Postbiotic Metabolites from a 31-Strain Lactobacillus/Bifidobacterium Co-Culture Attenuate DSS Colitis with Barrier- and Circadian-Linked Transcriptomic Signatures

Simple Summary

Ulcerative colitis is a chronic inflammatory disease of the large intestine that can cause diarrhea, bleeding, and abdominal pain. Safe, food-based approaches that help prevent flare-ups are strongly needed. In this study, we tested whether small molecules made during fermentation of soy by a mix of beneficial bacteria (without including any live bacteria) could protect the gut. Mice received this fermented soy extract before and during chemically induced colitis. Compared with untreated colitis mice, treated mice had milder disease symptoms, maintained longer colons, and showed clearer improvement in intestinal tissue damage under the microscope. Blood analyses indicated that immune messenger proteins shifted in a direction consistent with reduced inflammation. Gene activity patterns in the rectum suggested strengthened mucus and barrier functions and a calmer tissue repair state. In addition, the fermented soy extract changed the gut microbial community, increasing several bacteria often associated with gut health and reducing potentially harmful bacteria. These findings suggest that fermentation-derived bacterial metabolites may support intestinal barrier function and immune balance and may be useful as a preventive dietary strategy for maintaining gut health in inflammatory bowel conditions.

Abstract

Postbiotics produced by beneficial bacteria are emerging as safe dietary approaches to intestinal inflammation. We evaluated intestinal bacterial metabolites (IBM), a cell-free fermented soybean extract generated by co-culturing 31 Lactobacillus/Bifidobacterium-related strains, for prophylactic protection in 3% dextran sulfate sodium (DSS)-induced colitis. Male C57BL/6NJ mice received oral IBM (0.4 or 2 mL/kg/day) or vehicle for 7 days before and during 7 days of DSS. Disease activity index (DAI), colon length, and histopathology were assessed, and endpoint serum cytokines were quantified by a multiplex bead assay. DSS-independent responses were examined in healthy mice after 7 days of IBM by rectal RNA sequencing and cecal 16S rDNA profiling, and direct epithelial effects were tested in HCT-116 and DLD-1 cells treated with 2% IBM. IBM attenuated colitis, improving DAI, preventing colon shortening, and ameliorating histopathology, with decreased IL-23 and IL-17A and increased IFN-β and GM-CSF. Rectal transcriptomics showed modulation of circadian programs, upregulation of mucosal/barrier genes, and reduced extracellular-matrix remodeling signatures. IBM increased junctional proteins and barrier-related transcripts in vitro and shifted the microbiota, increasing Lactobacillus and Roseburia while decreasing Streptococcus and Staphylococcus. These coordinated clinical, immunological, transcriptomic, epithelial, and microbiome changes support prophylactic protection by IBM against DSS colitis.

1. Introduction

Inflammatory bowel disease (IBD), including ulcerative colitis (UC), is a chronic relapsing inflammatory disorder characterized by mucosal injury, impaired epithelial barrier function, and dysregulated immune responses to intestinal microbial and environmental cues. Despite advances in biologics and small molecules, many patients experience incomplete responses, loss of response, adverse effects, and substantial long-term disease burden, highlighting the need for complementary, safe, diet-based strategies that support mucosal homeostasis and help prevent exacerbations. A central concept emerging from both human and experimental studies is that intestinal inflammation is shaped by interactions among host immunity, epithelial barrier integrity, microbial communities, and microbe-derived metabolites, which together determine the inflammatory “tone” of the gut.

The intestinal epithelial barrier is a key determinant of susceptibility to and recovery from colitis. Barrier breakdown involves disruption of tight and adherens junction organization, altered epithelial renewal and differentiation, and loss or dysfunction of protective mucus layers, which collectively permit luminal antigens and bacteria to access the mucosa and amplify inflammation [1,2,3]. Junctional proteins such as ZO-1 and E-cadherin are integral to epithelial integrity and mucosal repair programs; their dysregulation has been linked to impaired healing in IBD and experimental inflammation [2,3]. In parallel, the mucus layer—composed largely of mucins and other secreted factors—serves as a first-line defense by spatially segregating microbes from the epithelium and supporting immunological tolerance [4,5]. Consistent with these concepts, pathways supporting epithelial secretion/absorption and barrier function (e.g., claudins and epithelial transport systems) and mucosal defense programs have gained attention as mechanistic and therapeutic nodes in colitis.

IBD pathogenesis also involves cytokine networks that coordinate innate and adaptive immune responses. Among these, the IL-23/Th17 axis is a well-established driver of intestinal inflammation, promoting the expansion and maintenance of Th17 responses and downstream effector cytokines including IL-17 family members, and it has been strongly implicated in both IBD mechanisms and therapeutic targeting [6,7,8]. Changes in inflammatory mediators detectable systemically can reflect and/or contribute to mucosal pathology; therefore, broad cytokine profiling approaches can provide an integrated view of immune modulation in experimental colitis.

In addition to immune and barrier pathways, the intestinal circadian clock has emerged as an additional layer of regulation relevant to IBD. Circadian programs shape epithelial turnover, barrier integrity, and mucosal immune tone, and circadian disruption has been linked to altered susceptibility to intestinal inflammation. In clinical cohorts, impaired sleep quality has been associated with increased risk of IBD flare, and short sleep duration has been linked to a higher risk of incident ulcerative colitis, supporting the relevance of circadian/sleep dysregulation to human disease [9,10,11].

Mechanistically, experimental studies indicate that perturbation of epithelial clock function can modulate colitis severity and inflammatory signaling. A recent report showed that genetic disruption of the core clock gene Bmal1 in intestinal epithelium reduces colonic inflammation in DSS colitis, highlighting direct epithelial-clock control of inflammatory outputs [12]. Conversely, disruption of colonic epithelial circadian rhythms has been reported to worsen DSS-induced colitis and to associate with changes in inflammatory mediators and microbiota-derived metabolites, emphasizing context-dependent roles of clock programs across cell types and perturbations [13]. Moreover, interventions that target intestinal circadian timing, such as meal-timing approaches, can ameliorate experimental intestinal inflammation, providing precedent that diet–microbe-derived cues may influence colitis partly via circadian-linked pathways [14].

Dietary fermentation products and microbe-derived metabolites are increasingly studied as modulators of intestinal homeostasis. “Postbiotics” have been defined by an International Scientific Association of Probiotics and Prebiotics (ISAPP) consensus as preparations of inanimate microorganisms and/or their components that confer a health benefit on the host [15]. Beyond this strict definition, many studies also evaluate cell-free, metabolite-rich preparations derived from beneficial bacteria and fermented foods, which can exert immunomodulatory and barrier-supporting effects while potentially avoiding concerns associated with administering live microbes to vulnerable populations. For example, probiotic-derived cell-free supernatants can directly influence intestinal epithelial and immune responses in vitro and in vivo, supporting the concept that microbial metabolites and secreted factors contribute substantially to host benefits [16]. Fermented soy products represent a particularly relevant class of functional foods because fermentation can alter soy constituents, generate bioactive metabolites, and influence the gut microbiota; recent reviews and experimental studies have highlighted anti-inflammatory potential and gastrointestinal impacts of fermented soy products [17]. Notably, several reports using fermented soymilk or extracts derived from lactic fermentation have demonstrated beneficial effects in DSS-induced colitis models, supporting the feasibility of fermentation-derived interventions in mucosal inflammation [18,19,20].

