B. lactis attenuate DSS-induced body weight loss in mice

The changes in body weights for test animals during the experimental course are summarized in Fig. 2. In general, the body weight of mice decreased substantially at each periodic exposure to DSS, and the mice tentatively regained their body weight once the DSS treatment was suspended. From the second exposure to DSS onward, the body weight loss of the mice in the AOM/DSS, B. lactis P, and B. lactis S groups was diminished significantly, suggesting an adaptive and cooperative recovery from inflammation under the DSS stimulation. After the third exposure to DSS, the body weight of AOM/DSS mice decreased significantly in stark contrast to those of the mice groups receiving either B. lactis P or B. lactis S, which showed body weight loss to a lesser extent (p < 0.05), suggesting B. lactis P or B. lactis S exerts its effects on restoring the body weight from the DSS-induced weight loss (Fig. 2).

Fig. 2

Weight changes of mice among four groups. C57BL/6 J mice were given a single i.p. injection of AOM (10 mg/kg) on the first day and provided drinking water ad libitum containing 2% (w/v) DSS for 5 days, followed by 2 weeks of sterile water. Mice were subjected to three cycles of 2% DSS (5 days per cycle) treatment and regular water (14 days per cycle). The total experimental period was 79 days. B. lactis P (1 × 108 CFUs/day) or B. lactis S (80 μg/day) were orally inoculated 7 days before AOM treatment and continued supplement for 79 days. Data are expressed as the means ± SD. *p < 0.05 significantly different from the control group

Prevention of AOM/DSS-induced small and large intestine damage, aberrant crypt foci, and formation of colonic polyps by B. lactis

AOM/DSS-induced intestinal damage was evaluated by measuring the length and weight of small and large intestines, ACF, and colonic polyps. The length of the small intestine in AOM/DSS mice is shorter than those in the control, B. lactis P, and B. lactis S groups (p < 0.05, Fig. 3A, B). The colon lengths were 7.17 ± 0.60 cm, 5.63 ± 0.56 cm, 6.80 ± 1.02 cm, and 5.53 ± 0.69 cm for the control group, the AOM/DSS group, the B. lactis P group, and the B. lactis S group, respectively. On average, the colon length of the control mice was 27.35% longer than those of mice treated with AOM/DSS. Mice subjected to the AOM/DSS treatment exhibited a reduction in colon length, whereas those receiving the B. lactis P treatment improved their colon length by 20.72% when compared with that of the AOM/DSS group (p < 0.05, Fig. 3D, E) but comparable to that of the B. lactis S administration (p > 0.05, Fig. 3D, E). Our results suggested that the treatment with B. lactis P could reverse the AOM/DSS-induced colon shortening.

Fig. 3

B. lactis P and B. lactis S attenuated the AOM/DSS-induced colon shortening. Assessment of the small intestine and the colon length in AOM/DSS murine model on day 79. (A) Macroscopic images of the small intestine are shown. (B) The small intestine length of each mouse was measured. (C) The small intestine weight of each mouse was measured. (D) Macroscopic images of the colon are shown. (E) The colon length of each mouse was measured. (F) The colon weight of each mouse was measured. Data are expressed as the means ± SD. *p < 0.05 significantly different from the compared group

The colonic precancerous lesions in the mice under the treatment of AOM/DSS were assessed for their colon mucosa proliferation and polyp occurrence at macroscopic and microscopic levels (Fig. 4). Colonic polyps (with a characteristic opaque spot) emerged from colons in all AOM/DSS mice (100%, Fig. 4A); macroscopic observation revealed that mucosal protuberance was increased in the descending colon region (Fig. 4B). Interestingly, with addition of B. lactis P or B. lactis S, both number and size of polyps were reduced in colon (Fig. 4A, B).

Fig. 4

B. lactis P and B. lactis S prevented AOM/DSS-induced aberrant crypt foci and the formation of colonic polyps. A Representative gross macroscopic image of the colon. Red arrows indicate polyps. B Representative pictures of opened specimens by cutting along the bowel. Yellow arrows indicate polyps. C The number of polyps per mouse in different parts of small intestines. Data are expressed as the means ± SD. *p < 0.05 significantly different from the compared group. D Correlation between colon weight and polys number in colons. E Representative H&E sections of colons

