L. acidophilus (BNCC 186473) and P. acidilactici (BNCC 137601) were obtained as frozen vials from BNCC (Beijing, China). The strains were routinely cultured in MRS broth at 37 °C under static conditions. They were maintained by revival and subculturing according to the BNCC frozen-culture protocol and stored at −80 °C in MRS supplemented with 15% (w/v) glycerol.

For in vivo administration, both strains were cultured in MRS broth at 37 °C and harvested for 10 h prior to centrifugation (3,000 × g for 3 min at 4 °C), washed twice with sterile phosphate buffered saline (PBS), and resuspended in PBS to reconstituted to a density of approximately 5 × 10⁸ CFU/mL, as ascertained through colony counting on MRS plate, in preparation for subsequent experiments. Mouse received an oral administration of 1 × 10⁸ CFU in 0.2 mL via oral gavage daily.

All animal procedures were performed in compliance with the Guidelines for the Ethics and Use of Agricultural Animals in Research and Teaching, as approved by the experimental animal ethics committee of Yanbian University (ethical review acceptance number: YD20220718003).

C57BL/6 male mice (7–8 weeks old) were purchased from the experimental animal center of Yanbian University. The mice were housed under specific pathogen-free conditions with controlled temperature (23 ± 2 °C), humidity (55–65%), and a 12 h light/dark cycle. The mice were provided with ad libitum access to a standard chow diet and water.

During the 7-day adaptation period, all mice were allowed to acclimatize without any treatment. After the adaptation period, the experiment was divided into a 14-day feeding period and a 7-day colitis induction period (Figure 1A). Seventy mice were randomly divided into seven groups (n = 10 per group): negative control group (NC), positive control group (DSS, colitis induced by dextran sulfate sodium (DSS) treatment), L. acidophilus group (LA), P. acidilactici group (PA), L. acidophilus: P. acidilactici = 1: 1 group (T1), L. acidophilus: P. acidilactici = 2: 1 group (T2), and L. acidophilus: P. acidilactici group = 1: 2 (T3). During the feeding period (days 1–14), the treatment groups (including single and combined probiotic groups) were given an oral gavage of 1 × 10⁸ CFU of L. acidophilus, P. acidilactici, or their combination (T1, T2, and T3 ratios as defined above) in 200 μL PBS daily. Then, colitis was induced by administration of drinking water containing 2.5% DSS (MW 36,000–50,000 Da) for 7 consecutive days (days 15–21), while the probiotic administration continued in the same way as during the feeding period.

At the end of the experiment, the mice were anesthetized by inhalation of 3% isoflurane and the blood samples were collected by orbital puncture using a capillary tube and left to stand at low temperature for 2 h until stratification. The samples were then centrifuged at 3,500 × g for 10 min at 4 °C to separate the serum, which was stored at −20 °C until further analysis. Then, the mice were euthanized by CO₂ inhalation with a displacement rate of 20% chamber volume/min and the colon samples were collected and its length measured.

The disease activity index (DAI) of mice is evaluated by three aspects, which reflects the severity of colitis. Changes in body weight, fecal traits and hematochezia of the mice were recorded every day during the experiment. DAI was calculated by summing the scores for weight loss, stool consistency, and the degree of hematochezia for each mouse (Cooper et al., 1993).

Following sample collection, 1 cm of the middle colon tissue was dipped in a 4% paraformaldehyde solution and fixed for 24 h. Then, the colon tissue was excised for trimming, and the mesentery along with other attachments were removed. Dehydration, routine paraffin embedding, and H&E staining were conducted.

Intestinal morphology was evaluated by measuring mucosal thickness, crypt depth, and the ratio of mucosal thickness to crypt depth. Mucosal thickness is defined as the vertical distance from the luminal epithelial surface to the muscularis mucosae. Crypt depth is defined as the vertical distance from the crypt opening to its base.

The kits (Nanjing Institute of Bioengineering, Nanjing, China) were used to determine the amounts of malonaldehyde (MDA), myeloperoxidase (MPO), nitric oxide synthase (NOS), superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and total antioxidant capacity (T-AOC).

