The agar diffusion method was used to evaluate the antibacterial effects of Enterococcus faecium Kimate-X and Lactobacillus plantarum Kimate-F against common pathogens associated with canine enteritis. The pathogens selected for this experiment included Enterotoxigenic Escherichia coli ATCC 43888, Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028, and Yersinia enterocolitica ATCC 9610. Each pathogen was cultured in suitable media until reaching the logarithmic growth phase and then diluted to a concentration of 1 × 10⁸ CFU/mL. One milliliter of each pathogen suspension was spread evenly onto LB agar plates, and after standing for 1 min, any excess liquid was removed. Three wells were made in each plate using an 8-mm sterile punch. The first well was filled with 50 μL of Lactobacillus plantarum Kimate-F suspension (1 × 10⁸ CFU/mL), the second well was filled with the same concentration of Enterococcus faecium Kimate-X suspension, and the third well, serving as a control, was filled with 50 μL of sterile saline. Each condition was tested in triplicate. Plates were incubated anaerobically at 37°C for 24 h, and inhibition zones were measured in millimeters using calipers to quantify the inhibitory effects of Kimate-X and Kimate-F on each pathogen.

RAW 264.7 macrophages were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics. The cells were incubated at 37°C in a 5% CO₂ incubator. When cells reached 80–90% confluence, they were digested with 0.25% trypsin-EDTA solution and resuspended at 5 × 10⁵ cells/mL. Next, 100 μL of this suspension (approximately 5 × 10⁴ cells) was seeded into each well of a 96-well plate and incubated for 24 h to allow cell adhesion. Inflammation was induced by treating the macrophages with 0.5 μg/mL lipopolysaccharide (LPS). Six hours after LPS stimulation, 1 × 10⁵ CFU/mL of Kimate-X and Kimate-F were added to the respective treatment groups. After 12 h of incubation, supernatants were collected, and nitric oxide (NO) levels were measured using a commercial ELISA kit (Meilian Bio, Shanghai). NO levels in each group were compared to evaluate the effects of the probiotic suspensions on LPS-induced NO production.

For the evaluation of LPS-induced TNF-α production, the same cell culture methods and groupings were employed. RAW 264.7 cells were treated with 500 ng/mL LPS to induce inflammation, and 1 × 10⁵ CFU/mL of Kimate-X or Kimate-F was added to the treatment groups. The control group received sterile saline without LPS or probiotics, and the inflammation model group received only LPS. After 12 h of incubation, supernatants were collected and stored at −80°C. TNF-α levels were measured using a commercial ELISA kit (Meilian Bio, Shanghai) to assess the regulatory effects of probiotics on LPS-induced TNF-α production.

A DSS-induced colitis model was used to evaluate the effects of Kimate-X, Kimate-F, and their combination on colitis in mice. Thirty 5-week-old male C57BL/6 mice were randomly assigned to five groups (n = 6 per group): Control, DSS model, Kimate-X, Kimate-F, and mixed probiotics (Kimate-X + Kimate-F). The Kimate-X group received Enterococcus faecium Kimate-X, the Kimate-F group received Lactobacillus plantarum Kimate-F, and the mixed probiotics group received a 1:1 combination of both strains. The Control group received PBS. Mice were administered 200 μL of the appropriate suspension (1 × 10⁹ CFU/mL) daily by oral gavage. After 7 days, all groups except the Control were given 3% DSS (molecular weight 36–50 kDa) in their drinking water to induce colitis. During the experiment, body weight, stool consistency, and bleeding were recorded daily to calculate the disease activity index (DAI). Samples were collected once the DSS model group exhibited clear colitis symptoms.

Blood samples were collected from the orbital venous plexus, and serum was separated by centrifugation at 3,000 rpm for 10 min after standing at room temperature for 30 min. Serum levels of LPS, IL-10, and TNF-α were measured using commercial ELISA kits (Meilian Bio, Shanghai). After euthanasia, spleens were collected to calculate the spleen index, and colon tissues were harvested for histological analysis, including colon length measurement and inflammation scoring.