We then examined what effects could result with respect to the weight changes of the small/large intestines and the formation of colonic polyps when AOM and DSS were applied. The weight of the small intestine in all groups is similar (p > 0.05, Fig. 3C), suggesting that AOM/DSS did not cause a significant weight change in the small intestine. The colon weights for the control, AOM/DSS, B. lactis P, and B. lactis S groups were measured as 0.18 ± 0.04, 0.22 ± 0.05, 0.18 ± 0.046, and 0.18 ± 0.02 g, respectively, where the one in the AOM/DSS group is significantly higher than others (p < 0.05, Fig. 3F). The increased colon mass is proportional to the increased polyp mass, where the mean number of colonic polyps treated by AOM/DSS, AOM/DSS + B. lactis P, and AOM/DSS + B. lactis S is 12.2 (range 8–17), 4.5 (range 2–6), or 4.2 (range 2–7), respectively (Fig. 4A–C, p < 0.05). It is worth noting that the colon sections of AOM/DSS mice showed a noticeable increased number and size of polyps in agreement with the macroscopic observation. The number of colonic polyps reflects what type of treatment; likewise, the colon weight is positively proportional to the number of polyps in the colon (Fig. 4D, r = 0.746, p < 0.001). As shown in Fig. 4E, the H&E staining is consistent with the descriptions mentioned above; namely, the polyp load is decreased in the mice treated with B. lactis P and B. lactis S.

Mice treated with AOM/DSS commonly result in colonic precancerous lesions. Histopathological analysis was carried out for the hematoxylin-and-eosin-stained sections of the mice colon. Microscopically, colon samples from the control group show normal histology. In the AOM/DSS group, the colonic sections demonstrate a number of pathological changes: crypts changes in severe lesions throughout mucosa, alteration in the epithelial structure (larger crypts, a thicker and darker-staining epithelial lining, and a larger pericryptal zone), and high levels of inflammatory cell infiltration into the mucosal and submucosal areas. In the B. lactis P and B. lactis S mice, epithelial lesions and infiltration of inflammatory cells are low as seen in the left panel of Fig. 4E, and the mucosal architecture remains intact when compared to those of the AOM/DSS group, highlighting that B. lactis P and B. lactis S can prevent or slow down the development of colonic precancerous ACF lesions.

B. lactis modulate hematological, spleen immunological parameters, and inflammatory responses in AOM/DSS-treated mice

In the AOM/DSS group, the total RBC count, hemoglobin, and MCHC are significantly low when compared with those of other groups. In contrast, the total RBC counts, hemoglobin, and MCHC in the mice receiving either the B. lactis P or B. lactis S treatment remain similar to that of the control (p < 0.05, Fig. 5A), suggesting that B. lactis can maintain the level of AOM/DSS-induced RBC, hemoglobin, and MCHC at a steady level.

Fig. 5

Effects of B. lactis P and B. lactis S on the hematological and spleen immunological parameters in AOM/DSS-treated mice. A Red blood cell indices. B White blood cell indices. C Spleenocyte parameter. Data are expressed as the means ± SD. *p < 0.05 significantly different from the compared group

After the AOM/DSS treatment, the cell counts of absolute neutrophils, eosinophils, and monocytes were determined to be 244.4 ± 144.7, 37.5 ± 19.0, and 181.5 ± 111.0 × 109/L, respectively, significantly higher than the counterparts in control. After the administration of B. lactis P or B. lactis S, the leukocytosis effect was significantly palliated, especially for eosinophilia (p < 0.05; Fig. 5B). Figure 6E–H shows the alterations in subpopulation distributions of splenocytes of mice subject to various treatments. The percentage of CD8+ T cells and B cells in splenocytes is meaningfully increased in B. lactis P–treated mice (Fig. 5C), suggesting the B. lactis treatment promotes acquired immune responses.

Fig. 6

Effects of B. lactis P and B. lactis S on expression of inflammatory cytokine and NF-κB signaling in AOM/DSS-treated mice. A The protein levels of TNF-α, IL-1β, IL-6, IFN-γ, and IL-10 in the plasma of mice were detected by ELISA. B Semiquantitative analysis of colonic tissue protein levels of IκB-α, IKK-β, NF-κB, and GAPDH levels. GAPDH served as an internal control for equal loading. C The intensities of western blot bands were determined by the ImageJ. The intensity (mean ± SD) was normalized to the control group that was set to 1. Protein levels differed significantly among the groups, * indicates p < 0.05 vs. the compared group

To explore the inflammatory modulation of B. lactis, we further examined the effects of B. lactis P or B. lactis S on inflammatory cytokines and NF-κB pathway–related protein expression in the colon tissue of the AOM/DSS-treated mice. The AOM/DSS treatment gave rise to a noticeable elevated level of tumor necrosis factor-α (TNF-α), interleukin (IL)−1β, IL-6, and interferon-γ (IFN-γ) in the colon tissue (Fig. 6A). In contrast, inflammatory cytokines were significantly reduced in the groups with the treatment of B. lactis P or B. lactis S when compared with those in the AOM/DSS group (p < 0.05). In contrast to the AOM/DSS group, where IκBα was decreased and NF-κB was increased, the NF-κB-related inflammation proteins were down-regulated in the B. lactis P/B. lactis S group, counteracting the effect of AOM/DSS (Fig. 6B, C).