The dextran sulfate sodium (DSS) model remains one of the most widely used and reproducible experimental systems to study UC-like colitis. DSS exposure induces epithelial injury and barrier disruption, leading to acute mucosal inflammation that shares key pathological features with human disease, making it suitable for evaluating interventions that support epithelial integrity and modulate inflammation [21,22]. Importantly, because DSS is thought to exert direct toxicity to colonic epithelial cells, improvements in disease readouts may plausibly arise from barrier-protective actions, immune modulation, and/or microbiota-related effects [21,22].

In this context, we investigated the prophylactic effects of intestinal bacterial metabolites (IBM), a cell-free fermented soybean extract produced by co-culturing 31 strains of Lactobacillus/Bifidobacterium-related bacteria, in DSS-induced colitis. To better reflect the ecological nature of the gut microbiome, we used a defined multistrain co-culture rather than a single strain. In the intestine, beneficial functions often emerge from microbial networks through metabolic cross-feeding, and multistrain preparations can provide broader metabolic capacity and functional redundancy across diverse host/microbiota backgrounds. We designed a pre-treatment (prophylaxis) paradigm to assess whether IBM administration prior to and during DSS exposure modulates clinical and pathological indices of colitis. To obtain mechanistic insight, we combined endpoint serum cytokine profiling using a multiplex bead-based platform with rectal transcriptomics and pathway analyses following IBM administration in healthy mice. Because microbial metabolites may act both directly on the epithelium and indirectly through microbiota remodeling, we further examined IBM effects on epithelial junctional markers in human colorectal cell lines and analyzed cecal microbial community structure by 16S rDNA sequencing. Accordingly, our study was structured into three complementary experimental modules under a single objective: (i) a DSS-induced colitis cohort to establish prophylactic efficacy and systemic cytokine changes, (ii) a DSS-free healthy cohort to profile IBM-associated rectal transcriptomic signatures (including barrier- and circadian-linked programs) and microbiota remodeling without DSS-driven confounding, and (iii) in vitro epithelial assays to test direct effects of IBM on junctional proteins and barrier-related transcripts. Together, these modules were designed to clarify how IBM is associated with coordinated clinical, immunological, transcriptomic, epithelial, and microbiota changes relevant to mucosal homeostasis and prophylactic protection against experimental colitis.

2. Materials and Methods

2.1. Intestinal Bacterial Metabolites (IBM)

The postbiotic preparation used in this study was a fermented soybean extract, Fermented soybean extract SUPER SOPHIA (SOPHIA Co., Ltd., Tokyo, Japan). In this manuscript, the preparation is referred to as intestinal bacterial metabolites (IBM). For in vivo administration, IBM was diluted with sterile distilled water. For in vitro experiments, IBM was diluted with phosphate-buffered saline (PBS) to the indicated final concentrations. The multistrain co-culture was selected to maximize metabolic diversity and generate a reproducible, cell-free metabolite mixture. A list of the strains used for IBM production is provided in Supplementary Table S1. Metabolomic profiling of IBM was conducted by Human Metabolome Technologies, Inc. (Yamagata, Japan) and identified 787 compounds, including 529 peptides. The top 60 most abundant non-peptide compounds are summarized in Supplementary Table S2.

2.2. Animals and Housing Conditions

Male C57BL/6NJ mice (8 weeks old, total n = 36) were purchased from Jackson Laboratory Japan (Kanagawa, Japan). Mice were housed under a 12 h light/12 h dark cycle with free access to food and water and were acclimated for at least 1 week before experimentation. The animal room temperature was maintained at approximately 23 °C. All animal experiments were conducted in accordance with institutional guidelines and were approved by the Institutional Animal Care and Use Committee (IACUC) of Toyo University (Approval number: #2024-27).

2.3. Experimental Design for DSS-Induced Colitis

Dextran sulfate sodium (DSS)-induced colitis was established using DSS (MP Biomedicals, Irvine, CA, USA, 160110). DSS was dissolved in drinking water at a final concentration of 3% (w/v) and provided ad libitum for 7 consecutive days. DSS-containing water was freshly prepared and replaced every 2–3 days. Mice were randomly allocated into four groups (n = 6 per group): (1) Control (vehicle gavage; no DSS); (2) DSS + vehicle (vehicle gavage + 3% DSS); (3) DSS + IBM (0.4 mL/kg/day); (4) DSS + IBM (2 mL/kg/day).

IBM or vehicle (sterile distilled water) was administered twice daily by oral gavage using a feeding needle. IBM/vehicle administration started 7 days before DSS exposure and continued throughout the DSS period (total administration period: 14 days). Daily weight measurements were performed, and a weight loss of 20% or more was set as the humane endpoint.

2.4. Disease Activity Index (DAI)

Body weight was monitored daily and expressed as a percentage of the baseline body weight on Day 0. Disease severity was assessed using a disease activity index (DAI) calculated as the sum of three components (maximum total score: 9): body weight loss (0–3), stool consistency (0–3), and rectal bleeding (0–3).

Body weight loss was scored as: 0, <1% loss; 1, 1–5% loss; 2, 5–10% loss; 3, ≥10% loss. Stool consistency was scored as: 0, normal; 1, soft but firm; 2, soft; 3, diarrhea. Rectal bleeding (blood in stool/rectal hemorrhaging) was scored as: 0, no blood; 1, mild blood; 2, obvious blood; 3, severe blood.

2.5. Necropsy and Colon Length Measurement

At the endpoint, mice were euthanized and blood was collected prior to tissue harvesting. The entire colon was excised, gently flushed with PBS, and the colon length was measured using calipers. For histological analyses, colonic tissue was fixed in 4% paraformaldehyde in phosphate buffer (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan).

2.6. Histology (Hematoxylin and Eosin Staining)

Fixed colon tissues were processed into formalin-fixed paraffin-embedded (FFPE) blocks and sectioned for hematoxylin and eosin (H&E) staining by Sapporo General Pathology Laboratory (Sapporo, Japan). Stained sections were imaged using a Zeiss microscope (Vert.A1 FL-LED, Carl Zeiss, Oberkochen, Germany). Scale bars (200 μm) are indicated in the figure panels as appropriate.

2.7. Serum Collection and Multiplex Cytokine Profiling

Collected blood was allowed to clot, followed by centrifugation at 3000 rpm for 15 min at room temperature to obtain serum. Serum cytokines were quantified using LEGENDplex™ Mouse Inflammation Panel (13-plex) with V-bottom Plate (BioLegend, San Diego, CA, USA; 740446) according to the manufacturer’s instructions. Samples were acquired on an Attune NxT flow cytometer (Thermo Fisher Scientific, Waltham, MA, USA; blue and red lasers), and cytokine concentrations were calculated using LEGENDplex™ Cloud-based Data Analysis Software (v2025-05-01) (Qognit, Inc., Santa Clara, CA, USA).

