El1405 was recovered from healthy human feces collected and was verified by 16S rRNA gene sequencing and phylogenetic and phenotypic analyses (25). The culture method for El1405 is based on previous studies (25). The preservation number of El1405 was CGMCC NO. 31231.

Specific-pathogen-free (SPF) female C57BL/6J mice, aged 5 to 6 weeks and weighing between 16 and 18 grams, were obtained from Vital River Lab Animal Technology Co., Ltd. in Beijing, China. The mice were housed in an environment maintained at a temperature of 23 ± 2 °C, with relative humidity set at 55% ± 5%, and subjected to a 12-hour light-dark cycle, all under specific pathogen-free conditions. After a one-week acclimatization period, the mice were randomly divided into three groups (8 mice per group): a control group (NC group), a phosphate-buffered saline (PBS)-treated colitis group (dextran sulfate sodium (DSS) group), and an E. limosum El1405-treated colitis group (El1405 group). Mice were orally administered 0.2 mL of PBS or 0.2 mL of El1405 (1×10⁸ CFU/0.2 mL per mouse, bacteria resuspended in PBS) for 14 days (day -7 - day -1). The NC group received an equal volume of PBS during the same period. After seven days, both the DSS and El1405 groups were administered a 3% solution of DSS (S0798, MP Biomedicals) for a consecutive duration of seven days to induce colitis (from day 0 to day 6), while the control group did not receive any DSS treatment. Specifically, the 3% DSS (freshly prepared) was incorporated into the drinking water for daily ad libitum consumption. The body weight of the mice was monitored daily throughout the study period. On day 8, the mice were sacrificed. Euthanasia was performed using a graded CO₂ inhalation system (30% chamber displacement rate), followed by neck dislocation to ensure death. All procedures adhered strictly to the 2020 AVMA guidelines for animal euthanasia. The serum samples were collected and stored at -80 °C for the analysis of cytokines and metabolomics. The colon length was measured from the ileocecal junction to the anus, recorded, and photographed. The distal colons were harvested for histopathological examination, and cecal contents were collected for microbiota analysis.

All animal studies were approved by the Ethics Review Committee of the National Institute for Communicable Disease Control and Prevention at the Chinese Center for Disease Control and Prevention (Approval code: 2023–032).

The DAI is a widely utilized measure for evaluating the severity of IBD. DAI was regularly assessed during the administration of DSS activity by scoring three distinct clinical parameters: stool consistency, weight loss, and hematochezia. The DAI scores were calculated as described in a previous study (18, 26).

The distal colon tissue specimens were initially embedded in 4% paraformaldehyde and subsequently cross-sectioned perpendicular to the long axis of the colon for further pathological studies. Stained with hematoxylin-eosin (HE) and analyzed by histopathologists. The histological score was determined based on the criteria outlined in previous studies (27, 28). The total histological score for each mouse was calculated for comparison among the three groups. The tissue slices were scanned and imaged using the Pannoromic (3DHISTECH) panoramic section scanner. To achieve imaging at a magnification of 200-fold, the intestinal tissue area was selected using CaseViewer 2.4 (3DHISTECH) scanning software. Subsequently, five crypts were randomly chosen from each tissue section. For each selected crypt, we first measured its depth using Image-Pro Plus 6.0 (Media Cybernetics) analysis software and then counted the number of goblet cells it contained. The number of goblet cells per unit length (number/mm) was calculated using the formula: number of goblet cells per unit length = number of goblet cells/length of intestinal crypt.

Immunohistochemical assays were performed on the aforementioned tissue sections. The primary antibodies were incubated with mucoprotein 2 (MUC2), while the secondary antibodies were labeled with horseradish peroxidase specific to their respective species. The Aipathwell digital pathology image analysis software was utilized to automatically assess protein positivity, which was quantified using the histochemistry score (H-Score), which was determined based on criteria established in previous studies (29, 30). The H-Score quantifies the ratio of the positive area and staining intensity in each slice, converting these metrics into corresponding values. This approach facilitates a comprehensive semi-quantitative analysis of both the depth and degree of positive tissue immunostaining. The H-score was calculated using the formula H-score = ∑ (pi × i), where i represents the staining intensity grade score and pi denotes the percentage of cells positive for the corresponding grade. The staining intensity of positive cells is categorized into four grades: negative (no coloring, 0 points), weak positive (yellowish, 1), moderately positive (brownish-yellow, 2), and strong positive (tan, 3).

