group compared to that in the DSS control (Fig. 4C). Specifically, total SCFAs exhibited strong positive correlations with colon length, Claudin1, and MUC2 expression, while demonstrating negative correlations with DAI scores and histological damage scores (Fig. 4B). These results highlight a potential mechanism by which MAM influences host physiology through microbial metabolite production. However, the relationship is currently correlative. Future investigations should aim to establish causality via targeted interventions.

MAM regulates the autophagy signaling pathway in experimental colitis

Autophagy plays a pivotal role in maintaining intestinal homeostasis and promoting mucosal repair [13]. Autophagy is impaired in colitis, and its activation could alleviate inflammation and support intestinal barrier integrity [19]. Based on these findings, we investigated whether MAM intervention could modulate autophagic responses in a DSS-induced colitis model.

We conducted qPCR on colon samples, assessing the mRNA expression levels of markers associated with autophagy (including mTOR, Atg5, Atg16l1, LC3b, and P62), tight junction proteins (ZO-1, Claudin1, and Occludin), mucin secretion (Relmb, MUC2, and TFF3), and pro-inflammatory cytokines (IL-6, IL-18, IL-1β, and TNF-α) (Fig. 5A). Compared to mice from the DSS model group, MAM-treated mice exhibited enhanced expression of barrier-protective and mucin-related genes, downregulation of inflammatory mediators, and upregulation of autophagy-related transcripts.

Fig. 5.

MAM intervention influences autophagy levels in mice. (A). Heatmap analysis of the expression of genes related to mucin secretion, tight junction protein expression, inflammation, and autophagy (n = 5–6 per group); (B-E). Quantitative analysis of Atg16L1 (B), P62 (C), Beclin1 (D), and LC3II/I protein levels (n = 3 per group). (F). Representative western blots of autophagy-related protein (Atg16L1, P62, Beclin1, and LC3II/I) and β-actin protein expression (n = 3 per group); (G). Immunohistochemical staining of ATG5 and Beclin1 in colon tissues (n = 3 per group); (H). Quantitative analysis of immunohistochemistry scoring for ATG5 and Beclin1(n = 4–6 per group); (I). Quantitative analysis of p-mTOR/t-mTOR level (n = 3 per group); (J) Representative western blots of mTOR, ULK1, AMPK, and GAPDH (n = 3 per group); (K). Quantitative analysis of p-ULK1/t-ULK1 level; (L) Quantitative analysis of p-AMPK/ t-AMPK level (n = 3 per group). Data were normalized to loading controls and presented as fold change relative to the control. Statistical analyses: one-way ANOVA with Tukey’s post hoc test for panels B–E, I, K, L; unpaired t-test for panel H. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

These transcriptional changes were further corroborated by western blotting (Fig. 5B-F) and IHC (Fig. 5G-H) staining, which revealed the increased expression of autophagy-related proteins in the MAM group. These findings suggest that the therapeutic effects of MAM on colitis may be associated with the activation of autophagic pathways. To explore the upstream signaling mechanisms underlying MAM-induced autophagy activation, we performed western blot analysis to evaluate the phosphorylation status of key regulators in the autophagy initiation cascade, including AMPK, mTOR, and ULK1.

As shown in Fig. 5I–L, MAM treatment significantly increased the phosphorylation of AMPK at Thr172, as indicated by a higher p-AMPK/t-AMPK ratio, suggesting the activation of this energy-sensing kinase. Concurrently, we observed a marked decrease in the phosphorylation of mTOR at Ser2448, with a reduced p-mTOR/t-mTOR ratio, indicating the inhibition of mTOR complex 1, a well-established negative regulator of autophagy. Downstream of AMPK and mTOR, ULK1 phosphorylation at Ser555 was significantly enhanced in the MAM group, reflected by an elevated p-ULK1 / t-ULK1 ratio consistent with ULK1 activation and autophagosome initiation.