2.8. Rectal Tissue Collection and RNA Extraction for RNA Sequencing

To investigate DSS-independent responses under steady-state conditions, a separate cohort of healthy male C57BL/6NJ mice was used. This design was chosen to identify IBM-associated rectal transcriptional changes while minimizing the confounding effects of DSS-induced epithelial injury and inflammation on gene expression. Healthy male C57BL/6NJ mice were administered IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6 per group). At necropsy, ~2 mm of rectal tissue was collected and placed in RNAlater (Invitrogen, Thermo Fisher Scientific). Total RNA was extracted using FastGene RNA Premium Kit (FastGene, FG-81050, Tokyo, Japan) according to the manufacturer’s protocol. RNA concentration was measured using DeNovix DS-11 FX (DeNovix, Wilmington, DE, USA), and RNA integrity was confirmed with an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA); all samples used for library preparation had RIN = 10.

2.9. Library Preparation, Sequencing, and Differential Expression Analysis

RNA-seq libraries were prepared using the NEBNext Ultra II RNA Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA) according to the manufacturer’s protocol. Sequencing was performed on an Illumina NovaSeq X Plus system (Illumina, San Diego, CA, USA) using paired-end reads with a target yield of approximately 6 Gb per sample.

Adapter trimming, quality trimming, and downsampling were conducted using FASTQ Toolkit (v2.2.6). For downsampling, 20 million reads per sample were randomly selected. Secondary transcriptome analysis was performed using DRAGEN Differential Expression (v4.3.7) (Illumina) with mm10 as the reference genome. Differential expression was assessed within the DESeq2 framework (DESeq2 v1.38.3); for each gene, p-values were calculated using the Wald test based on raw counts. For exploratory visualization, PCA was conducted using DESeq2 on variance-stabilized (VST) gene expression values generated from raw count matrices (genes), and group separation was assessed on PC1 and PC2.

2.10. Cecal Microbiota Analysis by 16S rDNA Amplicon Sequencing

To evaluate IBM-associated microbiota changes under DSS-free conditions, a separate cohort of healthy male C57BL/6NJ mice was used. This design was chosen to assess DSS-independent microbiota responses to IBM and to minimize confounding effects of DSS-induced epithelial injury and inflammation on microbial community structure. Healthy male C57BL/6NJ mice were administered IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6 per group). IBM was given by gavage twice daily (morning and evening) to achieve the daily dose. The following day, mice were euthanized and cecal contents were collected. 16S rDNA amplicon sequencing was outsourced to TechnoSuruga Laboratory Co., Ltd. (Shizuoka, Japan).

Sequencing was performed on the Illumina MiSeq i100 Series platform (Illumina). Primer sequences were removed using Cutadapt (v3.5), and paired-end reads were joined using fastq-join (v1.3.1). Quality filtering and representative sequence inference were performed using QIIME 2 (v2020.6) with dada2 (v1.10.0). Taxonomic assignment was conducted against the SILVA database (v138). Diversity and statistical analyses were performed in QIIME 2, and visualization was conducted in R (v3.4.1) using qiime2R (v0.99.13) and tidyverse (v1.2.1). Genus-level relative abundances from 16S rDNA sequencing were CLR-transformed after adding a small pseudocount. Rectal RNA-seq expression values were visualized using rlog-transformed counts. Spearman’s rank correlations were computed between selected genera and transcripts across individual mice, and p-values were adjusted for multiple testing using the Benjamini–Hochberg method. Correlation matrices were visualized as clustered heatmaps.

2.11. Cell Culture and IBM Treatment In Vitro

Human colorectal cancer cell lines HCT-116 (ATCC, Manassas, VA, USA; CCL-247) and DLD-1 (ATCC; CCL-221) were cultured in RPMI 1640 (Nacalai Tesque, Kyoto, Japan; 5176-25) supplemented with 10% fetal bovine serum (Gibco, Thermo Fisher Scientific; 10270106, lot 42F7201K), 1% penicillin–streptomycin (FUJIFILM Wako; 168-23191), 1% GlutaMAX (Gibco; 35050-061), and 1% sodium pyruvate (100 mM) (Gibco; 11360-070). Cells were maintained at 37 °C in a humidified incubator with 5% CO₂. Rationale for cell line selection. We selected HCT-116 and DLD-1 as two widely used human colorectal epithelial cell lines with distinct genetic backgrounds to test whether IBM-induced barrier-related responses are robust across different CRC genotypes. HCT-116 is characterized by microsatellite instability (MSI) and activating KRAS mutation, whereas DLD-1 is a microsatellite-stable (MSS) line harboring APC mutation and KRAS mutation. Using both lines allowed us to evaluate IBM effects on epithelial junction/barrier markers and inflammatory transcriptional outputs across divergent differentiation and signaling contexts.

For IBM treatment, cells were seeded and allowed to adhere overnight. IBM was added to a final concentration of 2% (v/v). For Western blotting, cells were treated for 24 h. For qPCR, cells were treated for 6 h.

2.12. Western Blotting

Protein lysates were prepared using RIPA Lysis Buffer System (Santa Cruz Biotechnology, Dallas, TX, USA; SC-24948) according to the manufacturer’s recommendations. Protein concentrations were quantified using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific; 23225). Equal amounts of protein were separated by SDS-PAGE and transferred to PVDF membranes using iBlot™ 3 Transfer Stacks, mini, PVDF (Thermo Fisher Scientific; IB34002). Membranes were blocked with PVDF Blocking Reagent for Can Get Signal (TOYOBO, Osaka, Japan; NYPBR01) for 30 min at room temperature.

Primary antibodies were diluted 1:1000 in Can Get Signal Solution 1 (TOYOBO) and incubated for 1 h at room temperature: anti-ZO-1 (Cell Signaling Technology (CST), Danvers, MA, USA; #13663), anti-E-cadherin (CST; #3195), anti-β-catenin (CST; #8480), and anti-β-actin (CST; #4967). The secondary antibody was anti-rabbit IgG HRP-linked (CST; #7074) diluted 1:2000 in Can Get Signal Solution 2 (TOYOBO) and incubated for 1 h at room temperature. Signals were developed using Pierce™ ECL Western Blotting Substrate (Thermo Fisher Scientific; 32209) and imaged with iBright CL1000 (Thermo Fisher Scientific). Band intensities were quantified using ImageJ v1.53e (NIH, Bethesda, MD, USA).

2.13. RNA Extraction and Quantitative PCR (qPCR)

Total RNA was extracted using ReliaPrep™ RNA Miniprep Systems (Promega, Madison, WI, USA; Z6010 according to the manufacturer’s protocol). RNA was quantified using DeNovix DS-11 FX. Reverse transcription was performed using PrimeScript RT reagent Kit (TaKaRa, Kusatsu, Japan; RR037A). qPCR was conducted using TB Green Premix Ex Taq II (TaKaRa; RR820A) on a TaKaRa TP1010 real-time PCR system. Gene expression was calculated using the 2^−ΔΔCt method with RPLP0 as an internal control. Primer sequences are listed in Supplementary Table S3. All experiments were performed in triplicate.