Cytokine levels in the serum and colon, including TNF-α, IL-6, IL-17, IL-1β, TGF-β, LPS, IL-21, IL-22, and Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) were evaluated by ELISA kits (Dogesce Beijing, China). Serum was diluted twofold and then tested by ELISA. For the colon, 0.1 g of colon tissue was weighed, and then 1 ml of PBS was added for grinding. Subsequently, ELISA detection was performed.

RAW264.7 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, USA) supplemented with 10% FBS (Sijiqing, China) at 37 °C and 5% CO₂. The cell concentration was adjusted to 5×10⁵ cells/mL. Then, 100 μL of cells were seeded into each well of 96-well plates and cultured for 12 hours. The 0.01mg/mL IAA and ILA were added to RAW 264.7 cells, respectively, while 1 μg/mL lipopolysaccharide (LPS) was added to the cells, and co-cultured for 6 (TNF-α) or 20 (IL-1β and IL-6) hours. Cells with no treatment were used as a negative control, while cells treated with only LPS were used as a positive control. ELISA kit (Dogesce Beijing, China) was used to determine IL-1β, TNF-α, and IL-6 levels in cell supernatants.

Cecal content samples from each mouse were collected on day 16, snap-frozen in liquid nitrogen, and then stored at -80 °C. The total microbial genomic DNA of cecal contents (Tiangen, China) was extracted using a DNA isolation kit. After extraction, the 16S rRNA gene full-length primers were designed as follows: Forward primer 27F: AGRGTTTGATYNTGGCTCAG/Reverse primer 1492R: TASGGHTACCTTGTTASGACTT (31).

The PCR products were purified with AMpure PB beads, quantified using a Qubit@ 2.0 Fluorometer (Thermo Scientific), and quality-assessed with the Agilent 2100 Bioanalyzer system (Agilent, USA). The qualified PCR products were sequenced on a Sequel II sequencer (PacBio, USA) (32). After sequencing, the quality inspection was conducted on the formed sequencing library, and the obtained high-quality circular consensus sequence (CCS) was processed serially, including barcode identification. The generated optimized CCS was clustered at the 97% similarity level (USEARCH, version 10.0), and its species classification was determined based on the serial composition of the operational taxonomic unit (OTU) (33). This study used the BMKCloud platform (https://www.biocloud.net) for the bioinformatics analysis. The highly qualified OTU was used to calculate the Alpha diversity, which presented the species richness and diversity of the samples, the Shannon index and the ACE index. The beta diversity was estimated through Principal Coordinates Analysis (PCoA), which was measured by calculating the Unweighted-Unifrac distances. The screening criteria of the linear discriminant analysis effect size (LEfSe) analysis was over 4. The one-way ANOVA statistical method was used to compare the differential abundance analysis of bacterial species. The 16S rRNA gene sequencing data from this study have been deposited in the NCBI Sequence Read Archive (SRA) database (Bioproject No: PRJNA1261401).

According to the manufacturer's instructions, the colon RNA was extracted using TRIzol Reagent (Life Technologies, USA). The concentration and purity of the RNA were measured by NanoDrop 2000 (Thermo Fisher Scientific, USA). The RNA integrity was assessed by Bioanalyzer 2100 (Agilent Technologies, USA). Qualified RNA samples were utilized to construct cDNA libraries. The library construction process encompasses several steps, including mRNA enrichment, cDNA synthesis, fragmentation, end repair, adapter ligation, and PCR amplification, followed by quality control. The PCR products were subsequently purified using the AMPure XP system, and the quality of the library was assessed using the Agilent Bioanalyzer 2100 system. Following the manufacturer’s instructions, the libraries were sequenced on the Illumina NovaSeq platform, producing 150 bp paired-end reads. Sequence quality was evaluated using FastQC (v0.11.9), with low-quality bases and adapter contamination removed through fastp (v0.23.4), filtering the raw data in Fastq format to get clean data. The effective data were compared to the reference genomic sequences, with sequences that are either exactly matched or contain a single mismatch undergoing further analysis and annotation. The Hisat2 tool (v2.0.4) is employed for alignment with the reference genome (34). StringTie (v2.2.1), which utilizes the reference annotation-based transcript (RABT) assembly method, was used to identify known transcripts and predict new transcripts based on the results from the Hisat2 alignment (35). Differential expression analysis was conducted between the two groups using DESeq2 (36). Genes with a corrected p value < 0.05 and a fold change ≥ 1.5, as analyzed by DESeq2, were designated as differentially expressed (37). Subsequent analyses and data mining were performed on BMKCloud (www.biocloud.net). The heatmap of differentially expressed genes was generated using Morpheus (https://software.broadinstitute.org/morpheus). The raw data were deposited in the NCBI SRA database (Bioproject number: PRJNA1202937).

