Introduction

Salmonella Typhimurium (S. Typhimurium) is a major foodborne zoonotic pathogen that can cause diarrhea and colitis in humans and animals (1, 2). It has been reported that Salmonella infections often occur in both developing and developed countries, which poses a substantial risk to public health and causes huge economic loss (3, 4). At present, antibiotic therapies is one of the most effective strategies for Salmonella infections. However, numerous studies have reported that the prolonged use or misuse of antibiotics can lead to environmental pollution, an increase in multi-drug-resistant (MDR) bacteria, as well as gastrointestinal microbiota dysbiosis (5–7). Therefore, there is an emergent search for novel antimicrobial agents to combat Salmonella infections.

Recently, an increasing number of studies have shown that gut microbiota plays an important role in alleviating intestinal inflammation (8) and repairing the intestinal barrier (9, 10). The use of probiotics is a popular approach to modulate intestinal microbiota nowadays. Among all probiotics, Bacillus is one of the most popular probiotic species because of their ability to form endospores that survive gastric transit, storage and delivery conditions (11). In addition, Bacillus species can also produce antimicrobial metabolites against enteric pathogens and provide extracellular enzymes, thereby reducing infection risk and improving nutrient utilization (12, 13).

Bacillus velezensis (B. velezensis), a newly discovered species of the Bacillus genus, can produce an abundance of secondary metabolites with broad-spectrum antibacterial activity (14). Previous studies have reported that B. velezensis exhibits varying degrees of probiotic effects on plants, livestock, poultry and fish. For instance, the combined use of B. velezensis SQR9 mutant and FZB42 improved root colonization and the production of secondary metabolites that are beneficial to cucumber (15). B. velezensis isolated from yaks was shown to enhance growth performance and ameliorate blood parameters related to inflammation and immunity in mice (16). The dietary B. velezensis supplementation can regulate the innate immune response in the intestine of crucian carp and reduce the degree of intestinal inflammation damage induce by A. veronii (17). However, most studies on B. velezensis were mainly biocontrol of plant fungal diseases, with few reports on the treatment or prevention of S. Typhimurium-induced colitis.

In this study, we have screened a strain of B. velezensis HBXN2020 with broad-spectrum antibacterial activity from a large number of Bacillus strains. We aimed to investigate the effectiveness of B. velezensis HBXN2020 in alleviating S. Typhimurium-induced mouse colitis and assessed the biological safety of B. velezensis HBXN2020. Notably, B. velezensis HBXN2020 can alleviate symptoms and colon tissue damage in experimental colitis, as indicated by markers such as S. Typhimurium loads, TNF-α and ZO-1 levels. Moreover, B. velezensis HBXN2020 was also enhanced the homeostasis of colonic microbiota and the abundance of beneficial bacteria (Lactobacillus and Akkermansia). Therefore, this study provides support for the development of probiotic-based microbial products as an alternative prevention strategy for Salmonella infection.

Results

Genomic Characteristics

The Illumina Hiseq and PacBio platforms were used to perform whole genome sequencing of HBXN2020, yielding a total of 8,194,986 Illumina clean reads (1,237,442,886 bases) and 225,593 PacBio reads (1,807,288,759 bases). Following a quality assessment via k-mer and a GC-depth analysis, the whole genome was assembled de novo using Unicycler (Version 0.4.8). The results showed that HBXN2020 has a circular chromosome of 3,929,792 bp (Fig. 1A) with a GC content of 46.5%. The genome of HBXN2020 was predicted to contain 3,857 genes, including 3,744 protein-coding genes (CDS), 86 tRNA genes, and 27 rRNA genes (Table S1).

Fig. 1

A phylogenetic tree based on genome-wide data from all 14 Bacillus strains demonstrated that the HBXN2020 belongs to the B. velezensis species (Fig. 1B). To further understand the classification status of HBXN2020, the online tool JSpeciesWS was used to calculate the average nucleotide identity (ANI) based on the BLAST (ANIb) method. As shown in Fig. 1C, HBXN2020 was found to be a member of the B. velezensis species due to the high percentage of ANIb (more than 97%).

Growth curve of B. velezensis HBXN2020 and its in vitro resistance against environmental assaults

The growth of B. velezensis HBXN2020 was assessed in flat-bottomed 100-well microtiter plates by measuring the values of OD₆₀₀ every hour using an automatic growth curve analyzer. The OD₆₀₀ value of B. velezensis HBXN2020 showed slight fluctuations within the first 2 h, followed by a logarithmic growth phase that reached a stable phase after 10 h (Fig. 2A). Despite environmental restrictions such as limited space and nutrients, the OD₆₀₀ value of the bacteria was not significantly reduced even after 22 h, indicating that the growth of B. velezensis HBXN2020 was relatively stable.