2.14. Statistical Analysis

Data are presented as mean ± SD unless otherwise stated. For comparisons between two groups, Welch’s t-test (two-tailed) was used. For comparisons among multiple groups, one-way ANOVA was performed, followed by Dunnett’s multiple comparisons test using the DSS + vehicle group as the reference. For longitudinal body weight data, a two-way repeated-measures ANOVA (group × time) was used. When variance heterogeneity was detected, Welch’s ANOVA was applied with appropriate post hoc testing. p < 0.05 was considered significant. For microbiota β-diversity, group differences in community structure were assessed using ANOSIM based on Bray–Curtis distances. Differential expression analysis for RNA-seq was performed within the DESeq2 framework as described above. A p-value < 0.05 was considered statistically significant. Statistical analyses and graph generation were performed using R v3.4.1 and Microsoft Excel v16.105.1 (Microsoft Corporation, Redmond, WA, USA).

3. Results

To improve clarity of the overall experimental logic, the Results are presented according to the three predefined experimental modules/cohorts. First, we report the in vivo prophylactic efficacy of IBM in the DSS-induced colitis cohort together with endpoint serum cytokine profiling (Section 3.1 and Section 3.2; Figure 1 and Figure 2). Next, we present DSS-free host and microbiota profiling in a separate healthy cohort to identify IBM-associated rectal transcriptomic signatures and microbiota remodeling under steady-state conditions (Section 3.3 and Section 3.4; Figure 3 and Figure 4). Finally, we describe in vitro epithelial experiments assessing direct effects of IBM on junctional proteins and barrier-related gene expression in colorectal epithelial-like cell lines (Section 3.5; Figure 5).

3.1. IBM Prophylaxis Ameliorated DSS-Induced Colitis Severity Without Altering Body Weight (DSS Colitis Cohort)

To evaluate the prophylactic potential of IBM, male C57BL/6NJ mice were administered vehicle or IBM (0.4 or 2 mL/kg/day) for 7 days prior to 3% DSS exposure and throughout the DSS period (Figure 1A). During the experimental period (Day 0–14), body weight trajectories were comparable among DSS-exposed groups, and IBM administration did not significantly affect body weight (Figure 1B).

In contrast, DSS exposure induced macroscopic and clinical signs of colitis that were mitigated by IBM. Colon length was shortened in DSS-treated mice, whereas IBM administration significantly attenuated DSS-associated colon shortening (Figure 1C). Consistently, disease activity index (DAI) scores were reduced by IBM, and the high-dose IBM group (2 mL/kg/day) showed a significant decrease compared with the DSS + vehicle group (Figure 1D). Histological assessment of H&E-stained colon sections revealed pronounced mucosal damage and inflammatory features in DSS-treated mice, including epithelial disruption and thickening of the muscular layer, which were visibly alleviated in IBM-treated groups (Figure 1E). Together, these findings indicate that IBM prophylaxis is associated with improvement of colitis severity indices (DAI, colon length, and histopathology) despite no detectable differences in body weight loss patterns under the present conditions.

Figure 1. IBM prophylaxis ameliorates DSS-induced colitis severity. (A) Experimental scheme and study modules. The study consisted of three predefined experimental modules. Module 1: DSS colitis cohort (Figure 1 and Figure 2)—male C57BL/6NJ mice (8 weeks old) received IBM (0.4 or 2 mL/kg/day) or vehicle by oral gavage twice daily starting 7 days before DSS exposure and continuing during DSS treatment; colitis was induced by 3% DSS in drinking water for 7 days (total dosing period, 14 days; necropsy at Day 14). Module 2: Healthy DSS-free omics (Figure 3 and Figure 4)—a separate cohort of healthy mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (necropsy at Day 7) for rectal transcriptomics and cecal microbiota profiling. Module 3: In vitro epithelial model (Figure 5)—HCT-116 and DLD-1 cells were treated with 2% IBM, followed by mRNA analysis at 6 h and protein analysis at 24 h. Colitis was induced by 3% DSS in drinking water for 7 days (Day 7–14), followed by necropsy at Day 14. (B) Body weight changes from Day 0 to Day 14. (C) Colon length at necropsy (Day 14). (D) Disease activity index (DAI) at Day 14. DAI was calculated as the sum of body weight loss (0–3), stool consistency (0–3), and rectal bleeding (0–3) (maximum 9). Body weight loss: 0, <1%; 1, 1–5%; 2, 5–10%; 3, ≥10%. Stool consistency: 0, normal; 1, soft but firm; 2, soft; 3, diarrhea. Rectal bleeding: 0, no blood; 1, mild blood; 2, obvious blood; 3, severe blood. (E) Representative H&E-stained colon sections from each group at Day 14. Scale bar, 200 μm. Groups: Control (vehicle, no DSS), DSS + vehicle, DSS + IBM (0.4 mL/kg/day), DSS + IBM (2 mL/kg/day). Data are shown as mean ± SE (n = 6/group). Statistical analyses: (B) two-way repeated-measures ANOVA; (C,D) one-way ANOVA with Dunnett’s multiple comparisons test using the DSS + vehicle group as the reference. * p < 0.05, *** p < 0.001 vs. DSS + vehicle.

Figure 1. IBM prophylaxis ameliorates DSS-induced colitis severity. (A) Experimental scheme and study modules. The study consisted of three predefined experimental modules. Module 1: DSS colitis cohort (Figure 1 and Figure 2)—male C57BL/6NJ mice (8 weeks old) received IBM (0.4 or 2 mL/kg/day) or vehicle by oral gavage twice daily starting 7 days before DSS exposure and continuing during DSS treatment; colitis was induced by 3% DSS in drinking water for 7 days (total dosing period, 14 days; necropsy at Day 14). Module 2: Healthy DSS-free omics (Figure 3 and Figure 4)—a separate cohort of healthy mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (necropsy at Day 7) for rectal transcriptomics and cecal microbiota profiling. Module 3: In vitro epithelial model (Figure 5)—HCT-116 and DLD-1 cells were treated with 2% IBM, followed by mRNA analysis at 6 h and protein analysis at 24 h. Colitis was induced by 3% DSS in drinking water for 7 days (Day 7–14), followed by necropsy at Day 14. (B) Body weight changes from Day 0 to Day 14. (C) Colon length at necropsy (Day 14). (D) Disease activity index (DAI) at Day 14. DAI was calculated as the sum of body weight loss (0–3), stool consistency (0–3), and rectal bleeding (0–3) (maximum 9). Body weight loss: 0, <1%; 1, 1–5%; 2, 5–10%; 3, ≥10%. Stool consistency: 0, normal; 1, soft but firm; 2, soft; 3, diarrhea. Rectal bleeding: 0, no blood; 1, mild blood; 2, obvious blood; 3, severe blood. (E) Representative H&E-stained colon sections from each group at Day 14. Scale bar, 200 μm. Groups: Control (vehicle, no DSS), DSS + vehicle, DSS + IBM (0.4 mL/kg/day), DSS + IBM (2 mL/kg/day). Data are shown as mean ± SE (n = 6/group). Statistical analyses: (B) two-way repeated-measures ANOVA; (C,D) one-way ANOVA with Dunnett’s multiple comparisons test using the DSS + vehicle group as the reference. * p < 0.05, *** p < 0.001 vs. DSS + vehicle.