Total RNA was extracted from colon tissues using TRIzol Reagent and reverse-transcribed into cDNA with the PrimeScript™ RT Reagent Kit (TaKaRa, Japan) following the manufacturer’s protocol. The primers were synthesized by Tsingke Biotechnology (Beijing). Quantitative reverse transcription PCR (qRT-PCR) was conducted using the SYBR Green Realtime PCR Master Mix (Toyobo, Japan). Gapdh served as the internal reference, using the fold gene change = 2^−ΔΔCT method to determine the expression levels of the relevant genes. The primer sequences are provided in Supplementary Table S1.

A solution of methanol, acetonitrile, and water (2:2:1, v/v) was added to 100 μL of serum. The mixture was vortexed and subjected to low-temperature ultrasound for 30 minutes. The mixture was first incubated at -20 °C for 10 minutes, after which it was centrifuged at 14,000 g for 20 minutes at 4 °C. The supernatant was collected and processed via vacuum drying. For subsequent mass spectrometry analysis, 100 μL of an acetonitrile-water solution (1:1, v/v) was added to re-dissolve the dried sample; this was followed by vortex mixing and a second centrifugation step at 14,000 g for 15 minutes at 4 °C. Finally, the supernatant from this centrifugation was used for sample analysis.

Samples were analyzed using an Agilent 1290 Infinity LC ultra-high performance liquid chromatography (UHPLC) system (Agilent Technologies), which was equipped with both HILIC and C18 columns. For mass spectrometric analysis, an AB 6500+ QTRAP mass spectrometer (AB SCIEX) was employed. Multiple reaction monitoring (MRM) data acquisition and processing were performed using Agilent Mass Hunter Workstation Software (Version B.08.00, Agilent Technologies), with the original MRM data serving as the foundation for calculating metabolite content. All identified metabolites were classified and subjected to statistical analysis based on their chemical taxonomy. Subsequently, principal component analysis (PCA) was utilized to investigate the overall distribution patterns across different groups. Metabolites that met the criteria of fold change > 1 and p value < 0.05 were designated as differential metabolites, which were then used to generate a heatmap (18).

All experimental results were performed using GraphPad Prism 9.0 software. Data are expressed as the mean ± Standard Error of the Mean (SEM). Differences among multiple groups were assessed using one-way ANOVA followed by Dunnett’s multiple comparison test. The correlation between inflammatory factor concentrations and microbiota abundance was analyzed using Spearman’s rank correlation test. p value < 0.05 was considered statistically significant. NS means no significance. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

To evaluate the beneficial effect of E. limosum on IBD, we employed a mouse model of DSS-induced colitis. Female C57BL/6J mice were pretreated with 1×10⁸ CFU E. limosum El1405 or PBS for 7 days. Following this pretreatment, the mice were administered the same dose of El1405 or PBS, along with oral treatment with 3% DSS for another 7 days (Figure 1A). An increasing trend in body weight was observed in the NC group; however, a decrease in body weight was noted in all mice from the DSS and El1405 groups (Figure 1B). Compared to the DSS group, the El1405 group showed significantly less weight loss and reduced DAI scores on day 6 (p < 0.05, Figures 1C, D). Compared to the NC group, the DSS group had significantly reduced colon length, whereas the El1405 group had significantly less colonic shortening compared to the DSS group (p < 0.01, Figures 1E, F). HE staining and histological analysis were performed to evaluate colonic mucosa injury systematically. The DSS group exhibited a significantly higher pathological histology score compared to the NC group (p < 0.001, Figure 1H). Specifically, the DSS group exhibited significant inflammation and cellular infiltration, along with a notable loss of goblet cells and crypts, hyperplasia of connective tissue, and diffuse edema within the submucosa (Figure 1G). Conversely, the El1405-treated group presented with a lower pathological histology score, characterized by diminished inflammation and cellular infiltration compared to the DSS group (p < 0.001, Figures 1G, H). Moreover, the El1405 group had a significantly increased number of goblet cells and the length of colon crypt, in comparison to the DSS group (p < 0.05, Supplementary Figures S1A, B). Subsequently, the expression level of MUC2 in intestinal epithelial cells was assessed using MUC2 immunohistochemical staining to elucidate its localization and relative abundance within intestinal tissue. The DSS group exhibited a significantly lower H-score compared to the NC group. Conversely, the El1405-treated group presented with a higher H-score, characterized by an increased number of goblet cells compared to the DSS group (p < 0.05, Figures 1I, J). These findings indicate that El1405 effectively ameliorated DSS-induced colitis in mice.