Fig. 2

Next, we evaluated the effect of physical, chemical, and biological sterilization conditions such as high temperature, strong acidity, and enzyme preparation on the survival of B. velezensis HBXN2020. The survival rate of B. velezensis HBXN2020 showed a decreasing trend with increasing temperature. At 65°C, the survival rate of B. velezensis HBXN2020 was less than 1%, while the spores of B. velezensis HBXN2020 exceeded 90% (Fig. 2B). The survival rate of B. velezensis HBXN2020 spores was also above 80% even under high-temperature conditions (95°C). Similar results were observed following exposure to a strong acid environment (pH 2.0) (Fig. 2C), whereby B. velezensis HBXN2020 spores were found to effectively tolerate strong acidic conditions and exhibited improved survival rate. Based on these results, we hypothesized that B. velezensis HBXN2020 spores might also exhibit improved survival in gastrointestinal tract environments. Therefore, we further evaluated the survival rate of B. velezensis HBXN2020 in bile salt (0.3%, v/v), simulated gastric fluid (pepsin 1 mg mL^-1, pH 1.2), and simulated intestinal fluid (trypsin 1 mg mL^-1, pH 6.8). As shown in Fig. 2D-F, B. velezensis HBXN2020 spores demonstrated significantly improved tolerance to these simulated gastrointestinal tract environments.

Antibiotic susceptibility of B. velezensis HBXN2020 and bacteriostasis effect in vitro

To evaluate the drug resistance of B. velezensis HBXN2020, 19 commonly used antibiotics in clinical practice were selected. The results indicated that only polymyxin B had a relatively small inhibition zone diameter (less than 15 mm). Ampicillin, meropenem, minocycline, ofloxacin, and trimethoprim had the strongest inhibition on B. velezensis HBXN2020, with an inhibition zone diameter exceeding 30 mm (Fig. 3A). In addition, the experimental results showed that B. velezensis HBXN2020 was extremely sensitive to β-lactams, tetracyclines, and quinolone drugs. Then, we performed an inhibition test of B. velezensis HBXN2020 against seven common clinical pathogenic bacteria. The results showed that B. velezensis HBXN2020 had a similar inhibitory effect on standard and wild strains of E. coli, Salmonella, S. aureus, and C. perfringens, as well as wild strains of S. suis, P. multocida, and A. pleuropneumoniae. Except for the wild strains of S. suis and A. pleuropneumoniae, the diameter of the inhibition zone of other strains was above 15 mm (Fig. 3B, S3), while the size of the inhibition zone of S. suis and A. pleuropneumoniae was also more than 12 mm.

Fig. 3

Biosafety evaluation of B. velezensis HBXN2020

To determine the safety of B. velezensis HBXN2020, we evaluated its biological safety in a mouse model. After gavage with B. velezensis HBXN2020 spores for two weeks, we observed no significant difference in the body weight of each group of mice (Fig. 4A). The gene expression levels of TNF-α, IL-1β, IL-6, and IL-10 of colon from all groups of mice exhibited no remarkable changes (Fig. 4B). However, in the colon, mRNA levels of the barrier proteins ZO-1 and occludin were trending towards an increase compared with the control group (Fig. 4C). Additionally, blood routine tests and serum biochemistry tests were performed for mice in the control group and H-HBXN2020 group on the 14^th day after oral administration of B. velezensis HBXN2020 spores multiple times. As shown in Fig. 4D and S1, the blood parameters of mice after B. velezensis HBXN2020 treatment, including red blood cells (RBC), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), hemoglobin (HGB), white blood cells (WBC), platelets (PLT), hematocrit (HCT), and mean corpuscular hemoglobin concentration (MCHC), were consistent with those of healthy mice. The serum biochemical parameters, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), albumin (ALB), total bilirubin (TBIL), serum creatinine (CREA), and blood urea nitrogen (BUN), were also within normal limits (Fig. 4E). The corresponding histological analysis of colon tissue from mice receiving low, medium, and high doses of B. velezensis HBXN2020 spores (L-HBXN2020, M-HBXN2020, and H-HBXN2020 groups, respectively) is presented in Fig. 4F. The colon tissue sections of mice in the test groups showed no significant differences compared to the control group. Additionally, there were no observable differences in the major organ tissues (heart, liver, spleen, lung, and kidney) of mice treated with the high dose of B. velezensis HBXN2020 spores compared to healthy mice (Fig. S2). Collectively, these experimental results indicate that B. velezensis HBXN2020 is safe for use and does not cause any adverse side effects in mice.