3.2. Serum Cytokinomics Indicated IBM-Associated Modulation of Inflammatory Mediators in DSS Colitis (DSS Colitis Cohort)

To broadly assess systemic inflammatory signals at the endpoint, serum cytokines were quantified using a bead-based multiplex assay. Compared with the DSS + vehicle group, IBM administration was associated with reduced IL-23 and IL-17A and increased IFN-β and GM-CSF (Figure 2A). These cytokine changes are consistent with IBM-associated modulation of pathways linked to intestinal inflammation and immune regulation in DSS colitis. To examine whether the IBM-associated cytokine profile relates to clinical disease severity under DSS exposure, we assessed the association between endpoint serum cytokine concentrations and the disease activity index (DAI) across DSS-exposed mice (DSS + vehicle, DSS + IBM 0.4 mL/kg/day, and DSS + IBM 2 mL/kg/day). Spearman correlation analysis showed modest positive correlations of DAI with IL-23 (ρ = 0.49) and IL-17A (ρ = 0.41), whereas IFN-β and GM-CSF exhibited inverse associations with DAI (ρ = −0.39 and ρ = −0.22, respectively) (Figure 2B). These relationships are consistent with higher IL-23/IL-17A levels tracking with greater clinical severity, while higher IFN-β/GM-CSF levels track with lower severity in this DSS setting. Given the limited sample size, these correlations should be interpreted as supportive associations rather than definitive predictors of severity.

Figure 2. Serum cytokine profiling reveals IBM-associated modulation of inflammatory mediators in DSS colitis. Endpoint serum cytokines were quantified using a bead-based multiplex assay (LEGENDplex™ Mouse Inflammation Panel (13-plex), BioLegend) and measured on an Attune NxT flow cytometer. (A) Bar graphs show serum concentrations of IL-23, IL-17A, IFN-β, and GM-CSF at Day 14 in DSS-exposed groups. (B) Scatter plots show the relationship between DAI and serum levels of IL-23, IL-17A, IFN-β, and GM-CSF measured at the experimental endpoint by a multiplex bead-based assay. Each point represents an individual mouse from DSS-exposed groups (DSS + vehicle, DSS + IBM 0.4 mL/kg/day, DSS + IBM 2 mL/kg/day). Spearman’s rank correlation coefficient (ρ) is indicated in each panel. The dotted line denotes a fitted trend line for visualization. Data are shown as mean ± SE (n = 6/group).

Figure 2. Serum cytokine profiling reveals IBM-associated modulation of inflammatory mediators in DSS colitis. Endpoint serum cytokines were quantified using a bead-based multiplex assay (LEGENDplex™ Mouse Inflammation Panel (13-plex), BioLegend) and measured on an Attune NxT flow cytometer. (A) Bar graphs show serum concentrations of IL-23, IL-17A, IFN-β, and GM-CSF at Day 14 in DSS-exposed groups. (B) Scatter plots show the relationship between DAI and serum levels of IL-23, IL-17A, IFN-β, and GM-CSF measured at the experimental endpoint by a multiplex bead-based assay. Each point represents an individual mouse from DSS-exposed groups (DSS + vehicle, DSS + IBM 0.4 mL/kg/day, DSS + IBM 2 mL/kg/day). Spearman’s rank correlation coefficient (ρ) is indicated in each panel. The dotted line denotes a fitted trend line for visualization. Data are shown as mean ± SE (n = 6/group).

3.3. Rectal Transcriptomics After IBM Administration in Healthy Mice Revealed Modulation of Circadian-Related Genes and Barrier-Associated Programs (Healthy Cohort; DSS-Free)

To explore DSS-independent host responses that could underpin prophylactic effects, rectal RNA sequencing was performed after 7 days of IBM administration (2 mL/kg/day) in healthy mice. Because DSS exposure itself induces profound transcriptional and microbiota perturbations, these omics analyses were conducted in healthy mice to capture IBM-associated signatures without DSS-driven confounding. PCA based on variance-stabilized (VST) expression values showed clear separation between the IBM and vehicle groups, indicating a robust global transcriptomic shift following IBM administration (Figure 3A). Unsupervised visualization of the top differentially expressed genes demonstrated a clear IBM-associated transcriptional shift (Figure 3B). PCA was performed using VST values for global structure, whereas rlog values are shown for visualization of selected genes. Notably, multiple circadian-related genes were altered, including increased expression of Dbp, Nr1d1/Nr1d2, Tef, and Hlf, alongside reduced expression of Arntl (Bmal1) (Figure 3B), suggesting IBM-associated modulation of rectal circadian-linked gene programs under controlled dosing and sampling conditions.

We next focused on genes relevant to mucosal barrier function and tissue remodeling. IBM administration increased the expression of epithelial/mucosal-associated genes, including Muc1, Zg16, Guca2a, Cldn15, Cldn3, Slc9a3, and Aqp1, while decreasing expression of genes linked to injury-associated epithelial signaling and extracellular matrix remodeling, including Areg, Ereg, Dusp6, Adamts4, Adamts9, Adamts15, Col1a1, Col1a2, and Fn1 (Figure 3C). Collectively, the rectal transcriptome profile is consistent with IBM-associated enhancement of mucosal/barrier-related features and attenuation of remodeling-associated signatures in the steady state.

Figure 3. Rectal transcriptomics after IBM administration in healthy mice. Healthy male C57BL/6NJ mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6/group), followed by rectal RNA sequencing. (A) Principal component analysis (PCA) of rectal RNA-seq profiles from Control and IBM-treated mice. PCA was performed on variance-stabilized (VST) gene expression values derived from raw count data (DESeq2). Each point represents one mouse. Percent variance explained by each principal component is indicated on the axes. (B) Heatmap of the top 30 differentially expressed genes between vehicle and IBM groups. Each column represents an individual mouse. (C) Box-and-whisker plots of rlog-transformed expression values for selected genes related to mucosal barrier/transport and inflammation/remodeling: Muc1, Zg16, Guca2a, Cldn15, Cldn3, Slc9a3, Aqp1, Areg, Ereg, Dusp6, Adamts4, Adamts9, Adamts15, Col1a1, Col1a2, and Fn1. Boxes indicate the interquartile range with median; whiskers indicate the range. Differential expression p-values were calculated using the Wald test within the DESeq2 framework based on raw counts, while rlog values are shown for visualization. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 3. Rectal transcriptomics after IBM administration in healthy mice. Healthy male C57BL/6NJ mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6/group), followed by rectal RNA sequencing. (A) Principal component analysis (PCA) of rectal RNA-seq profiles from Control and IBM-treated mice. PCA was performed on variance-stabilized (VST) gene expression values derived from raw count data (DESeq2). Each point represents one mouse. Percent variance explained by each principal component is indicated on the axes. (B) Heatmap of the top 30 differentially expressed genes between vehicle and IBM groups. Each column represents an individual mouse. (C) Box-and-whisker plots of rlog-transformed expression values for selected genes related to mucosal barrier/transport and inflammation/remodeling: Muc1, Zg16, Guca2a, Cldn15, Cldn3, Slc9a3, Aqp1, Areg, Ereg, Dusp6, Adamts4, Adamts9, Adamts15, Col1a1, Col1a2, and Fn1. Boxes indicate the interquartile range with median; whiskers indicate the range. Differential expression p-values were calculated using the Wald test within the DESeq2 framework based on raw counts, while rlog values are shown for visualization. * p < 0.05, ** p < 0.01, *** p < 0.001.