These results demonstrate that MAM modulates the AMPK/mTOR/ULK1 signaling axis in the colonic epithelium, leading to the activation of functional autophagy. Along with the observed improvements in barrier function and inflammation, these findings suggest that the therapeutic effects of MAM on experimental colitis are, at least in part, mediated through the regulation of this canonical autophagy signaling pathway.

MAM ameliorates aggravated colitis caused by autophagy Inhibition

To assess autophagic flux, we employed bafilomycin A1 (BafA1), a specific inhibitor of lysosomal acidification [24], to evaluate changes in LC3-II and p62 protein levels. Western blot analysis revealed that under conditions without BafA1 (–BafA1), the MAM-treated group exhibited improved LC3-I-to-LC3-II conversion and enhanced p62 degradation compared to that in the DSS model group (Fig. 6A). In the presence of BafA1 (+ BafA1), which blocks autophagosome-lysosome fusion, the LC3-II levels increased across all groups, indicating the effective inhibition of autophagic degradation. However, no statistically significant differences were observed among the groups (Fig. 6C), suggesting similar trends in autophagosome accumulation without intergroup variation. In contrast, p62 expression was markedly elevated in the DSS model group. MAM treatment significantly reduced p62 levels, particularly under + BafA1 conditions, where this reduction was more pronounced (Fig. 6B). These findings indicate that MAM not only promotes autophagosome formation but also enhances autophagic degradation capacity, thereby increasing functional autophagic flux.

Fig. 6.

MAM ameliorates colitis caused by autophagy inhibition. (A). Representative western blots showing the expression levels of P62, LC3I, and LC3II under -BafA1 and + BafA1 conditions (n = 3 per group); (B). Quantitative analysis of P62 protein levels normalized to GAPDH level; (C). Quantitative analysis of LC3II/GAPDH expression; (D). Experimental design. (n = 5 per group); (E). Body weight changes. (n = 5 per group); (F). Disease Activity Index (DAI). (n = 5 per group); (G). Macroscopic assessment of colon length; (H). Colon length; (I). Histological score; (n = 5 per group); (J). Histological analysis; Scale bar = 100 μm; (K-N). mRNA expression levels of tight junction proteins and inflammatory cytokines. Claudin1 (K), ZO-1 (L), IL-6 (M), and TNF-α (N); (O). Immunofluorescence staining of tight junction proteins. Nuclei were stained with DAPI. Data are presented as mean ± SD. Statistical analysis was performed using two-way ANOVA followed by Tukey’s post hoc test (E, F) and one-way ANOVA followed by Tukey’s post hoc test (B, C, H, I, K-N) *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

HCQ, an antimalarial agent, is widely recognized for its ability to inhibit autophagy in various pathological contexts [25]. To investigate whether MAM could exert protective effects even when autophagy is pharmacologically impaired, we co-administered HCQ and MAM or EMPTY in mice with DSS-induced colitis and evaluated disease severity across treatment groups. As shown in Fig. 6E-F, HCQ treatment exacerbated colitis symptoms, as evidenced by heightened weight loss and elevated DAI scores compared to those in the DSS group. MAM administration attenuated HCQ-induced exacerbation of colitis, but no protective effect was observed in the EMPTY group. Colon length measurements (Fig. 6G-H) and histopathological analysis (Fig. 6I-J) revealed that HCQ treatment led to pronounced colon shortening, epithelial damage, and inflammatory infiltration compared to that in the DSS group. These alterations were substantially ameliorated in response to MAM treatment. The mRNA expression of intestinal barrier-related proteins and pro-inflammatory cytokines was further assessed using qPCR. The results confirmed that MAM restored the expression of key barrier proteins Claudin1 and ZO-1 (Fig. 6K-L) and suppressed the HCQ-induced upregulation of IL-6 and TNF-α (Fig. 6M-N). These molecular findings were corroborated by IF staining, which revealed increased expression of Claudin1, MUC2, and ZO-1 in the MAM-treated group (Fig. 6O), indicating enhanced epithelial integrity and barrier function.