Fig. 4

Oral administration of B. velezensis HBXN2020 spores alleviated S. Typhimurium-induced colitis

To investigate whether B. velezensis HBXN2020 could inhibit S. Typhimurium ATCC14028 (STm), solid and liquid co-culture tests were conducted, and the results showed that B. velezensis HBXN2020 inhibited the growth of STm in a dose-dependent manner (Fig. 5A, S4A-B, S4D). Next, the therapeutic potential of B. velezensis HBXN2020 was evaluated in an STm-induced mouse colitis model. At days 1, 3, and 5 after STm infection, mice in the STm + HBXN2020 group were orally administered B. velezensis HBXN2020 spores by gavage (Fig. 5B), while the control group and STm + PBS group were treated with sterile PBS. Also, the number of excreted STm in feces was counted daily at the designated time points. As shown in Fig. 5C, there was a significant and continuous reduction in the number of STm in feces following treatment with B. velezensis HBXN2020 spores, while the number of B. velezensis HBXN2020 viable bacteria in feces is also gradually decreasing (Fig. S5A). The number of STm in mouse feces was reduced by 10 to 50 times compared to the STm + PBS group from day 2 to 7 post-treatment. B. velezensis HBXN2020 spore treatment also resulted in a 5 to 10-fold reduction in the number of STm in the ileum, cecum, and colon, respectively (Fig. 5D-E, S5B). The increased inhibition of STm after B. velezensis HBXN2020 spore treatment also resulted in less weight loss and a marked reduction (P < 0.001) in DAI scores in STm + HBXN2020 group mice (Fig. 5F-G). In addition, we also measured the colon length of the mice and found that the STm + PBS group had a significantly shorter colon length (P < 0.001) than the control group. However, compared to the STm + PBS group, the STm + HBXN2020 group exhibited a longer colon (P < 0.05) (Fig. 5H-I).

Fig. 5

To further assess the impact of B. velezensis HBXN2020 on the intestinal tract, we histologically analyzed their colonic tissues. The results showed that mice treated with B. velezensis HBXN2020 spores exhibited a lower histology score (total scoring) (P < 0.05) (Fig. S5C) than mice not treated with the spores and were similar to that of healthy mice (control group). The colon tissue of mice not treated with B. velezensis HBXN2020 spores displayed higher histological scores, with significant crypt deformation, severe mucosal damage, and obvious inflammatory cell infiltration. These findings suggest that B. velezensis HBXN2020 can alleviate colon tissue damage. Moreover, we measured the mRNA levels of inflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukin-1 β (IL-1 β), interleukin-6 (IL-6), and interleukin-10 (IL-10) in colon tissue using RT-qPCR analysis. As shown in Fig. 6B-D, the mRNA levels of TNF-α (P < 0.01), IL-1β (P < 0.05), and IL-6 (P < 0.05) in the colon tissue of STm-infected mice were significantly upregulated, indicating the successful establishment of the STm-induced colitis model. As expected, treatment with B. velezensis HBXN2020 spores led to a significant reduction in the mRNA levels of these pro-inflammatory cytokines, while the mRNA levels of the anti-inflammatory cytokine IL-10 were significantly increased (P < 0.01) (Fig. 6E), suggesting that B. velezensis HBXN2020 spores can alleviate the inflammatory reaction of the colon. Moreover, the mRNA expression levels of colonic epithelial barrier proteins, including ZO-1, ocludin, claudin, and Muc2, were also measured by RT-qPCR. Compared with the PBS-treated group, the B. velezensis HBXN2020 spores treatment group exhibited significantly upregulated mRNA expression levels of ZO-1 (P < 0.05), occludin (P < 0.05), claudin (P < 0.05), and Muc2 (P < 0.01) in the colon tissue (Fig. 6F-I), further demonstrating the ability of B. velezensis HBXN2020 to reduce damage to the colon epithelial barrier.

Fig. 6

Next, we further explored the effect of HBXN2020 spores on the intestinal microbiota composition of STm-treated mice via 16S rRNA gene sequencing. The alpha diversity of colonic microbiota was evaluated using Sobs, Chao1, Shannon, and Simpson indices (OTU levels). Compared with the control and HBXN2020 group, the Sobs, Chao1, and Shannon indices of the STm + PBS group clearly decreased (Fig. 7A-C), while the Simpson index significantly increased (P < 0.05, Fig. 7D). Principle component analysis (PCA) based on Bray–Curtis distance showed that compared with the STm + PBS group, the gut microbiota composition of the STm + HBXN2020 group and the control group were more closer together (R= 0.3988, P = 0.001, Fig. 7E). We analyzed the significance of differences among groups at different classification levels (Fig. 7F-G), and the results revealed that the composition of the gut microbiota changed markedly after STm-induced mice. At the phylum level, the STm + PBS group increased the abundance of Verrucomicrobiota and Proteobacteria and decreased the abundance of Firmicutes, Patescibacteria, Desulfobacterota, and Actinobacteriota compared to the Control group (Fig. 7F). The HBXN2020 group increased the abundance of Bacteroidota and Firmicutes and decreased the abundance of Verrucomicrobiota compared with the Control group. On the contrary, the relative abundance of Firmicutes, Patescibacteria, Desulfobacterota, and Actinobacteriota in the STm + HBXN2020 group was significantly higher than that in the STm + PBS group. At the genus level, infection with S. Typhimurium ATCC14028 in the STm + PBS group dramatically reduced the relative abundance of norank_f_Muribaculaceae and Lactobacillus (Fig. 7H-I) and enhanced the abundance of Bacteroides, Escherichia-Shigella, Salmonella, Alistipes, and Enterococcus compared to the Control group (Fig. 7J-K, S6B). The HBXN2020 group increased the abundance of norank_f_Muribaculaceae and Alloprevotella compared to the Control group. In contrast, STm + HBXN2020 group significantly improved the relative abundance of norank_f_Muribaculaceae and Lactobacillus compared with the STm + PBS group (P < 0.01, P < 0.001, Fig. 7H-I). In addition, LEfSe analysis was also performed to identify specialized microbial communities between the STm + HBXN2020 group and the STm + PBS group, and found that there was a clear difference in gut microbiota composition between them. The microbiota that played a major role in the STm + PBS group were harmful Enterobacterales (Fig. 7L), while the STm + HBXN2020 group was Candidatus_Saccharimonas (Fig. 7L).