The experimental design is shown in Figure 3A. Sixteen 5-month-old male Beagle dogs were randomly divided into two groups (n = 8 per group): Control and probiotic-treated (Kimate-X + Kimate-F). The Control group received a basic diet, while the probiotic-treated group received a basic diet plus a daily oral administration of mixed Kimate-X and Kimate-F suspension (1 × 10⁹ CFU/day for each strain) for the first 7 days. From days 8 to 14, the Control group was given 4% DSS (molecular weight 36–50 kDa) solution to induce colitis, while the probiotic-treated group received both probiotics and DSS. Body weight was recorded every 2 days, and the DAI was calculated. Blood and fecal samples, were collected on day 14. The mental state and fecal scoring were assessed according to the method described by Nestler et al. (42).

On day 14, 5 mL of blood was collected from the forelimb veins of each dog and centrifuged to obtain serum, which was stored at −80°C for subsequent analysis. Serum inflammatory cytokines (IL-1β, TNF-α, LPS, and IL-10) were measured using commercial ELISA kits (Meilian Bio, Shanghai). Fresh fecal samples were collected, immediately frozen at −80°C, and analyzed for SCFAs. All experimental procedures were approved by the Animal Ethics Committee of Nanjing Agricultural University (Approval Number: NJAU. NO20231227193).

SCFAs analysis was performed at Shanghai Zhongke New Life Biotechnology Co., Ltd. by gas chromatography-mass spectrometry (GC-MS). Fecal samples were thawed, resuspended in 20% phosphoric acid, and processed for GC-MS analysis, with methylvaleric acid as the internal standard. Samples were analyzed using an Agilent DB-FFAP capillary column, and SCFAs concentrations were quantified based on calibration curves.

Genomic DNA was extracted from fecal samples using the QIAamp PowerFecal Pro DNA Kit (QIAGEN, United States). DNA quality was assessed by agarose gel electrophoresis and quantified using a Qubit Fluorometer. Libraries were prepared using the NEBNext DNA Library Prep Kit, and sequencing was performed on an Illumina Novaseq 6000 platform. Raw reads were filtered, host DNA was removed, and non-host reads were assembled for gene prediction.

Taxonomic profiling was performed using MetaPhlAn4, providing species-level resolution of gut microbial composition, while functional annotations were carried out using EggNOG mapper to classify genes into metabolic pathways and biological functions. Alpha diversity indices (Shannon index, Inverse Simpson index) were calculated to assess microbial richness and evenness, while beta diversity was evaluated using Bray–Curtis dissimilarity and principal coordinate analysis (PCoA) to compare microbiota compositions across experimental groups. Group differences in microbial diversity and composition were assessed using Wilcoxon rank-sum tests for pairwise comparisons and PERMANOVA (permutational multivariate analysis of variance) for overall community structure variations. Functional pathway enrichment analysis was conducted using KEGG (Kyoto Encyclopedia of Genes and Genomes) and LEfSe (Linear Discriminant Analysis Effect Size) analysis to identify differentially enriched metabolic pathways between groups.

Spearman correlation analysis was conducted to examine associations between gut microbiota, functional gene pathways, SCFAs, and inflammatory markers. False discovery rate (FDR) correction was applied to control for multiple testing errors, with adjusted p-values <0.05 considered statistically significant. Correlation heatmaps were generated to visualize key relationships, while network analysis using Cytoscape identified highly correlated features (r > 0.5 or r < −0.5, p < 0.05), mapping microbial taxa, functional genes, and immune responses in an integrated network.

All statistical analyses were conducted using SPSS 26.0, and figures were generated using GraphPad Prism 8.0. Group comparisons were made using Student’s t-test, with statistical significance set at p < 0.05.