To validate the generalizability of these findings in a distinct colitis model, we extended our study to the TNBS-induced colitis model. As shown in Fig. 7, MAM alleviated HCQ-aggravated disease severity in TNBS-treated mice. Specifically, MAM reduced weight loss (Fig. 7B), improved survival (Fig. 7C), attenuated colon shortening (Fig. 7D-E), and ameliorated histological damage (Fig. 7F-H). These consistent protective effects across two independent colitis models, particularly under HCQ-induced pharmacological autophagy impairment, support the therapeutic potential of MAM in mitigating autophagy dysfunction-associated intestinal inflammation.

Fig. 7.

MAM alleviates TNBS-induced colitis in mice. (A) Experimental design; n = 6/group; (B) Survival curve; (C) Body weight change; *: TNBS + MAM- vs. TNBS + PBS / TNBS + EMPTY; #: Control vs. DSS + PBS; (D) Representative colon images; (E) Colon length; (F) FITC-dextran concentration in the serum; (G) Histological score; (H) HE staining of colon sections. Scale bars:500 μm, 100 μm. Data are presented as mean ± SEM. *, P < 0.05, **, P < 0.01, ***, P < 0.001, ****, P < 0.0001, as determined by two-way repeated measures ANOVA followed by Tukey’s multiple comparisons test (C) and one-way ANOVA followed by Tukey’s multiple comparisons test (E, F, G)

Collectively, the above data show that MAM attenuates HCQ-aggravated colitis. This improvement is associated with enhanced autophagic flux, restoration of intestinal barrier integrity, and suppression of inflammation.

Discussion

F prausnitzii, a gut symbiotic probiotic bacterium, has long been investigated owing to its critical role in maintaining intestinal homeostasis. However, because it is strictly anaerobic, in vitro cultivation is significantly challenging, which hinders mechanistic studies. F. prausnitzii and its metabolites possess potent anti-inflammatory properties with documented therapeutic benefits in IBD models. Our findings confirmed that both F. prausnitzii and its culture supernatants effectively alleviate DSS-induced colitis in mice. This is evidenced by reduced body weight loss, improved colon shortening, decreased DAI scores, and enhanced intestinal barrier integrity, consistent with prior findings [26, 27].

Among the bioactive components secreted by F. prausnitzii, the MAM protein has recently emerged as a key effector with substantial immunomodulatory activity. It can inhibit the aberrant activation of the NF-κB signaling pathway and promote the differentiation of regulatory T cells, thereby effectively alleviating intestinal inflammation in experimental colitis models [28].

In this study, by engineering L. lactis to express the MAM protein and administering it orally, the biological effects of F. prausnitzii supernatant were successfully recapitulated. The findings confirmed the central role of MAM in IBD treatment [28, 29].

In recent years, autophagy has garnered considerable attention as a key regulator of intestinal immune homeostasis and a critical element in IBD pathogenesis [30–32]. Polymorphisms in autophagy-related genes (including ATG16L2 and LC3) have been associated with increased susceptibility to IBD, suggesting that autophagy dysfunction may be closely related to IBD development [33–35].

While prior studies have established that MAM ameliorates intestinal inflammation by inhibiting NF-κB signaling and enhancing epithelial barrier integrity [28], our study revealed a previously unrecognized role of MAM in activating functional autophagy in colonic epithelial cells. Mechanistically, we found that MAM significantly upregulated Beclin1 expression and reduced p62 levels, indicating the enhancement of autophagic flux. BafA1 intervention helped confirm this finding. Critically, western blot analysis showed that MAM activates the canonical AMPK/mTOR/ULK1 axis: it increases the phosphorylation of AMPK and ULK1 while suppressing mTOR activity, thereby initiating autophagosome formation. Notably, this autophagy-inducing effect appears to complement the known anti-inflammatory and barrier-protective functions of MAM. Given that autophagy regulates both innate immune responses (e.g., by limiting inflammasome activation) and epithelial homeostasis (e.g., via the clearance of damaged organelles and tight junction maintenance), our findings position autophagy not as an isolated pathway but as an integral component that synergizes with NF-κB inhibition and barrier reinforcement to mediate MAM’s complete protective effect in colitis.