Fig. 7

Prophylactic B. velezensis HBXN2020 spores alleviated S. Typhimurium-induced colitis

Based on the improved therapeutic efficacy of B. velezensis HBXN2020 in the treatment of STm-induced colitis in mice, we further explored its potential for disease prevention by evaluating pre-treatment. Mice in the PBS + STm group and HBXN2020 + STm group were given the same amount of sterile PBS and B. velezensis HBXN2020 spores by gavage one week in advance. As shown in Fig. 8, compared with the PBS + STm group, oral administration of B. velezensis HBXN2020 spores as a preventive measure remarkably alleviated STm-induced colitis, including weight loss of mice (P < 0.01) (Fig. 8B), a significant reduction in DAI (P < 0.001) (the comprehensive score of weight loss, stool consistency, and blood in the feces, Fig. 8C), and the prevention of colon length shortening (Fig. 8D-E). However, the number of B. velezensis HBXN2020 viable bacteria in feces is also gradually decreasing with the time prolonging (Fig. S7A). Furthermore, compared with the PBS + STm group, oral administration of B. velezensis HBXN2020 spores not only reduced the number of STm in mouse feces (Fig. 8F) but also decreased STm colonization in the ileum, cecum, and colon (Fig. 8G-H, S7B).

Fig. 8

Histological analysis further revealed that oral administration of B. velezensis HBXN2020 significantly reduced the infiltration of inflammatory cells, mucosal damage and the overall histological score of colon tissue (Fig. 9A, S7C). Additionally, the mRNA expression levels of pro-inflammatory cytokines, including TNF-α (P < 0.05), IL-1β (P < 0.05) and IL-6 (P < 0.01), in the colon tissue of mice pretreated with B. velezensis HBXN2020 spores were significantly decreased (Fig. 9B-D), while the mRNA levels of anti-inflammatory cytokine (IL-10) were significantly higher (P < 0.05) (Fig. 9E) than mice pretreated with PBS. In addition, compared to the group pretreated with PBS, the group pretreated with B. velezensis HBXN2020 spores showed a significant increase in mRNA expression levels of ZO-1 (P < 0.001), ocludin (P < 0.05), and claudin (P < 0.05) in the colon tissue of mice (Fig. 9F-H). While there was no significant difference in Muc2 expression (p=0.252) (Fig. 9I), the results were consistent with those of the therapeutic experiments. Overall, these findings suggest that B. velezensis HBXN2020 can effectively alleviate the symptoms of STm-induced colitis and the associated damage to colon tissue.

Fig. 9

Next, we examined the impact of prophylactic B. velezensis HBXN2020 on the intestinal microbiota composition of STm-treated mice. Sobs, Chao1 index and Shannon index have changed to some extent between groups, but differences were not significant (Fig. 10A-C). Compared with the Control group, the Simpson index significantly decreased in the PBS+STm group (P < 0.05, Fig. 10D). However, compared to the PBS+STm group, Simpson index in the HBXN2020+STm group did not change (Fig. 10D). PCA analysis showed a significant separation between control and PBS+STm groups, while the HBXN2020+STm group, HBXN2020 group, and the control group were clustered together (Fig. 10E). Next, we analyzed microbial community composition at the phylum and genus levels (Fig. 10F-G). At the phylum level, the PBS + STm group increased the abundance of Bacteroidota, Proteobacteria, Actinobacteriota, and Deferribacterota and reduced the abundance of Verrucomicrobiota, Firmicutes, Patescibacteria, and Desulfobacterota compared with the Control group(Fig. 10F, S8A). The HBXN2020 group increased the abundance of Bacteroidota compared to the control group. In contrast, compared to the PBS+STm group, the abundance of Firmicutes in the HBXN2020+STm group was significantly increased (P < 0.05). At the genus level, a depletion of Lactobacillus and Akkermansia (Fig. 10H-I) and an enrichment of Alistipes, Bacteroides, Escherichia-Shigella, Alloprevotella, Helicobacter, Enterococcus, and Salmonella were observed in PBS + STm group compared with the control (Fig. 10J-M, 8SB). The enrichment with norank_f_Muribaculaceae and Alloprevotella were observed in HBXN2020 group compared to the control. In contrast, an enrichment of Lactobacillus, Akkermansia, and Lachnospiraceae_NK4A136_group and a sharp decrease of Alistipes, Bacteroides, Escherichia-Shigella, and Enterococcus were observed in HBXN2020 + STm group compared to the PBS + STm group. LEfSe analysis showed that the abundance of Lactobacillus in the intestinal microbiota of mice was significantly increased after B. velezensis HBXN2020 treatment, while the abundance of harmful Enterobacterales in the untreated group was clearly increased (Fig. 10N).