The antibacterial activity of Enterococcus faecium Kimate-X and Lactobacillus plantarum Kimate-F was evaluated using the agar diffusion method. As shown in (Figure 1A), both probiotic strains inhibited the growth of Escherichia coli, Salmonella enteritidis, and Yersinia enterocolitica, while no inhibition zones were observed in the control group. For E. coli, the inhibition zone diameter of Kimate-F (21 ± 1 mm) was significantly larger than that of Kimate-X (14.5 ± 0.5 mm, P < 0.001), indicating a stronger antibacterial effect of Kimate-F. Similar results were observed for S. enteritidis, with Kimate-F exhibiting a significantly greater inhibition zone (17.6 ± 0.6 mm) than Kimate-X (10.5 ± 0.5 mm, P < 0.001). Against Y. enterocolitica, Kimate-F again demonstrated superior antibacterial activity (17.5 ± 0.46 mm) compared to Kimate-X (14.3 ± 0.52 mm, P < 0.05).

In the LPS-induced RAW 264.7 macrophage inflammation model, both Kimate-X and Kimate-F significantly reduced NO and TNF-α levels. The NO levels in the Kimate-X group were significantly lower than those in the Kimate-F group (p < 0.001) (Figure 1B). In terms of TNF-α levels (Figure 1C), the Kimate-X group exhibited significantly lower levels compared to the Kimate-F group (p < 0.01).

In the DSS-induced colitis mouse model, the effects of Kimate-X, Kimate-F, and their combination on body weight, disease activity index (DAI), spleen index, colon length, histopathological changes, and inflammatory markers were evaluated. Regarding body weight changes (Figure 2A), the DSS group showed significant weight loss, whereas the control group experienced continuous weight gain. Kimate-X, Kimate-F, and the mixed probiotic group significantly alleviated weight loss, with the mixed probiotic group demonstrating the greatest weight maintenance (p < 0.001). In the DAI (Figure 2B), the DSS group exhibited significantly higher scores, while Kimate-X, Kimate-F, and the mixed probiotic group significantly reduced DAI scores, with the mixed probiotic group showing the most pronounced improvement (p < 0.05).

For the spleen index (Figure 2C), the DSS group exhibited the highest index, and Kimate-X, Kimate-F, and the mixed probiotic groups significantly reduced spleen indices, with the mixed probiotic group showing the greatest reduction (p < 0.05). In terms of colon length (Figure 2D), the DSS group exhibited significantly shortened colons, while Kimate-X, Kimate-F, and the mixed probiotic groups alleviated this shortening, with the mixed probiotic group showing the longest colon length, significantly superior to the single probiotic groups (p < 0.001).

In histopathological scoring (Figure 2E), the DSS group exhibited the most severe pathological damage, with significantly higher scores than other groups. Kimate-X, Kimate-F, and the mixed probiotic groups significantly reduced histopathological scores, with the mixed probiotic group showing markedly better tissue repair than the single probiotic groups (p < 0.01). Inflammatory marker analysis (Figures 2F–H) revealed that the DSS group exhibited significantly elevated serum LPS and TNF-α levels, while IL-10 levels were significantly reduced. In contrast, Kimate-X, Kimate-F, and the mixed probiotic groups significantly lowered LPS and TNF-α levels and increased IL-10 levels, with the mixed probiotic group demonstrating superior results compared to the single probiotic groups (p < 0.001).

In the DSS-induced colitis experiment in dogs (Figure 3A), the weight gain in the mixed probiotic group was significantly higher than in the DSS group (Figure 3B, p < 0.01), while mental status scores were significantly lower in the mixed probiotic group compared to the DSS group (Figure 3C, p < 0.05). Fecal analysis showed that the water content and fecal scores in the mixed probiotic group were significantly lower than those in the DSS group (Figures 3D,E, p < 0.05). Serum inflammatory markers revealed that LPS, TNF-α, and IL-1β levels in the mixed probiotic group were significantly lower than in the DSS group (Figures 3F–H, p < 0.05, p < 0.01, and p < 0.001, respectively). The level of the anti-inflammatory immune marker IL-10 in the mixed probiotic group was significantly higher than that in the DSS group (Figure 3I, P < 0.01).