Colon tissue was homogenized in a centrifuge tube, and total RNA was extracted from the colon using the Trizol method. cDNA synthesis was performed using a cDNA reverse transcription kit (Shanghai Enzyme Biotechnology Co., Ltd., China). The three-step real-time PCR reaction program was as follows: 5 μL SYBR Green, 0.8 μL primers (10 μmol/L), 1 μL cDNA (1,000 ng/μL), and distilled water to make up a 10 μL system. The reaction conditions included preheating at 95 °C for 30 s, followed by the cycling phase consisting of denaturation at 95 °C for 5 s, annealing at 59 °C for 1 min, and extension at 72 °C for 30 s, repeated for 40 cycles. The melting curve was conducted at 65 °C for 5 s, followed by 95 °C for 5 s. The 2(−ΔΔCt) method, using Gapdh as the internal control, was used to quantify the mRNA levels of inflammation-related genes in colon tissue. The sequences of target gene primers are shown in Table 1.

Levels of interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α in colonic tissue homogenates were measured using ELISA kits (all from Jiangsu Enzyme-Labeled BioTECH, Yancheng, China) according to the manufacturer’s instructions.

Genomic DNA was extracted from samples using the cetyltrimethylammonium bromide (CTAB) method. The V4 region of the bacterial 16S rRNA gene was amplified using specific primers 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and 806R (5′-GGACTAC HVGGGTWTCTAAT-3′). The PCR amplification was performed under the following conditions: initial denaturation at 98 °C for 1 min; followed by 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and extension at 72 °C for 30 s, with a final extension at 72 °C for 5 min. The PCR products were purified using a Qiagen gel extraction kit (QIAGEN, Germany) and quantified by fluorescence-based techniques. The purified DNA was used to construct sequencing libraries with the TruSeq® DNA PCR-free sample preparation kit (Illumina, United States). The library quality was assessed using a Qubit® 2.0 Fluorometer (Thermo Scientific) and an Agilent Bioanalyzer 2100 system. Paired-end sequencing (2 × 250 bp) was performed using the Illumina NovaSeq platform (Novogene, China). Raw sequencing data were processed with the QIIME software package (v1.9.1).

The concentration of SCFAs was measured by the external standard method using a gas chromatograph (GC-7890A, Agilent Technologies Inc., Santa Clara, CA, United States). The injection volume was set at 1 μL, with injector and detector temperatures set at 170 °C and 200 °C, respectively. The column temperature was maintained at 120 °C and nitrogen was used as the carrier gas at a flow rate of 26.43 mL/min. For sample preparation, the fermentation supernatant was centrifuged at 13,400 × g for 10 min. Then, 0.8 mL of the supernatant was transferred into a 1.5 mL centrifuge tube containing 0.2 mL of 25% metaphosphoric acid. Then, the mixture was allowed to stand for 30 min, followed by centrifugation at 10,000 × g for 15 min. The resulting supernatant was filtered through a 0.22 μm membrane before GC analysis. Quantification was performed using the external standard method with five-point calibration curves constructed for each SCFA. The linearity of the calibration curves was robust, with all correlation coefficients at R² > 0.99.

The data obtained in the above experiments were statistically analyzed by SPSS 17.0 software. One-way ANOVA for analysis of variance, and LSD and Duncan methods were used for pairwise comparison, p < 0.05 was considered as a significant level, and data were expressed as mean ± standard deviation. Bar plots were generated by using GraphPad Prism 10.2.3. To integrate related host biomarkers into a comprehensive analysis, principal component analysis (PCA) was performed. Prior to PCA, all variables were Z-score standardized to eliminate scaling differences. PCA was conducted on the protein concentrations of TNF-α, IL-1β, and IL-6 from colonic tissue homogenates and separately on the data from six antioxidant markers (SOD, GSH-Px, T-AOC, MPO, MDA, and NOS). The first principal component (PC1), was extracted to serve as a variable for subsequent correlation analysis. To explore the intrinsic associations among the microbiota, metabolites, and host phenotypes, a Spearman’s rank correlation analysis was performed based on all individual sample data. The variables analyzed included the relative abundance of Akkermansia, cecal butyrate concentration, the PC1 scores derived from cytokines and antioxidant markers, and the relative mRNA expression of ZO-1 and occludin. The correlation matrix (Spearman r) and corresponding p-values are presented as a heatmap and scatter plots of key significant correlations in the supplementary materials. Key significant correlations were highlighted and plotted as scatter plots. Significance levels are denoted as follows: *p < 0.05, **p < 0.01, and ***p < 0.001.