Fig. 10

Discussion

S. Typhimurium is an important intestinal pathogen that can cause invasive intestinal diseases such as bacterial colitis (18, 19). Although antibiotics are commonly used to treat bacterial-induced colitis, their overuse has significantly increased bacterial resistance in recent years, resulting in serious environmental pollution. A recent study showed that antibiotic-resistance genes in probiotics could be transmitted to the intestinal microbiota, which may threaten human health (20). Therefore, the source of probiotics is particularly important. Black pigs are an excellent domestic breed in China, with strong environmental adaptability and disease resistance (21). In this study, B. velezensis HBXN2020 was isolated from the free-range feces of black piglets in a mountain village in Xianning City (Hubei, China) and exhibited excellent antibacterial activity. Previous studies have demonstrated that the difficidins isolated from B. subtilis can inhibit bacterial protein synthesis, giving it broad-spectrum antibacterial activity against Gram positive and Gram negative bacteria. Therefore, we speculate that B. velezensis HBXN2020 has the ability to synthesize Difficidin or its derivatives (22). Otherwise, in vitro tolerance assays showed that B. velezensis HBXN2020 spores have good tolerance to high temperature, strong acids, bile salts, and simulated gastric and intestinal fluid, which was similar to previous research results (23). Furthermore, antibiotic susceptibility tests showed that B. velezensis HBXN2020 does not exhibit antibiotic resistance.

Probiotics are unique in that they are alive when consumed, and unlike food or drug, they have the potential to produce toxins or infections in the body (24, 25). Therefore, assessing the biosafety of probiotics is crucial. In this study, the safety of B. velezensis HBXN2020 was evaluated by measuring the body weight, intestinal barrier proteins, and inflammatory cytokines of the experimented mice, as well as conducting routine blood and biochemical tests. The results showed that the expression levels of ZO-1 and occludin in the colon of mice increased after treatment with different doses of B. velezensis HBXN2020 spores. It has been reported that ZO-1, occludin and claudin are the three most important tight junction proteins in intercellular connections, which play a crucial role in maintaining the intestinal epithelial barrier (26, 27). Therefore, upregulating the expression levels of ZO-1 and Occludin might enhance intestinal barrier function.

Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are important aminotransferases in animals and are considered key parameters for evaluating liver function injury (28, 29). Under normal circumstances, the levels of ALT and AST are in dynamic balance without notable changes. However, when the liver is damaged or becomes dysfunctional, the levels of aminotransferases (ALT and AST) increase significantly (30). Previous studies have shown that CCL4-induced liver injury strongly increases plasma ALT/AST levels (31). Zhang et al. reported that the addition of Bacillus subtilis to chicken diets significantly decreased the serum levels of ALT and AST (32). In this present study, the levels of ALT and AST in the probiotic-treated group were slightly lower than those in the control group, indicating that B. velezensis HBXN2020 had no negative effect on liver health in mice. Furthermore, routine blood tests found that the blood indices of the B. velezensis HBXN2020 treatment group were within the normal reference range, similar to that of the control group.