To investigate the changes in microbiota composition, we conducted 16S rRNA gene sequencing on the cecal contents from the three groups of mice. The results indicated a significant reduction in gut microbiota diversity in the DSS group compared to the NC group, whereas the diversity in the El1405 treatment group was partially maintained (p < 0.05, Figure 2A, Supplementary Figures S2A, B). PCoA revealed that, based on the intestinal microbiota composition, the three groups of cecal contents samples were divided into three groups, and the samples from the El1405 group were more similar to those from the NC group (p < 0.05, Figure 2B). At the phylum level, Proteobacteria exhibited the highest relative abundance (42.46%) in the DSS group, Firmicutes was the predominant phylum (69.96%) in the NC group, and Bacteroidota was the predominant phylum (38.04%) in the El1405 group (Figure 2C). Compared with the NC group, the Firmicutes/Bacteroidota (F/B) ratio of the DSS group showed an increasing trend; compared with the DSS group, the F/B ratio of the El1405 group showed a decreasing trend. However, neither of these trends reached statistical significance (Supplementary Figure S2C). As shown in Figure 2D, the top three species with the highest relative abundance in the NC group were Faecalibaculum rodentium, Lactobacillus johnsonii, and Bacteroides acidifaciens. In the DSS group, the relative abundance of F. rodentium was reduced, while the E. coli group was the most dominant, accounting for 41.63%. The relative abundance of B. acidifaciens and Bacteroides thetaiotaomicron was increased in the El1405 group. LEfSe analysis showed that at the species level, E. coli and Enterococcus faecalis were significantly enriched in the DSS group, conversely, B. acidifaciens, B. thetaiotaomicron, Mucispirillum schaedleri, Phocaeicola vulgatus, and Akkermansia muciniphila were enriched in the El1405 group (Figure 2E). Furthermore, one-way ANOVA analysis confirmed that the relative abundance of E. coli and E. faecalis in the DSS group was significantly higher than that in the other two groups, in contrast, the relative abundance of B. acidifaciens and E. limosum in the El1405 group was higher than that observed in the other two groups (p < 0.05, Figure 2F). These results demonstrated that El1405 restored the altered gut microbiota composition in DSS-induced colitis mice.

RNA sequencing was conducted on colon tissues to elucidate the molecular mechanisms of El1405 in reducing colitis. Differentially expressed genes (DEGs) were screened in comparison to the respective control groups (p < 0.05, fold change > 1.5). Compared with the NC group, there were 757 up-regulated genes and 1404 down-regulated genes in the DSS group (Supplementary Figure S3A). KEGG enrichment analysis revealed that compared to the NC group, the up-regulated DEGs were significantly enriched in cell cycle, HIF, p53, IL-17, and TNF signaling pathways in the DSS group (Figure 3A). Compared with the DSS group, 174 genes were up-regulated and 290 genes were down-regulated in the El1405 group (Supplementary Figure S3B). Among the DEGs downregulated in the El1405 group, those in the TGF-β, IL-17, and p53 signaling pathways were significantly enriched (Figure 3B). In addition, gene set enrichment analysis (GSEA) showed that IL-17 and p53 signaling pathways were up-regulated (|NES| >1, p < 0.05) in the DSS group compared with the NC group (Figure 3C, Supplementary Figure S4A). In contrast, compared with the DSS group, IL-17, p53, MAPK, and TGF-β signaling pathways were down-regulated (|NES| >1, p < 0.05) in the 1405 group (Figures 3D, E, Supplementary Figure S4B). The Venn plot shows that 79 overlap genes were up-regulated in the DSS group and down-regulated in the El1405 group (Figure 3F). Next, we selected the genes related to the IL-17 pathway from these 79 DEGs and plotted them onto a heatmap (Figure 3G, Supplementary Table S2). The heatmap illustrated the four core DEGs in the IL-17 signaling pathway, including Lcn2, Il1β, Ptgs2, and Mmp13. Subsequently, qRT-PCR was used to verify these four core DEGs. It was noteworthy that the expression levels of these genes were significantly lower in the El1405 group compared to the DSS group, which aligned with the results of RNA sequencing (p < 0.05, Figure 3H). These results indicated that El1405 effectively inhibited the increase of the IL-17 signaling pathway in DSS-induced colitis in mice.