3.4. IBM Altered Gut Microbiota Community Structure and Shifted Specific Taxa After 7-Day Administration (Healthy Cohort; DSS-Free)

Because IBM may also exert effects via microbiota remodeling, cecal contents were analyzed after 7 days of IBM administration (2 mL/kg/day) in healthy mice. β-diversity analysis based on Bray–Curtis distances demonstrated a significant IBM-associated shift in community structure, as supported by ANOSIM (Figure 4A). Principal coordinate analysis further showed separation of the IBM and control groups, indicating distinct microbial community configurations after IBM intake (Figure 4B).

Taxonomic profiling identified multiple taxa with altered relative abundance between groups (Figure 4C). At the genus level, IBM administration was associated with increased abundance of several taxa often considered beneficial, including Lactobacillus, Roseburia, Acetitomaculum, and Blautia, and decreased abundance of taxa regarded as potentially unfavorable in this context, including Staphylococcus and Streptococcus (Figure 4D). These findings demonstrate that IBM intake is associated with measurable remodeling of gut microbial communities, which may contribute to its prophylactic effects in DSS colitis. To integrate the microbiota shifts with host transcriptional responses, we performed a cross-omics correlation analysis across individual mice (n = 12), relating genus-level abundances to rectal rlog-transformed gene expression (Figure 4E). Because this correlation matrix involves multiple comparisons, it is presented as an exploratory integrative analysis to highlight coherent association patterns for follow-up validation. After the centered log-ratio (CLR) transformation of genus abundances, Spearman correlation analysis revealed two opposing association patterns. IBM-enriched genera, including Roseburia (and Blautia), showed positive correlations with circadian regulators (Dbp, Nr1d1/2, Tef, Hlf) and mucosal/transport-associated transcripts (Slc9a3, Guca2a, Zg16, Aqp1), whereas genera reduced by IBM, including Streptococcus and Staphylococcus, tended to correlate with epithelial stress/remodeling-related genes (Areg/Ereg, Adamts4/9, Fn1, Col1a1/2). These results indicate coordinated microbiota–host expression associations consistent with an IBM-linked shift toward mucosal homeostasis.

Figure 4. IBM alters gut microbiota community structure and shifts specific taxa. Healthy male C57BL/6NJ mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6/group). Cecal contents were collected and subjected to 16S rDNA amplicon sequencing, followed by QIIME 2–based analyses. (A) β-diversity comparison using Bray–Curtis distances assessed by ANOSIM (box plot). (B) Three-dimensional principal coordinate analysis (PCA) plot based on Bray–Curtis distances, showing separation between vehicle (blue) and IBM (red) groups; each point represents one mouse. (C) Species-level relative abundance profiles (relative frequency) displayed as box plots; the 28 most abundant species in the vehicle group are listed in descending order. (D) Selected taxa shown as the ratio relative to the vehicle group, highlighting increases in putatively beneficial genera (Lactobacillus, Roseburia, Acetitomaculum, Blautia) and decreases in potentially unfavorable taxa (Staphylococcus, Streptococcus). (E) Heatmap showing Spearman’s rank correlation coefficients (ρ) between genus-level microbial abundances in cecal contents and rectal RNA-seq expression values across individual mice (n = 12; Control vs. IBM). Genus abundances were centered log-ratio (CLR) transformed (after addition of a small pseudocount), and host transcript levels are shown as rlog-transformed expression values. Rows indicate selected genera and columns indicate selected transcripts related to circadian regulation, mucosal barrier/transport, and inflammation/remodeling. Hierarchical clustering was applied to both genera and transcripts. Color scale denotes ρ (red, positive correlation; blue, negative correlation). Correlations reflect inter-individual associations and are hypothesis-generating rather than causal. For (A), statistical significance was evaluated by ANOSIM. For (D), between-group comparisons were evaluated using Welch’s t-test (two-tailed). * p < 0.05, ** p < 0.01.

Figure 4. IBM alters gut microbiota community structure and shifts specific taxa. Healthy male C57BL/6NJ mice received IBM (2 mL/kg/day) or vehicle by oral gavage for 7 days (n = 6/group). Cecal contents were collected and subjected to 16S rDNA amplicon sequencing, followed by QIIME 2–based analyses. (A) β-diversity comparison using Bray–Curtis distances assessed by ANOSIM (box plot). (B) Three-dimensional principal coordinate analysis (PCA) plot based on Bray–Curtis distances, showing separation between vehicle (blue) and IBM (red) groups; each point represents one mouse. (C) Species-level relative abundance profiles (relative frequency) displayed as box plots; the 28 most abundant species in the vehicle group are listed in descending order. (D) Selected taxa shown as the ratio relative to the vehicle group, highlighting increases in putatively beneficial genera (Lactobacillus, Roseburia, Acetitomaculum, Blautia) and decreases in potentially unfavorable taxa (Staphylococcus, Streptococcus). (E) Heatmap showing Spearman’s rank correlation coefficients (ρ) between genus-level microbial abundances in cecal contents and rectal RNA-seq expression values across individual mice (n = 12; Control vs. IBM). Genus abundances were centered log-ratio (CLR) transformed (after addition of a small pseudocount), and host transcript levels are shown as rlog-transformed expression values. Rows indicate selected genera and columns indicate selected transcripts related to circadian regulation, mucosal barrier/transport, and inflammation/remodeling. Hierarchical clustering was applied to both genera and transcripts. Color scale denotes ρ (red, positive correlation; blue, negative correlation). Correlations reflect inter-individual associations and are hypothesis-generating rather than causal. For (A), statistical significance was evaluated by ANOSIM. For (D), between-group comparisons were evaluated using Welch’s t-test (two-tailed). * p < 0.05, ** p < 0.01.

3.5. IBM Directly Increased Junctional Proteins and Induced Barrier-Related Transcripts in Colorectal Epithelial-like Cell Lines (In Vitro Epithelial Model)

To determine whether IBM can act directly on epithelial-like cells, HCT-116 and DLD-1 cells were treated with IBM (2%). Western blotting revealed increased levels of junctional proteins, including ZO-1, E-cadherin, and β-catenin, following IBM exposure compared with controls (Figure 5A), and densitometric analyses confirmed these increases across independent experiments (Figure 5B). Uncropped immunoblot images corresponding to Figure 5A are provided in Supplementary Figure S1.

In addition, short-term IBM treatment (6 h) induced barrier-related transcripts in both cell lines. qPCR analysis showed increased expression of OCLN, CLDN1, and CLDN3 in HCT-116 (Figure 5C) and DLD-1 cells (Figure 5D). AREG transcript levels were also increased in vitro in both cell lines after IBM exposure (Figure 5C,D). These results support a direct epithelial response to IBM that includes enhanced junctional marker abundance and induction of barrier-related gene expression, while AREG increased in vitro, rectal tissue transcriptomics showed reduced Areg/Ereg signatures (Figure 3C), suggesting that regulation of this axis may differ between the in vitro epithelial context and the intact tissue environment.