To further evaluate the potential of B. velezensis HBXN2020 spores in alleviating S. Typhimurium-induced colitis, mice were challenged with S. Typhimurium ATCC14028 and then treated with B. velezensis HBXN2020 spores via oral administration. Our experimental results demonstrated that oral treatment with B. velezensis HBXN2020 spores could alleviate S. Typhimurium-induced colitis, as evidenced by decreased weight loss, DAI, and histological damage, which is similar with prior studies (33, 34). Previous research showed that macrophages are the first line of host defense against bacterial infection. They release pro-inflammatory cytokines, which are critical in initiating adaptive immune responses (35, 36). Nevertheless, pro-inflammatory cytokines have immunological properties that can be beneficial for the host to resist the invasion of bacteria and other microbes in the surrounding environment (37). For instance, pre-treatment with recombinant murine TNF-α was shown to protect mice against lethal bacterial (E. coli) infection (38). PJ-34 exerted protective effects on intestinal epithelial cells against invasive Salmonella infection by upregulating IL-6 expression through the ERK and NF-κB signal pathways (39). However, some studies have shown that excessive immune cell activation can lead to tissue damage, multi-organ malfunction, systemic or chronic inflammation, and autoimmune disorders (40, 41). IL-10 is a recognized anti-inflammatory mediator that plays a crucial role in maintaining intestinal microbe-immune homeostasis, regulating the release of inflammatory mediators, and inhibiting pro-inflammatory responses of innate and adaptive immunity (42–44). For instance, previous studies have shown that Clostridium butyrate can induce the production of IL-10 in the intestine, thereby alleviating experimental colitis in mice (45). Similar results were observed in our study, where oral administration of B. velezensis HBXN2020 spores reduced the expression levels of TNF-α, IL-1β and IL-6 while increasing the levels of IL-10 in the colon of mice. Furthermore, oral administration of B. velezensis HBXN2020 spores improved the functional damage and integrity of the intestinal barrier caused by S. Typhimurium infection, as evidenced by the upregulated expression levels of intestinal barrier proteins and reduced histological damage.

Subsequently, we further analyzed the colon microbiome and it was noteworthy that oral B. velezensis HBXN2020 markedly increased alpha diversity of the colonic microbiota. As reported in prior studies, elevated alpha diversity associates with enhanced anti-interference ability (46). PCA analysis Results showed that oral administration of B. velezensis HBXN2020 significantly affected the structure of the intestinal microbiota. Further analysis at the phylum/genus level revealed that B. velezensis HBXN2020 treatment reduced the increase of Bacteroides, Escherichia-Shigella, and Salmonella (harmful to intestinal homeostasis) in colitis (47), but enhanced the relative abundance of Lactobacillus (known for enhancing the epithelial barrier by increasing mucus secretion and upregulating the expression of tight junction proteins such as claudin-1, occludin, and ZO-1) (48) and Akkermansia (known for modulating intestinal immune response by producing short chain fatty acids and reducing the secretion of pro-inflammatory cytokines) (49). Bacteroides is generally ‘friendly’ commensals in the intestine, but they can enter normal tissues through the intestinal mucosa when intestinal barrier function becomes compromised (50), leading to inflammation, diarrhea, and abdominal abscess (51). Escherichia-Shigella is a common intestinal pathogen, and its relative abundance is associated with pro-inflammatory states (52). Salmonella can secrete bacterial virulence proteins, which are directly delivered to the host cell cytosol through the Type III secretion system and induce the expression of pro-inflammatory cytokines (53), leading to inflammation. Overall, the structure of the colon microbiota was disrupted by S. Typhimurium ATCC14028 infection and increased the relative abundance of harmful bacteria, while therapeutic B. velezensis HBXN2020 enhanced the abundance of beneficial bacteria, reduced inflammation, and improved intestinal barrier function.

Due to the improved efficacy of treating S. Typhimurium-induced colitis, we further investigated the potential of B. velezensis HBXN2020 spores for disease prevention by evaluating pre-treatment. Similarly to the therapeutic effect of oral B. velezensis HBXN2020, pre-treatment with B. velezensis HBXN2020 spores effectively alleviated the symptoms of S. Typhimurium-induced colitis and reduced the expression levels of pro-inflammatory cytokines and increased anti-inflammatory cytokine levels in the colon of S. Typhimurium-infected mice. Furthermore, pre-treatment with B. velezensis HBXN2020 spores also promoted the expression levels of intestinal barrier proteins in the colon of S. Typhimurium-infected mice. In addition, the microbiota composition was also dramatically modulated by prophylactic B. velezensis HBXN2020 in mice with colitis, although the community richness and diversity revealed by alpha diversity index was not significantly impacted in mice with colitis by prophylactic-treatment. Further analysis at the genus level revealed that prophylactic B. velezensis HBXN2020 clearly increased SCFAs-producing bacteria including Lactobacillus and Akkermansia, which was consistent with the results of therapeutic B. velezensis HBXN2020. Similarly, harmful bacteria such as Alistipes, Bacteroides, Escherichia-Shigella, and Enterococcus were significantly enriched in the PBS+STm group. Alistipes is a facultative - pathogenic bacteria and promotes the development of inflammation, obesity, and colorectal cancer (54). Based on those results, we found that prophylactic B. velezensis HBXN2020 alleviates S. Typhimurium induced colitis by increasing the abundance of beneficial bacteria, attenuating pro-inflammatory response of the intestine, and increasing intestinal barrier function.

Conclusion

In conclusion, infection with S. Typhimurium ATCC14028 disrupts the intestinal mucosal barrier, and leading to intestinal inflammation. Both therapeutic and prophylactic, supplementing B. velezensis HBXN2020 spores alleviates S. Typhimurium-induced colonic injury by reducing Salmonella colonization, and enhancing the abundance of beneficial bacteria (especially Lactobacillus and Akkermansia) and intestinal barrier integrity, and decreasing inflammation.