Weight loss is a well-established and objective parameter of DSS-induced colitis in mice (Britto et al., 2019). To evaluate whether the combined probiotics ameliorate colitis symptoms, we monitored body weight (Figure 1B) and daily weight gain (Figure 1C). From days 15 to 21 of the experiment, the daily weight gain in all DSS-treated groups decreased compared to the NC group due to the progression of colitis. At day 21, the body weight of the T2 group was significantly higher than that of the DSS group (p < 0.05); all other groups significantly alleviated the decrease in daily weight gain (p < 0.05). Overall, the T2 group showed less weight loss compared to the other groups.

To further assess inflammation severity and tissue damage, we evaluated DAI scores, colon length, and histological changes. The DAI scores (Table 2) showed the severity of colitis, including the degree of weight loss, the presence of blood in the stool, and stool consistency. The DSS group mice showed a significant increase in the index of DAI score on the second day of the experiment (Figure 1D). Compared to the DSS group, the DAI indexes of the mice in each treatment group showed a reduction. Severe inflammation may result in symptoms such as bloody stool, leading to completely unformed colon contents. The colon contents, while not distinctly granular, remained well-formed, and the presence of bloody stool showed significant improvement in the T2 group (Figure 1E).

As expected, there was a significant (p < 0.05) reduction in colon length in the DSS group. All treatment groups significantly alleviated DSS-induced colon shortening in mice compared to the DSS group (p < 0.05). The colon length in the T2 group was significantly greater than in the DSS group (p < 0.05) and showed the most substantial improvements compared to other treatment groups, although it was not statistically different from the other combination groups (Figures 1E,F). The T2 group demonstrated superior efficacy compared to other treatment groups in alleviating DSS-induced intestinal injury.

Next, we examined the effect of L. acidophilus combined with P. acidilactici on colonic pathology in mice (Figure 2). As expected, Figure 2A showed that the NC group had well-developed mucosa, normal crypts, intact gland structures, and abundant goblet cells with no evidence of immune cell infiltration. The DSS group exhibited significant mucosal epithelium damage, characterized by extensive gland destruction, loss of crypt structures, submucosal edema, and substantial immune cell infiltration. It was noting that the inflammatory condition improved across all treatment groups receiving lactic acid bacteria. The T2 group showed the most significant improvement in the alleviation of pathological structural damage to the colon. The vast majority of epithelial structures retained their integrity, with a reduction in the destruction of crypts and glandular structures, along with a significant decrease in the cell infiltration after 7 days of DSS induction.

Figures 2B–D illustrates the effects of L. acidophilus combined with P. acidilactici on the thickness of colon mucosa and the depth of the crypts, and their respective ratio. The DSS group had a significantly greater crypt depth than the NC group (p < 0.05). After intervention with lactic acid bacteria, all treatment groups showed effective relief, with no significant difference observed compared to the NC group (p > 0.05). The mucosal thickness and the ratio of mucosal thickness to crypt depth in the DSS group were significantly lower than those in the NC group (p < 0.05). Following intervention with lactic acid bacteria, all treatment groups showed improvement, with the T1 and T2 groups performing better improvement in mucosal thickness compared to other treatment groups.