To further evaluate the effect of El1405 on the inflammatory response, we measured the levels of typical inflammatory cytokines (IL-6, IL-17, TNF-α, IL-21, IL-22, and GM-CSF) in the serum and colon of mice, which are associated with the IL-17 pathway based on RNA sequencing results. Specifically, IL-21, IL-22, and GM-CSF, which are cytokines secreted by Th17 cells and contribute to the Th17 cell-mediated inflammatory response. Compared to the NC group, the concentrations of IL-6, IL-17, TNF-α, IL-21, IL-22, and GM-CSF were significantly elevated in both the serum and colon of the mice following DSS treatment (p < 0.05). However, treatment with El1405 resulted in a significant decrease in these pro-inflammatory cytokines (p < 0.05, Figures 4A, B). Moreover, serum LPS and IL-1β levels were significantly higher in the DSS group than in the NC group (p < 0.01, Supplementary Figures S5A, B), while the El1405 group showed a significant reduction in these factors compared to the DSS group (p < 0.05). We also observed that TGF-β levels were lower in El1405-treated mice compared to DSS-treated mice, which was consistent with the RNA sequencing results indicating a decrease in TGF-β signaling pathway activity (p < 0.01, Supplementary Figure S5C). These findings suggest that El1405 inhibits the secretion of pro-inflammatory cytokines in both the serum and colon of the mice.

To explore the relationship between gut microbiota and immune factors, we analyzed the correlations between the relative abundance of significantly different species and serum/colon cytokine levels. Both E. coli and E. faecalis exhibited positive correlations with colon TNF-α, IL-6, and IL-17 levels, suggesting their potential role in promoting colitis development. In contrast, the microbiota enriched in the NC group, including F. rodentium, L. johnsonii, uncultured Bacteroidales bacterium, Lactobacillus reuteri, Muribaculum intestinale, Lactobacillus intestinalis, and Ruminococcus champanellensis, was negatively associated with TNF-α, IL-6, and IL-17 in the colon (p < 0.05, Supplementary Figure S6). These findings indicate that both E. coli and E. faecalis exhibited positive correlations with pro-inflammatory cytokines. This suggests that the El1405 intervention decreases the abundance of pathogenic bacteria associated with pro-inflammatory responses, thereby mitigating the effects of chronic colitis.

Targeted metabolomics was employed to quantify the serum metabolic profiles of mice. A total of 397 metabolites were identified and classified (Supplementary Table S3). PCA revealed significant differences in serum metabolite profiles among these three groups (Figure 5A). A total of 73 metabolites were identified as significantly up-regulated, while 85 metabolites were found to be significantly down-regulated in the DSS group compared to the NC group (fold change > 1 and p < 0.05) (Figure 5B). Additionally, compared to the DSS group, the El1405 group demonstrated a significant increase in the levels of 84 metabolites, whereas the levels of 8 metabolites exhibited a significant decrease (fold change > 1 and p < 0.05) (Figure 5C). Analyzing the classification of 92 metabolites with significant differences in concentrations between the El1405 and DSS groups, we found that six of the differential metabolites were indole derivatives (Figure 5D). The heatmap illustrated the differences in serum indole derivatives between the DSS and El1405 groups of mice. Mice treated with El1405 exhibited significantly elevated serum levels of n-acetylserotonin, 5-hydroxy-tryptophan, ILA, 5-hydroxyindole-3-acetic acid (5-HIAA), and IAA when compared to those treated with DSS (p < 0.05, Figures 5E, F). Notably, IAA and ILA levels were significantly lower in the DSS group than the NC group, but higher in the El1405 group than the DSS group (Figure 5E, Supplementary Tables S4, S5). Based on the conclusion we previously drew from in vitro experiments regarding the production of ILA and IAA by El1405 (25), these results suggest that IAA and ILA may be the main metabolites of EI1405 in alleviating IBD. Furthermore, we found that IAA markedly inhibited the secretion of IL-1β, IL-6, and TNF-α in LPS-induced RAW264.7 cells (p < 0.05, Supplementary Figure S7). However, ILA showed no inhibitory effect on the secretion of these cytokines. We also observed that in the serum of mice treated with El1405, the contents of metabolites such as vitamin B1, vitamin B2, trans-ferulic acid, vanillic acid, and ornithine were significantly increased (p < 0.05, Figures 5G, H). Previous studies have reported that these metabolites were closely related to the alleviation of IBD (38–42).