In the metagenomic analysis of canine fecal samples, microbial diversity and community composition were compared between the DSS group and the Kimate-X + Kimate-F group. The Shannon diversity index revealed significantly higher microbial diversity in the Kimate-X + Kimate-F group compared to the DSS group (Figure 4A, p < 0.05), although the InvSimpson index did not show a significant difference (p > 0.05). Principal coordinate analysis (PCoA) (Figure 4B, left) demonstrated partial separation of microbial community structures between the two groups, indicating structural differences. However, Bray–Curtis distance analysis (Figure 4B, right) did not show statistically significant differences between the two groups (p > 0.05).

Microbial abundance analysis (Figure 4C) revealed distinct microbial compositions at the phylum, family, genus, and species levels in the Kimate-X + Kimate-F group compared to the DSS group. Specifically, the Kimate-X + Kimate-F group exhibited a higher relative abundance of Firmicutes, with significant increases in Bifidobacterium. LEfSe analysis (Figure 4D) further highlighted that Limosilactobacillus oris was significantly enriched in the Kimate-X + Kimate-F group, whereas Tyzzerella nexilis, Coprococcus comes, and Prevotella copri clade A were more abundant in the DSS group.

Comparing gut metabolic pathways between the DSS and Kimate-X + Kimate-F groups revealed significant differences. PCoA results (Figure 5A) showed distinct metabolic pathway compositions between the two groups (p = 0.019), indicating differences in metabolic pathway diversity and composition. Figure 5B displays functional pathway differences across various levels, with the Kimate-X + Kimate-F group showing higher relative abundances in pathways related to metabolism, environmental information processing, and cellular processes, particularly key metabolic pathways such as amino acid, carbohydrate, and lipid metabolism. LDA analysis (Figure 5C) revealed significant differences in specific metabolic pathways. The DSS group was enriched in pathways such as purine metabolism, the phosphotransferase system (PTS), glycolysis/gluconeogenesis, and pyruvate metabolism, while the Kimate-X + Kimate-F group exhibited higher enrichment in fatty acid metabolism, amino acid biosynthesis, secondary metabolite biosynthesis, cysteine and methionine metabolism, and longevity regulation pathways.

Figure 5D illustrates the correlations between differential gut microbiota and various metabolic pathways in DSS-induced canine colitis. Several bacterial species show significant positive or negative correlations with specific metabolic pathways. Bifidobacterium animalis demonstrates positive correlations with carbohydrate metabolism, specifically pentose and glucuronate interconversions (p < 0.05), fatty acid biosynthesis, arginine biosynthesis, and several amino acid biosynthesis pathways, including those for valine, leucine, and isoleucine, as well as phenylalanine, tyrosine, and tryptophan (p < 0.05). It is also positively correlated with the PPAR signaling pathway while being negatively correlated with the AMPK signaling pathway (p < 0.05). Blautia hansenii is positively correlated with cysteine and methionine metabolism (p < 0.05), and with the PPAR signaling pathway, but shows a negative correlation with the AMPK signaling pathway. Coprococcus comes is similarly positively correlated with the PPAR signaling pathway and negatively correlated with the AMPK signaling pathway. Dorea longicatena is primarily involved in amino acid synthesis and degradation, showing positive correlations with arginine biosynthesis, valine, leucine, and isoleucine biosynthesis, phenylalanine, tyrosine, and tryptophan biosynthesis (p < 0.05), and histidine metabolism (p < 0.01), while being negatively correlated with lysine degradation (p < 0.05). In terms of inflammatory pathways, it also exhibits a positive correlation with the PPAR signaling pathway and a negative correlation with the AMPK signaling pathway. Fusicatenibacter saccharivoran shows positive correlations with arginine biosynthesis, valine, leucine, and isoleucine biosynthesis, and phenylalanine, tyrosine, and tryptophan biosynthesis (p < 0.05), alongside other amino acid metabolic pathways like cysteine and methionine metabolism, and arginine and proline metabolism (p < 0.05). It is negatively correlated with lysine degradation (p < 0.05), positively correlated with the PPAR signaling pathway (p < 0.01), and negatively correlated with the AMPK signaling pathway (p < 0.05). Limosilactobacillus oris shows negative correlations with the biosynthesis of several amino acids, including phenylalanine, tyrosine, and tryptophan, and arginine biosynthesis (p < 0.01), as well as valine, leucine, and isoleucine biosynthesis (p < 0.05). It also negatively correlates with amino acid metabolic pathways such as cysteine and methionine metabolism, histidine metabolism, and C5-branched dibasic acid metabolism (p < 0.05). Additionally, it is negatively correlated with the PPAR signaling pathway and positively correlated with the AMPK signaling pathway. Prevotella copri clade A shows positive correlations with histidine metabolism (p < 0.05), and with the PPAR signaling pathway, while being negatively correlated with the AMPK signaling pathway. Finally, Tyzzerella nexilis is positively correlated with the PPAR signaling pathway (p < 0.05) and negatively correlated with the AMPK signaling pathway.