Intestinal homeostasis relies on dynamic equilibrium among intestinal barrier integrity, immune homeostasis, and microbial composition [36, 37]. Beyond its direct immunomodulatory effects, MAM significantly impacts gut microbiota composition. 16 S rRNA gene amplicon sequencing revealed that MAM reshaped the microbial community structure. In particular, the relative abundance of beneficial bacteria, such as Lactobacillus and Lachnospiraceae_NK4A136_group, increased significantly. The abundance of Lachnospiraceae_NK4A136_group showed a positive correlation with colon length and body weight and a negative correlation with DAI and histopathological scores, suggesting its potential role in MAM-mediated intestinal protection. These results indicate that MAM can modulate the gut microbiome, providing compelling evidence for “probiotic-microbiota interactions.” The changes are likely a combination of direct ecological modulation and indirect effects via improved mucosal homeostasis. However, whether MAM exerts its anti-inflammatory effects through a “host-microbiota interaction” network is yet to be determined.

Notably, our data reveal a striking concurrence between MAM-induced autophagy activation in colonic epithelial cells and the enrichment of beneficial, SCFA-producing bacterial taxa. We therefore propose that MAM may exert its therapeutic effects through a coordinated network involving host autophagy pathway activation and gut microbiota functional remodeling, forming a synergistic anti-inflammatory axis. However, it is critical to emphasize that the current evidence supports only a strong association and potential interplay between these two processes; a direct causal relationship.

Using MAM-based therapy in human IBD has key translational challenges. First, F. prausnitzii is one of the most important and abundant bacterial species in the healthy human gut microbiota, accounting for approximately 5%–15% of the total microbial community [7, 8, 38]. Multiple studies have shown that its abundance is reduced in patients with IBD, and the degree of depletion correlates with disease severity and relapse risk [39, 40]. This dysbiotic state may limit the responsiveness of patients to MAM-based interventions [41]. Therefore, in future clinical applications, patients could be categorized based on their baseline F. prausnitzii levels. Alternatively, combination therapies—such as co-administration with prebiotics (e.g., resistant starch) or fecal microbiota transplantation (FMT)—may be used to restore a favorable microbial niche before MAM administration. Second, the stability and bioavailability of MAM in the human gastrointestinal tract remain important considerations. E. Quévrain et al. [10] demonstrated that the MAM plasmid persists in the mouse gut and retains transcriptional activity. However, its long-term stability in the intestinal environment is uncertain, and unprotected bacteria may be rapidly cleared, thereby reducing therapeutic efficacy. To address this, advanced encapsulation technologies, such as pH-responsive alginate-chitosan microcapsules or enteric-coated hydrogels [42], could be used to protect the engineered bacteria during transit and enable targeted colonic release [43]. Furthermore, lyophilized formulations supplemented with cryoprotectants may extend shelf life and enhance bacterial survival in vivo. Third, early-phase clinical trials of MAM should adopt a phase I/IIa design in patients with mild-to-moderate ulcerative colitis who are refractory to conventional therapy. The inclusion criteria could include confirmed diagnosis, active disease (Mayo score 4–8), and a stable medication regimen. Primary endpoints may include changes in fecal F. prausnitzii abundance, fecal calprotectin levels, serum CRP, and intestinal autophagy markers (e.g., LC3-II in rectal biopsies). Safety monitoring should focus on bacteremia, immune activation, and the off-target effects of sustained MAM expression.