Figure 5. IBM directly increases junctional proteins and induces barrier-related transcripts in colorectal cell lines. (A) Representative immunoblots of ZO-1, E-cadherin, β-catenin, and β-actin in HCT-116 and DLD-1 cells treated with vehicle or IBM (2%) for 24 h. β-actin served as a loading control. (B) Densitometric quantification of ZO-1, E-cadherin, and β-catenin normalized to β-actin. (C,D) qPCR analysis of OCLN, CLDN1, CLDN3, and AREG in (C) HCT-116 and (D) DLD-1 cells treated with vehicle or IBM (2%) for 6 h. Expression levels were calculated using the 2^−ΔΔCt method with RPLP0 as an internal control. All experiments were performed as three independent biological replicates (n = 3), and data are presented as mean ± SD. Statistical significance was determined using Welch’s t-test (two-tailed). * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 5. IBM directly increases junctional proteins and induces barrier-related transcripts in colorectal cell lines. (A) Representative immunoblots of ZO-1, E-cadherin, β-catenin, and β-actin in HCT-116 and DLD-1 cells treated with vehicle or IBM (2%) for 24 h. β-actin served as a loading control. (B) Densitometric quantification of ZO-1, E-cadherin, and β-catenin normalized to β-actin. (C,D) qPCR analysis of OCLN, CLDN1, CLDN3, and AREG in (C) HCT-116 and (D) DLD-1 cells treated with vehicle or IBM (2%) for 6 h. Expression levels were calculated using the 2^−ΔΔCt method with RPLP0 as an internal control. All experiments were performed as three independent biological replicates (n = 3), and data are presented as mean ± SD. Statistical significance was determined using Welch’s t-test (two-tailed). * p < 0.05, ** p < 0.01, *** p < 0.001.

4. Discussion

This study demonstrates that prophylactic oral administration of a cell-free intestinal bacterial metabolite preparation (IBM) derived from a 31-strain Lactobacillus/Bifidobacterium co-culture mitigates DSS-induced colitis severity, as evidenced by improved colon length, reduced DAI, and attenuated histopathology, together with systemic cytokine shifts, rectal transcriptomic reprogramming, and marked microbiome remodeling. Conceptually, IBM aligns with the emerging “postbiotic” paradigm—health effects mediated by microbial products rather than live organisms—consistent with the ISAPP consensus definition and scope of postbiotics [15].

4.1. Positioning IBM Within UC Pathobiology and the DSS Model

UC is characterized by chronic mucosal inflammation linked to epithelial barrier impairment, dysregulated innate/adaptive immune activation, and altered host–microbe interactions. DSS colitis is particularly useful for probing barrier-protective and mucosa-stabilizing interventions because disease initiation is driven by epithelial injury and increased permeability, followed by innate immune activation and secondary inflammatory cascades [21,22,23]. In our study, IBM prophylaxis improved local disease indices (DAI, colon shortening, histological injury) without a detectable difference in body-weight trajectory, a pattern consistent with interventions that primarily reduce mucosal damage/inflammation rather than systemic sickness behavior in acute DSS settings [21,22,23].

4.2. IBM Shifts Systemic Cytokine Tone Away from IL-23/Th17-Skewed Inflammation

The bead-array cytokinomics identified decreased serum IL-23 and IL-17A in IBM-treated mice. The IL-23 pathway is central to IBD pathogenesis, supporting pathogenic Th17/innate lymphoid programs and sustaining neutrophil-dominated inflammation [7,24,25]. Clinically, the therapeutic success of IL-23 pathway blockade in UC (e.g., ustekinumab, mirikizumab, guselkumab, and risankizumab) underscores the translational relevance of directionality consistent with our findings [26,27,28,29]. Importantly, IL-17 biology in the gut is context dependent: direct IL-17 pathway blockade can worsen Crohn’s disease or precipitate IBD-like disease in susceptible contexts, despite IL-23 blockade being broadly beneficial [7,30,31,32]. Thus, our serum data are most conservatively interpreted as IBM dampening the upstream IL-23 inflammatory circuit and/or reducing tissue damage that fuels IL-23/IL-17 amplification, rather than suggesting that indiscriminate IL-17 suppression is universally protective [7,24,25,30,31,32].

4.3. Increased IFN-β and GM-CSF May Reflect Protective Innate Programming

IBM increased serum IFN-β and GM-CSF at the DSS endpoint. Type I interferons can confer tissue-protective effects in experimental colitis by modulating myeloid activation, epithelial stress responses, and mucosal repair programs, though clinical outcomes with exogenous IFN-β have been mixed—highlighting the importance of timing, disease context, and dosing [33,34,35]. Likewise, GM-CSF can support mucosal innate defense and myeloid competence; recombinant GM-CSF (sargramostim) has shown clinical activity in Crohn’s disease, and GM-CSF–related immune phenotypes (including anti-GM-CSF autoantibodies) have been linked to barrier dysfunction and complicated disease courses [36,37,38,39]. While we cannot assign causality from serum alone, the combined cytokine pattern is compatible with IBM promoting a “less IL-23/Th17-driven” systemic milieu while supporting innate programs that may enhance mucosal resilience.

4.4. Rectal Transcriptomics Support Epithelial Stabilization and Reduced Injury-Remodeling Tone

RNA-seq after 7 days of IBM alone (pre-DSS) revealed coordinated upregulation of mucosal/epithelial genes (Muc1, Zg16, Guca2a, Cldn3/Cldn15, Slc9a3, Aqp1) and downregulation of genes associated with injury-induced repair signaling and extracellular matrix (ECM) remodeling (Areg, Ereg, Dusp6, Adamts4/9/15, Col1a1/Col1a2, Fn1). These changes align with a model in which IBM preconditions the mucosa toward a barrier-competent baseline and reduces pro-injury regenerative/mesenchymal remodeling programs that are typically elevated during active inflammation and wound repair [40,41,42,43]. Barrier integrity is regulated by tight/adherens junctions, mucus, epithelial transport programs, and immune–epithelial crosstalk; disruption of these systems is a key amplifier of IBD pathogenesis [40,41,42]. Notably, EGFR ligands (amphiregulin/epiregulin) are commonly induced in epithelial injury states; their reduction in IBM-only animals is consistent with a lower “injury-repair” tone rather than impaired repair capacity per se [44,45]. In parallel, reduced collagen/fibronectin and ADAMTS remodeling signatures suggest dampened ECM turnover/mesenchymal activation, which can accompany attenuation of inflammation-associated tissue remodeling [43].

4.5. Circadian Gene Reprogramming as a Plausible Integrator of Metabolism, Barrier, and Inflammation

Among the top IBM-responsive rectal genes were circadian regulators (increased Dbp, Nr1d1/2 (REV-ERBα/β), Tef, Hlf, and decreased Arntl/Bmal1). Circadian clocks coordinate intestinal epithelial renewal, barrier function, antimicrobial defense, and immune responsiveness, and circadian disruption can exacerbate experimental colitis [46,47,48]. Moreover, the gut microbiome itself exhibits diurnal oscillations that feed back onto host metabolic and immune rhythms; dietary timing and microbial metabolites are major entraining signals [49,50]. Therefore, IBM-associated circadian rewiring—observed under controlled dosing and sampling timing—raises the testable hypothesis that IBM influences the “gut clock–metabolism–barrier” axis via metabolite-sensing pathways (e.g., bile acids, SCFAs, indoles) and/or by altering microbial rhythmicity [49,50,51]. This does not prove causality, but it provides a mechanistic framework for follow-up experiments integrating chronobiology, targeted metabolomics, and barrier function assays.