In the comparison of SCFAs between the two groups of dogs (Figure 6), the Kimate-X + Kimate-F group showed significantly higher levels of acetic acid (p < 0.05), butyric acid (p < 0.05), valeric acid (p < 0.01), isovaleric acid (p < 0.01), and hexanoic acid (p < 0.05) compared to the DSS group. This indicates that the Kimate-X + Kimate-F group produced higher overall amounts of SCFAs. Although no significant differences were observed in propionic and butyric acid levels between the two groups, the Kimate-X + Kimate-F group exhibited higher overall SCFAs concentrations.

In the DSS-induced colitis model in dogs, significant correlations were observed between specific microbiota and SCFAs (e.g., acetic acid, propionic acid, butyric acid, isovaleric acid, valeric acid, hexanoic acid), as well as pro-inflammatory markers (LPS, TNF-α, IL-1β) (Figure 7A). For example, Bifidobacterium animalis showed significant positive correlations with isovaleric acid and hexanoic acid (p < 0.01), and negative correlations with LPS (p < 0.05). Blautia hansenii was positively correlated with isovaleric acid, valeric acid, LPS (p < 0.05), and TNF-α (p < 0.01). Coprococcus comes was positively correlated with isovaleric acid, valeric acid (p < 0.05), and LPS (p < 0.05). Dorea longicatena showed significant positive correlations with isovaleric acid (p < 0.05) and hexanoic acid (p < 0.01). Both Fusicatenibacter saccharivorans and Limosilactobacillus oris showed significant positive correlations with hexanoic acid (p < 0.05). Prevotella copri clade A showed significant positive correlations with isovaleric acid (p < 0.01) and LPS (p < 0.05). Tyzzerella nexilis was positively correlated with valeric acid (p < 0.01) and isovaleric acid (p < 0.01).

In the DSS-induced colitis model in dogs, SCFAs are associated with various metabolic pathways (Figure 7B). Acetate is positively correlated with branched-chain amino acid synthesis, phenylalanine, tyrosine, and tryptophan biosynthesis, fatty acid metabolism, fatty acid degradation, and glycolysis/gluconeogenesis. Propionate shows positive correlations with branched-chain amino acid metabolism and fatty acid metabolism. Butyrate is positively correlated with arginine biosynthesis and phenylalanine metabolism (p < 0.05), and also with branched-chain amino acid synthesis and fatty acid synthesis pathways. Isovalerate is positively correlated with phenylalanine, tyrosine, and tryptophan biosynthesis, phenylalanine metabolism, branched-chain amino acid synthesis, and fatty acid metabolism (p < 0.05). Pentanoate is positively correlated with fatty acid synthesis (p < 0.05). Hexanoate shows significant positive correlations with fatty acid metabolism and amino acid metabolism, including arginine, proline, cysteine, and methionine metabolism, as well as branched-chain amino acid synthesis (p < 0.05).

Regarding pro-inflammatory markers, LPS is negatively correlated with fatty acid metabolism, cysteine and methionine metabolism, histidine metabolism, and arginine biosynthesis (p < 0.05). TNF-α is negatively correlated with fatty acid metabolism, glycolysis/gluconeogenesis, and several amino acid metabolic pathways (p < 0.05). Similarly, IL-1β is negatively correlated with fatty acid synthesis and multiple amino acid metabolic pathways (p < 0.05).