Experimental procedures

Animals and bacterial strains

Female specific-pathogen-free (SPF) C57BL/6 mice aged 6 to 8 weeks were purchased from the Animal Experimental Center of Huazhong Agricultural University. All mice experiments were conducted in the standard SPF facility of Huazhong Agricultural University, with 12 h of light and 12 h of darkness at a temperature of 25°C and ad libitum access to food and water. The use of animals in this experiment was approved by the Animal Care and Ethics Committee of Huazhong Agricultural University (Ethics Approval Number: HZAUMO-2023-0089).

The strains used in this study are listed in Table S2, and all the primers used in Table S3. Clostridium perfringens were cultured in a fluid thioglycollate medium (FTG) (Hopebio, Qingdao, China) at 45°C. Streptococcus suis, Pasteurella multocida, and Actinobacillus pleuropneumoniae were cultured in tryptic soy broth (TSB) (BD Biosciences, MD, United States) medium supplemented with 5% (v/v) sheep serum (Solarbio, Beijing, China) at 37°C, while the other bacterial strains were cultured in Luria Bertani (LB) medium (Solarbio, Beijing, China) at 37°C. In addition, the Difco sporulation medium (DSM) was used for inducing the sporulation of B. velezensis HBXN2020 via the nutrient depletion method (55).

Stain isolation, growth conditions and strain identification

In this study, we isolated 362 Bacillus strains from the feces of healthy black piglets that were free-ranging in a mountain village in Xianning City, Hubei Province, China, and 4 Bacillus strains were screened through initial antibacterial tests. Then, a strain with broad-spectrum antimicrobial activity was ultimately screened out by the antibacterial spectrum assay (Table S4), and named as B. velezensis HBXN2020. The strain was stored in 28% glycerol at −80°C. For culturing the bacterial cells, Luria-Bertani (LB) media or agar plates supplemented with 1.5% agar were used and incubated at 37°C.

A single bacterial colony was inoculated into 10 mL of LB medium and incubated overnight at 37°C with shaking (180 rpm). Then, the 16S rRNA sequence was amplified by the polymerase chain reaction (PCR) using forward primer 27F and reverse primer 1492R and similarity analysis with the National Center for Biotechnology Information (NCBI) alignment tool (BLAST) was used to identify the species.

Genome sequencing, annotation and analysis

The complete genome of B. velezensis HBXN2020 was sequenced using the Illumina HiSeq PE and PacBio RSII (SMRT) platforms (Shanghai Majorbio Bio-pharm Technology Co., Ltd.). Briefly, the short reads from the Illumina HiSeq PE were assembled into contigs using SPAdenovo (http://soap.genomics.org.cn/). The long reads from the PacBio RSII and the Illumina contigs were then aligned using the miniasm and Racon tools in Unicycler (version 0.4.8, https://github.com/rrwick/Unicycler) to generate long-read sequences. During the assembly process, sequence correction was performed using Pilon (version 1.22, https://github.com/broadinstitute/pilon/wiki/Standard-Output). Lastly, a complete genome with seamless chromosomes was obtained.

The coding sequences (CDs) of the B. velezensis HBXN2020 genome were predicted using Glimmer (version 3.02, http://ccb.jhu.edu/software/glimmer/index.shtml). The tRNA and rRNA genes were predicted using TRNAscan-SE (version 2.0, http://trna.ucsc.edu/software/) and Barrnap (version 0.8, https://github.com/tseemann/barrnap), respectively. The functional annotation of all CDSs was performed using various databases, including Swiss-Prot Database (https://web.expasy.org/docs/swiss-prot_guideline.html), Pfam Database (http://pfam.xfam.org/), EggNOG Database (http://eggnog.embl.de/), GO Database (http://www.geneontology.org/), and KEGG Database (http://www.genome.jp/kegg/). The circular map of the B. velezensis HBXN2020 genome was generated using CGView (version 2, http://wishart.biology.ualberta.ca/cgview/) (56).

Comparative genomic analysis

To elucidate the phylogenetic relationships from a whole-genome perspective, a phylogenetic tree based on the whole genome was constructed using the Type (Strain) Genome Server (TYGS) online (https://ggdc.dsmz.de/) (57). The average nucleotide identity (ANI) values of genome to genome were calculated using the JSpeciesWS online service (58), and a heatmap was then generated using TBtools (version 1.113) (59).

Growth curves of B. velezensis HBXN2020

The growth curve of B. velezensis HBXN2020 was recorded in flat-bottomed 100-well microtiter plates via detecting optical density at 600 nm (OD₆₀₀) at 1 h intervals using the automatic growth curve analyzer (Bioscreen, Helsinki, Finland).