Oxidative stress is a major contributor to intestinal tissue injury and has been widely implicated in the pathogenesis of inflammatory bowel disease (Jin et al., 2025). Given that local intestinal inflammation often leads to systemic redox imbalance, we investigated the systemic antioxidative effects of the combined probiotics. Key oxidative markers in the serum were quantified to evaluate the host’s overall redox status. The effects of L. acidophilus combined with P. acidilactici on serum oxidative stress in mice with DSS-induced colitis are shown in Figure 3. The induction of DSS significantly decreased the SOD, T-AOC, and GSH-PX levels in the mice (p < 0.05). The intervention with combined lactic acid bacteria increased all indicators, with the levels of serum SOD, T-AOC, and GSH-PX in groups T2 and T3 significantly higher than those in the DSS group (p < 0.05). Following treatment with combined lactic acid bacteria, significant improvements were observed, particularly in group T2, which exhibited the most favorable outcomes, as all indicators showed significant reduction compared to the DSS group.

Next, we examined the impact of the combined probiotics on the inflammatory response associated with colitis by measuring pro-inflammatory cytokine levels in colonic tissues. Elevated pro-inflammatory cytokines are indicative of inflammation and are correlated with the severity of colitis (Zhang et al., 2022). Figures 4A–C illustrates the impact of L. acidophilus and P. acidilactici on inflammatory cytokines in mice with DSS-induced colitis. In comparison to the NC group, the levels of inflammatory cytokines in the DSS group significantly elevated (p < 0.05). Following intervention with lactic acid bacteria, all treatment groups exhibited a significant (p < 0.05) reduction in IL-1β, IL-6, and TNF-α at the protein level, with the combined groups demonstrating the most pronounced effect.

Tight junction proteins are essential for maintaining intestinal epithelial barrier integrity, disruption of these proteins is considered a critical feature of colitis (Chelakkot et al., 2018b). We examined the expression of tight junction proteins to evaluate the effect of the treatment on mucosal barrier integrity. Figures 5A,B illustrates the impact of L. acidophilus and P. acidilactici on the mRNA expression of colonic tight junction proteins in mice with DSS-induced colitis. In comparison to the NC group, the DSS group exhibited a significant reduction in the relative mRNA expression levels of ZO-1 and occludin (p < 0.05). Following intervention with lactic acid bacteria, all treatment groups exhibited significant increases in gene expression (p < 0.05), with the T2 group demonstrating the most pronounced effect.

Next, we further explored the underlying regulatory mechanisms by analyzing changes in the gut microbiota and metabolites. Gut microbiota dysbiosis and changes of SCFA production play a critical role in the development of colitis by disrupting barrier integrity and promoting inflammation (Haneishi et al., 2023). Figure 5C illustrates that the rarefaction curves for all samples reached a plateau, indicating that the current sequencing depth was sufficient to encompass microbial diversity, thus confirming the reliability of the sequencing data. Analyses of alpha and beta diversity are presented in Figures 5D,E. Compared to the NC group, the DSS group showed an increasing trend in the Chao1 index. Following intervention with lactic acid bacteria, the Chao1 indexes of the T1, T2, and T3 groups exhibited a substantial elevation (p < 0.05) compared to the DSS group, with T2 showing the most pronounced increase. The Chao1 indices for the LA and PA groups, however, exhibited a decline (p > 0.05). The NMDS results demonstrated that DSS caused substantial changes in gut microbial composition. The microbial communities in colitis-induced mice treated with combined lactic acid bacteria were distinctly clustered in contrast to mice with colitis.

The impact of L. acidophilus combined with P. acidilactici on relative abundance is shown in Figures 5F,G, where the relative abundance of gut microbiota at both the phylum and genus levels was analyzed. The predominant phyla identified were Firmicutes, Bacteroidota, and Verrucomicrobiota. Compared to the NC group, the DSS group showed a significant (p < 0.05) reduction in the relative abundance of Firmicutes and an increase in Bacteroidota. Following the intervention with lactic acid bacteria, these trends were reversed, resulting in an increase in Firmicutes and an a decrease in Bacteroidota. In addition, colitis induction significantly diminished Verrucomicrobiota (p < 0.05); however, its relative abundance was increased in all treatment groups upon intervention with lactic acid bacteria. The T1, T2, and T3 groups exhibited significantly elevated levels of Verrucomicrobiota when compared to the DSS group (p < 0.05). After intervention with lactic acid bacteria, the prevalence of Akkermansia increased in all treatment groups, with the T1, T2, and T3 groups showing significantly higher levels than the DSS group (p < 0.05).