4.6. Microbiome Remodeling Suggests an Additional Indirect Pathway Consistent with Mucosal Protection

IBM robustly altered cecal community structure (β-diversity separation) and shifted specific taxa, increasing Lactobacillus and several genera commonly associated with SCFA production (e.g., Roseburia, Blautia) while reducing taxa often expanded in inflammatory conditions. Importantly, our cross-omics correlation analysis (Figure 4E) provides internal consistency between the IBM-associated ecological shift and host transcriptional programs. Genera increased by IBM, particularly Roseburia and Blautia, were positively associated with circadian-linked transcripts and mucosal/transport markers (e.g., Dbp, Nr1d1/2, Tef, Hlf; Slc9a3, Guca2a, Zg16, Aqp1), whereas Streptococcus/Staphylococcus aligned with remodeling/stress-response genes (e.g., Areg/Ereg, Adamts4/9, Fn1, Col1a1/2). This pattern supports a model in which IBM-associated enrichment of SCFA-related taxa coincides with a host mucosal state favoring barrier integrity and reduced tissue remodeling. In UC, dysbiosis is frequently characterized by reductions in butyrate-producing bacteria and diminished fecal SCFAs; notably, Roseburia hominis and Faecalibacterium prausnitzii reductions correlate with disease activity [52]. SCFAs (particularly butyrate) are well-established immunometabolic mediators that expand colonic Treg pools, support epithelial metabolism, and protect against colitis in multiple models [53,54,55]. Thus, IBM-driven enrichment of SCFA-associated taxa offers a plausible indirect route by which IBM could reinforce mucosal tolerance and barrier function. Conversely, inflammatory environments can favor facultative anaerobes by increasing oxygenation and alternative electron acceptors, promoting blooms of opportunistic taxa; interventions that restore epithelial/immune homeostasis can counteract this ecological drift [56,57]. Although our study did not quantify SCFAs, bile acids, or indole derivatives directly, these well-supported mechanistic links motivate targeted metabolomics and gnotobiotic/antibiotic perturbation experiments to dissect microbiome-dependent versus direct epithelial components of IBM efficacy.

4.7. Direct Epithelial Effects Are Supported by In Vitro Junctional and Transcriptional Responses

In HCT-116 and DLD-1 cells, IBM increased ZO-1, E-cadherin, and β-catenin protein levels and upregulated tight junction–related transcripts (Ocln, Cldn1, Cldn3), supporting direct epithelial reinforcement. Tight junction/adherens junction integrity is central to mucosal homeostasis, and junctional proteins can also regulate epithelial repair competence beyond baseline permeability [2,40,41,42]. While tumor cell lines do not fully recapitulate normal epithelium, these data provide proof-of-principle that IBM contains bioactive components capable of engaging epithelial barrier programs. The observed in vitro induction of Areg contrasts with reduced Areg/Ereg in vivo after IBM alone; this may reflect differences between acute direct epithelial stimulation in vitro and the multicellular in vivo context where injury cues are reduced at baseline (IBM-only condition), or simply kinetics (early adaptive response vs. sustained injury repair signal) [44,45]. The concordant induction of barrier-related transcripts/proteins in both HCT-116 (MSI) and DLD-1 (MSS, APC-mutant) supports that IBM can directly modulate epithelial barrier programs across different colorectal epithelial genetic contexts, suggesting that the observed effects are not restricted to a single cell line background.

4.8. Fermented Soybean–Derived Bioactives Provide Additional Biological Plausibility for IBM

IBM was produced from co-culture in soybean substrate, and fermented soybean products are increasingly recognized as sources of bioactive metabolites and peptides that modulate inflammation and barrier function. Multiple studies report preventive or therapeutic effects of fermented soymilk/soy products in chemically induced colitis models, including improvements in clinical indices and colon pathology together with barrier- and inflammation-related readouts [20,58,59,60,61]. Soybean-derived bioactive compounds and fermentation-derived peptides can modulate inflammatory pathways relevant to IBD (e.g., NF-κB and cytokine networks) and may contribute to the functional activity of complex postbiotic preparations [61,62,63]. Furthermore, microbial tryptophan metabolism and indole derivatives (and related metabolite networks) can shape mucosal immunity and barrier function; recent work highlights links between indole metabolites, equol-related pathways, and DSS outcomes, underscoring how microbial metabolite ecosystems can influence colitis susceptibility [64]. These lines of evidence support the plausibility that IBM contains multiple bioactive classes (organic acids, peptides, transformed polyphenols, and other microbial metabolites) capable of orchestrating the integrated immune–epithelial–microbial effects observed here.

4.9. Limitations and Future Directions

Key limitations include: (i) IBM is chemically complex, and the active constituents were not defined; (ii) cytokines were measured in serum rather than colon tissue; (iii) immune-cell phenotyping in lamina propria/MLN was not performed; (iv) only male C57BL/6NJ mice were studied; (v) the in vitro systems used cancer cell lines rather than primary epithelium; (vi) the animal cohort size (n = 6/group) was sufficient to detect differences in DSS colitis phenotypes, but the transcriptomic and microbiome analyses should be considered exploratory and warrant validation in larger and/or independent cohorts. Addressing these limitations with targeted metabolomics, barrier permeability assays, epithelial restitution models, organoids, and microbiome-dependency experiments (antibiotics, fecal transfer, defined consortia) will be essential to pinpoint causal pathways and optimize IBM dosing schedules (including chronobiology-informed administration) [46,47,48,49,50,51,53,54,55].

5. Conclusions

In this study, prophylactic oral administration of a cell-free postbiotic preparation, intestinal bacterial metabolites (IBM) derived from a 31-strain Lactobacillus/Bifidobacterium co-culture in a soybean matrix, attenuated DSS-induced colitis in mice. IBM significantly improved colon shortening, reduced disease activity index (DAI) scores, and alleviated histopathological damage, while body weight changes were not significantly different among DSS-exposed groups. Serum multiplex cytokine profiling suggested IBM-associated modulation of inflammatory mediators, characterized by decreased IL-23 and IL-17A and increased IFN-β and GM-CSF at the endpoint. Mechanistically, rectal transcriptomics after IBM administration in healthy mice indicated coordinated changes in circadian-related genes and an upregulation of epithelial/mucosal barrier-associated programs with downregulation of injury-associated remodeling signatures. In vitro, IBM directly enhanced junction-related proteins and induced barrier-related gene expression in human colorectal cell lines. In addition, IBM altered gut microbiota community structure and shifted the relative abundance of multiple taxa, increasing several genera often considered beneficial while decreasing potentially unfavorable taxa. Collectively, these findings support IBM as a promising postbiotic strategy that may enhance mucosal resilience through coordinated epithelial, immune, and microbiota-associated pathways, warranting further studies to identify active components, clarify microbiome-dependent versus direct mechanisms, and evaluate optimal dosing schedules and therapeutic potential after disease onset.