This study had several limitations. First, the small sample size (n = 5–6 per group) limited the statistical power to detect subtle but biologically relevant effects, especially in microbiota and molecular analyses. Second, all experiments were performed only in male mice. Given the known sex differences in immunity and gut microbiota, our findings may not be generalizable to females. Third, although MAM appears to enhance autophagic flux, the mechanism remains unclear: BafA1 treatment showed increased p62 degradation with MAM but led to inconsistent LC3-II accumulation. Autophagy inhibition did not fully reverse MAM’s benefits, suggesting that autophagy may not be strictly required. Moreover, the unavailability of early time-course data (e.g., within 24 h) did not allow us to confirm whether autophagy activation precedes inflammation resolution. Critically, while western blot analysis revealed activation of the AMPK/mTOR/ULK1 pathway, the precise molecular targets of MAM have yet to be identified, beyond autophagy, MAM may also act via non-autophagic pathways—such as barrier enhancement, oxidative stress modulation, or immune regulation—the molecular targets for which are unknown. Additionally, the microbiota changes observed were correlative and require functional validation via FMT or in gnotobiotic models. Future studies using autophagy-deficient mice, time-series analyses, and target identification approaches, such as affinity purification coupled with mass spectrometry (AP-MS), CRISPR-based screening, or receptor knockout models, are required to fully elucidate the mechanism of action of MAM and assess its efficacy across IBD subtypes, long-term safety, and translational potential.

Conclusions

In conclusion, our findings revealed a novel mechanism by which MAM ameliorates IBD through the activation of the autophagy pathway, providing a theoretical foundation for the development of precision therapeutics based on F. prausnitzii or its functional components. Engineered probiotics or recombinant protein-based formulations of MAM hold promise as next-generation interventions for IBD and other chronic inflammatory conditions.

Abbreviations

MAM

Microbial anti-inflammatory molecule

IBD

Inflammatory bowel disease

DSS

Dextran sulfate sodium

TNBS

2,4,6-trinitrobenzene sulfonic acid

HCQ

Hydroxychloroquine

BafA1

Bafilomycin A1

SCFA

Short-chain fatty acid

GC-MS

Gas chromatography–mass spectrometry

CD

Crohn’s disease

UC

Ulcerative colitis

DMSO

Dimethyl sulfoxide

PEG300

Polyethylene glycol 300

SPF

Specific pathogen-free

i.p

Intraperitoneal

HE

Hematoxylin-eosin

FITC-dextran

Fluorescein isothiocyanate–dextran

RT-qPCR

Reverse Transcription Quantitative Polymerase Chain Reaction

IHC

Immunohistochemistry

DAB

Diaminobenzidine

IF

Immunofluorescence

OTUs

Operational taxonomic units

LDA

THE Linear Discriminant Analysis

Author contributions

X.G, W.J.C. designed the study and drafted the manuscript; Y.Z, J.X, and J.H. L. performed animal experiments, statistical analysis and interpretation of the data; H.M.X, Y.T.L, Y.C.Z, N.J, Y.Y.L.P participated in animal experiments, recorded general status of experimental animals, collected samples and related tests; T.L.Z, Y.J.Z, Y.Q.N. interpreted the data and revised the manuscript; Y.Q.L, C.H, and Y.J.Z. designed and organized the study, interpreted the data and revised the manuscript. The author(s) read and approved the final manuscript.

Funding

This work was supported by the grants from the National Natural Science Foundation of China (82370552, 82200574), Natural Science Foundation of Guangdong Province (2023A1515030214), and Guangzhou Key Laboratory of Digestive Diseases (2022–2023) (KY17010003).

Data availability

The data used to support the findings of this study are available in NCBI-SRA under accession number PRJNA1265921 (https://www.ncbi.nlm.nih.gov/sra/?term=PRJNA1265921).

Declarations

Ethics statement

The animal studies (Grant No.:S-2022-218) were reviewed and approved by the Animal Ethics Committee of the Second Affiliated Hospital of South China University of Technology (Guangzhou First People’s Hospital).

Consent for publication

Yes.

Competing interests

The authors declare no competing interests.