Figure 5H shows that LEfSe analysis identified significant differences in biomarkers at the genus level among several treatment groups (LDA score > 3.5, p < 0.05). The NC group exhibited a high prevalence of Turicibacter, Ligilactobacillus, and Enterorhabdus. The DSS group had an abundance of Dubosiella and Romboutsia. The LA group showed enrichment of Faecalibaculum, while the PA group was enriched with Candidatus_Saccharimonas and Limosilactobacillus. Among the combined lactic acid bacteria intervention groups, the T1 group was significantly enriched in Akkermansia and Bacteroides; the T2 group in Colidextribacter and Lachnospira; and the T3 group in Faecalibacterium, Alistipes, and Blautia.

As shown in Figures 5I–K, the cecal concentrations of acetic acid, propionic acid and butyric acid were significantly decreased in the DSS group compared to the NC group (p < 0.05). Following the intervention in mice with lactic acid bacteria, all treatment groups showed significantly increased butyric acid levels (p < 0.05), with the T2 group demonstrating the most substantial improvement.

Given that the combined probiotic treatments enriched the relative abundance of Akkermansia and elevated cecal butyrate levels, we sought to investigate whether these specific microbiota and metabolite changes were statistically linked to the improvements in host phenotypes. A Spearman’s rank correlation analysis was performed on all individual animal samples in Figure 6. Principal component analysis (PCA) was performed to integrate host phenotypes. The first principal component (PC1) extracted from cytokines explained 86.943% of the total variance and exhibited strong positive loadings for all three pro-inflammatory cytokines, indicating that higher concentrations of these cytokines contribute to a higher PC1 score. Therefore, the PC1 score was defined as the “Inflammation Index,” where a higher score represents a stronger overall inflammatory status. Similarly, for antioxidant markers, the PC1 extracted explained 84.729% of the total variance. This PC1 score correlated positively with antioxidant enzymes (SOD, GSH-Px, T-AOC) and negatively with oxidative products (MPO, MDA, NOS). Thus, the PC1 score was defined as the “Antioxidant Index,” where a higher score represents a stronger overall antioxidant capacity. The overall correlation heatmap revealed a distinct pattern of interactions (Figure 6A). As shown in Figures 6B–F illustrates that, at the microbial level, the relative abundance of Akkermansia is significantly positively correlated with cecal butyrate concentration (p < 0.05). This finding suggests a potential synergistic relationship in which the colonization of Akkermansia may favor a butyrogenic environment. The abundance of Akkermansia was also significantly negatively correlated with the inflammation index (p < 0.05), indicating that the restoration of Akkermansia is closely tied to the alleviation of inflammatory signaling. It also reached statistical significance for direct correlation with both tight junction genes, ZO-1 and occludin (p < 0.01), implying a potential role of Akkermansia in reinforcing the mucosal physical barrier.

Parallel to the microbiota findings, cecal butyrate concentration has been identified as a key metabolite regulator, showing robust correlations with host phenotypes (Figures 6G–J). As illustrated in the scatter plots, there was a significant positive correlation between butyrate and the antioxidant index (p < 0.001) and an apparent negative correlation with the inflammation index (p < 0.001). Furthermore, increased butyrate levels were positively correlated with the relative mRNA expression of both tight junction protein genes, ZO-1 (p < 0.001) and occludin (p < 0.001). The strong associations indicate that butyrate serves as a central metabolite hub for host-protective effects, linking probiotic-induced modulation of the microbiota to the observed enhancement in gut integrity, antioxidant capacity, and immune homeostasis.