B. lactis mediate the structural and functional compositions of gut microbiota

Since several studies have revealed that there is a close relation between the composition of gut microbiota and inflammation or colonic precancerous lesions (Ryan et al. 2020; Polimeno et al. 2020), we then came to investigate whether gut microbiota of mice were changed upon the B. lactis P or B. lactis S supplementation during colonic precancerous lesions. The taxonomic profiles at the phylum level revealed that the gut bacteria community in the control mice was dominated by Bacteroidetes (64.05 ± 1.86%), Firmicutes (31.60 ± 1.12%), and Deferribacterota (3.43 ± 1.62%) (Fig. 7A). In the AOM/DSS-treated mice, we found a significant reduction of Bacteroidetes (48.93 ± 6.34%) and an increase of Firmicutes (42.19 ± 8.16) in the gut microbiota. Clearly, treatment with B. lactis P or B. lactis S promotes a decrease in Firmicutes and Deferribacterota while increasing Bacteroidetes compared to the AOM/DSS group, ultimately restoring the microbial balance to that of a healthy control group (Fig. 7A). Given that both Firmicutes and Bacteroidetes are deemed as influential regulators in human gut microbiota (Flemer et al. 2017), the ratio of Firmicutes to Bacteroidetes (F:B ratio) acts as an index reflecting intestinal homeostasis. With the AOM/DSS treatment, it significantly increased the population abundance of Firmicutes, but the population abundance of Bacteroidetes decreased, thus leveraging the F:B ratio toward Firmicutes when compared with that in the control mice (p < 0.05; Fig. 7B). Treatment with B. lactis P or B. lactis S significantly reduced the F:B ratio compared to the AOM/DSS group, restoring it to the healthy control group level, with B. lactis S exhibiting a more pronounced effect. This further suggests that B. lactis S supplementation promotes the growth of Bacteroidetes in the mouse gut (p < 0.05; Fig. 7B). The beta diversity (PCA), the covariance matrix of relative abundance profiles at the genus level, suggested that there is a clear distinction between the control and AOM/DSS groups or the B. lactis S and AOM/DSS groups (Fig. 7C). Alpha diversity analysis indicated that the values of the observed species richness estimator Chao1 suggest there are significant differences between the AOM/DSS and B. lactis P groups, as well as the AOM/DSS and B. lactis S groups (Fig. 7D). However, the values of the Simpson and Shannon diversity indices show no significant differences among four groups (Fig. 7D).

Fig. 7

Effects of B. lactis P and B. lactis S on the structural and functional composition of gut microbiota. A Changes in the relative phylum level abundances of gut microbiota components in AOM/DSS-treated mice. B The ratio of Firmicutes to Bacteroidetes (F/B) boxplot showing gut microbiota in AOM/DSS-treated mice. Boxes contain 50% of all values and whiskers represent the 25th and 75th percentiles. *p < 0.05 significantly different from the compared group. C PCA plot based on the covariance matrix of genus-level relative abundance of gut microbiota components in AOM/DSS-treated mice. D Box plots showing species richness estimator (Chao1 and observed features) and species evenness estimator (Shannon and Simpson) of alpha diversity. *p < 0.05 significantly different from the compared group. E, F Heatmap depicting the relative abundance of the most abundant family (E) and genera (F) (> 0.1%) of gut microbiota from different treatments. The color intensity in each sample is normalized to represent its relative ratio in AOM/DSS-treated mice. A range of colors, from green to red, indicates the relative values of microbiota (0–1). G The relative abundance of Muribaculaceae, Prevotellaceae_UCG-001, Anaerostipes, Ruminococcaceae, Mucispirillum, Clostridia_UCG-014, and Clostridia_vadinBB60. Data are expressed as the means ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001 significantly different from the compared group

A heat map from hierarchical clustering analysis based on the top 23 different family taxa shows that there are intersample changes amid the four groups. The relatively highly abundant species include Butyricicoccaceae, Oscillospiraceae, RF39, Clostrida_ vadinBB60_group, Acholeplasmataceae, Deferribacteraceae, Clostridia_UCG-014, and Akkermansiaceae likely as a result of the AOM/DSS treatment, while they were decreased after oral administration of B. lactis P or B. lactis S (Fig. 7E). In contrast, the relatively less abundant species include Prevotellaceae, Ruminococcaceae, Muribaculaceae, Rikenellaceae, and Tannerellaceae likewise as a result of the AOM/DSS treatment, which was reversed after oral administration of B. lactis P or B. lactis S.

At the genus level, we used a heat map, which was derived from hierarchical clustering analysis on the basis of the top 31 different genus taxa, to summarize intersample changes among the four groups. We found that there is a clear separation in the gut microbiota between the AOM/DSS mice and other groups (Fig. 7F). We then selected 7 differently expressed genera and quantified their expression changes. The application of AOM/DSS treatment significantly reduced the relative abundance of well-characterized beneficial bacteria, including Muribaculaceae, Prevotellaceae_UCG-001, Anaerostipes, and Ruminococcaceae, in the gut of mice. However, their relative abundance was significantly increased in mice gavaged with B. lactis P or B. lactis S (p < 0.05; Fig. 7G). In contrast, the abundance of pathogenic bacteria, such as Mucispirillum, Clostridia_UCG-014, and Clostridia_vadinBB60, was significantly increased in the AOM/DSS group. However, this trend was reversed when B. lactis P or B. lactis S was supplemented (p < 0.05; Fig. 7G).

Discussions

In this study, we affirmed that the prophylactic administration of B. lactis P or B. lactis S is capable of alleviating the AOM/DSS-induced weight loss, lowering the expression of inflammatory cytokines, and diminishing ACF and colonic polyps in an AOM/DSS mouse model. These results are in good agreement with previous probiotics reports (Genaro et al. 2019; Ni et al. 2017 Rifkin et al. 2020). Interestingly, we identified here that the effect of viable probiotic bacteria and non-viable postbiotic metabolites is comparable, in which the latter achieves a similar inhibitory effect as the former on the AOM/DSS-induced weight loss, inflammation, and precancerous lesions. We reasoned that this outcome is because of the formation of postbiotic metabolites from a long fermentation process of B. lactis, whereby some risk factors are drained out and beneficial factors are converged. Postbiotics are metabolites of microorganisms including short-chain fatty acids (SCFAs), bacterial polysaccharides, and vitamins (B and K), which may act as antioxidant, anti-inflammatory agents, and/or anticancer agents in addition to facilitating the growth of probiotics (Tsilingiri and Rescigno 2013). Although numerous beneficial phenomena for both prebiotics and postbiotics have been learned, new research is still needed as to how and what given metabolites from the cell-free supernatant of B. lactis affect the prevention of inflammation and precancerous lesions.

ACF, colon mucosa proliferation, and polyp occurrence are thought to be useful biomarkers/indices for neoplastic lesions in the colon carcinogenetic model (Wargovich et al. 2010). The present study established that high ACF, severe lesions in the mucosa, alterations in epithelial structure, and high-level inflammatory cell infiltration into mucosal and submucosal areas can be induced in the AOM/DSS-treated mice. In contrast, the epithelial lesions, the infiltration of inflammatory cells, and the integrity of mucosal architecture can be significantly improved in the B. lactis P and B. lactis S mice, suggesting that B. lactis P and B. lactis S are in a position to compensate the AOM/DSS-induced colonic precancerous lesions. It is known that colonic epithelial inflammation can result in persistent immune dysregulation and neoplastic mucosa; the development of inflammation-associated bowel cancer then follows. A previous study has reported that synbiotic intervention with L. acidophilus was able to inhibit colon carcinogenesis by regulating inflammation along with decreasing the number of precancerous lesions on 1,2-dimethylhydrazine (DMH)/DSS-induced colonic precancerous lesions and tumors in mice (Deol et al. 2017; Lee et al. 2019; Li et al. 2019). Moreover, the administration of Bifidobacterium was reported able to alleviate intestinal inflammation and carcinogenesis in free fatty acid receptor 2 (Ffar2)-knockout mice (Sivaprakasam et al. 2016), suggesting that bifidobacteria counters oncogenesis through modulating inflammation, restoring compromised mucus layers, and suppressing unfavorable microbiota.

Several lines of evidence have shown that gut microbiota plays an important role in maintaining intestinal homeostasis, as it is closely associated with the development of bowel cancer and bowel precancerous conditions (Polimeno et al. 2020; Ryan et al. 2020). In the present study, we observed that there is a significant population shifting of gut microbiota between the B. lactis P/B. lactis S mice and AOM/DSS mice, in close correlation to the ratio of Firmicutes versus Bacteroidetes (Liu et al. 2022). At the phylum level, we found that the F:B ratio was elevated in the AOM/DSS group, suggesting that AOM/DSS treatment may influence gut microbiota composition during bowel cancer formation by potentially inhibiting the growth of Bacteroidetes while favoring the growth of Firmicutes. Both B. lactis P and B. lactis S supplementation, however, can reverse the microbiota imbalance caused by AOM/DSS. At the genus level, the abundance of Muribaculaceae, Prevotellaceae_UCG-001, Anaerostipes, and Ruminococcaceae was significantly decreased in the AOM/DSS group. Muribaculaceae was shown capable of inhibiting CD8 + T-cell activation to modulate immunity stimulation, therefore being regarded as an anti-inflammatory bacterium (Shang et al. 2021). Anaerostipes, Prevotellaceae, and Ruminococcaceae are considered butyrate-producing bacteria in a way to enhance colonic defense, while the depletion of Anaerostipes, Prevotellaceae, and Ruminococcaceae often leads to the intestinal barrier and gut microbiota dysbiosis (Chen et al. 2022; Dunn et al. 2022; Xie et al. 2022). Lower butyrate production in mucosa and feces of patients is a typical manifestation of bowel cancer, wherefore manipulation of the butyrate level with rebalance of microbiota may point out a new direction in cancer treatment or prevention. The number and abundance of Muribaculaceae, Prevotellaceae_UCG-001, Anaerostipes, and Ruminococcaceae were significantly increased with the B. lactis P or B. lactis S supplementation, suggesting that B. lactis P or B. lactis S has a positive role in mediating immune imbalance, colonic inflammation, and development of colonic precancerous lesions. In contrast, Mucispirillum, Clostridia_UCG-014, and Clostridia_vadinBB60 were significantly increased in the AOM/DSS group, while the consequence can be corrected with the B. lactis P or B. lactis S supplementation. Mucispirillum is deemed as a mucus-dwelling pathobiont weakening the intestinal barrier integrity as a result of degradation of host-derived mucin. On the other hand, both the genera Clostridium_UCG-014 and Clostridia_vadinBB60_group belong to the Clostridiaceae family, which are likely involved in colonic inflammation or cancer (Lin et al. 2022), aging (Saeedi Saravi et al. 2021), and high-fat diet murine models (Zhao et al. 2019) in agreement with our results. Fortunately, the AOM/DSS-induced dysbiosis can be reversed with the B. lactis P or B. lactis S supplementation.

The schematic representation of the action of B. lactis on AOM/DSS-induced colonic precancerous lesions is shown in Fig. 8. The compromised body weight, inflammation, ACF, polyps, and gut dysbiosis were demonstrated in the AOM/DSS-treated mice, for which the B. lactis P or B. lactis S supplementation significantly reverses the carcinogenesis process and gut dysbiosis, therefore mitigating the weight loss, colon shortening, and inflammatory responses.

Fig. 8

Schematic of the representative mechanisms for the action of B. lactis on AOM/DSS-induced colonic precancerous lesions. In AOM/DSS-treated mice, the body weight, inflammations, ACF, polyps, and gut dysbiosis were enhanced, while the B. lactis P or B. lactis S administration significantly reversed the carcinogenesis process and gut dysbiosis, as well as mitigated weight losing, colon shorting, and inflammatory response. The present study provided information regarding the B. lactis P or B. lactis S administration in AOM/DSS-induced colonic precancerous lesions and the regulation of inflammation and gut dysbiosis

In conclusion, the present work sheds new light on the modulatory effect of B. lactis P or B. lactis S on counteracting inflammation and gut dysbiosis resulting from the AOM/DSS-induced colonic precancerous lesions. Three possible pathways were proposed to account for the B. lactis P- or B. lactis S–mediated effects against the AOM/DSS-induced colonic precancerous lesions. First, B. lactis P or B. lactis S is in a position to slow down body weight loss, colon shortening, and formation of ACF and polyps. Second, B. lactis P or B. lactis S is capable of modulating immune responses via suppression of inflammatory immune cells, pro-inflammatory cytokines, and the NF-κB signaling pathway. Third, B. lactis P or B. lactis S is adept at rebalancing the gut microbiota reversing the AOM/DSS-induced gut dysbiosis. The overall mechanism exerted by B. lactis P or B. lactis S for the alleviation of the AOM/DSS-induced colonic precancerous lesions is outlined in Fig. 8. We believe that the beneficial effects of B. lactis P and B. lactis S demonstrated herein should pave a new avenue for better diagnosis, treatment, prognosis, and prevention of colorectal cancer.