Antimicrobial Assays

The in vitro antagonistic activity of B. velezensis HBXN2020 was tested using the agar well-diffusion method against 18 indicator strains (pathogens), which included 7 standard strains (E. coli ATCC 25922, E. coli ATCC 35150, S. Typhimurium ATCC 14028 (STm), S. Typhimurium SL1344, S. aureus ATCC 29213, S. aureus ATCC 43300, and C. perfringens CVCC 2030) and 11 clinical isolates (E. coli EC024, S. Enteritidis SE006, S. aureus S21, C. perfringens CP023, C. perfringens CP002, S. suis SC19, S. suis SS006, P. multocida PM002, P. multocida PM008, A. pleuropneumoniae APP015, and A. pleuropneumoniae APP017). All plates were cultured at 37°C for 16 h before observing the inhibition zone, and the diameter of the inhibition zone was measured using a vernier caliper. The presence of a clear zone indicated antagonistic activity.

Antibiotic susceptibility assays

Antimicrobial susceptibility testing was performed using the Kirby-Bauer (KB) disk diffusion method in accordance with the Clinical Laboratory Standards Institute (CLSI) guidelines (60). The antimicrobial agents tested were: Ampicillin, Meropenem, Piperacillin, Gentamycin, Tetracycline, Doxycycline, Minocycline, Erythromycin, Enrofloxacin, Ofloxacin, Sulfamethoxazole, Trimethoprim-sulfamethoxazole, Polymyxin B, Teicoplanin, Trimethoprim, Florfenicol, Spectinomycin, Nitrofurantoin, and Rifampicin. The diameter of the inhibition zone was measured using a vernier caliper.

In vitro resistance assay of B. velezensis HBXN2020

B. velezensis HBXN2020 spores (100 μL) or vegetative cells (100 μL) were separately resuspended in 900 μL of LB medium supplemented with different pH values (2, 3, 4, 5 or 6), bile salts (0.85% NaCl, 0.3%), simulated gastric fluid (SGF, HCl, pH 1.2) containing 10 g L^-1 of pepsin in 0.85% NaCl solution, or simulated intestinal fluid (SIF, NaOH, pH 6.8) containing 10 g L^-1 of trypsin in 0.05 M KH2PO4 solution, and incubated at 37 °C. A normal LB medium (pH 7.0) was used as the control. At predetermined time points, 100 μL was taken from each sample, serially diluted 10-fold with sterile PBS (pH 7.2), and then spread onto LB agar plates. The plates were incubated overnight in a constant temperature incubator at 37°C, and the bacterial colonies were counted. The survival rate was calculated using the following formula: Survival rate = (number of bacteria in the treatment group/number of bacteria in the control group) × 100%.

One milliliter of B. velezensis HBXN2020 spores or vegetative cells were separately placed in water baths at different temperatures (37°C, 45°C, 55°C, 65°C, 75°C, 85°C or 95°C) for 20 min, with a 37°C water bath used as the control. The survival rate was calculated as described above.

In vitro bacterial competition

To investigate whether B. velezensis HBXN2020 directly inhibited the growth of STm, we performed spot-on lawn and agar-well diffusion assays, as well as co-culture assays in liquid culture medium. In the spot-on lawn antimicrobial assays, we prepared double layers of agar by first pouring LB agar into the plate as the bottom layer. The top layer consisted of 10 mL TSB broth containing 0.7% agar with STm overnight culture. Then, 10 μL of B. velezensis HBXN2020 overnight culture and cell-free supernatant (CFS) were respectively spotted onto TSB agar and incubated at 37°C for 12 h to measure the inhibition zone. A transparent zone of at least 1 mm around the spot was considered positive.

The antagonistic effect of B. velezensis HBXN2020-CFS against STm was determined using the agar-well diffusion assays. To collect B. velezensis HBXN2020-CFS, the culture was centrifuged at 9,000g for 15 min at 4°C, and the supernatant was filtered through a 0.22 μm membrane filter (Millipore, USA). The STm lawn medium was prepared by mixing 10 mL of TSB broth containing 0.7% agar with STm overnight culture and then poured into a sterile plate covered with LB agar and Oxford cups, 8 mm diameter wells were prepared in the TSB agar after removing the cups. The wells were filled with 100 μL of B. velezensis HBXN2020-CFS, LB medium or ampicillin (100 µg mL^-1). The plates were incubated at 37°C for 14 h, and the inhibition zone was measured.

The co-culture assay was conducted by incubating B. velezensis HBXN2020 and S. Typhimurium ATCC14028 (carry pET28a (+), kanamycin resistance) (33) separately overnight at 37°C and diluting them to 10⁴ CFU/mL. Then, the two strains were mixed at different ratios (1:1, 1:10, 1:50 or 1:100) and co-cultured at 37°C with shaking (180 rpm). At predetermined time points, serial 10-fold dilutions were prepared for all samples and spread onto selective (kanamycin 50 µg mL^-1) LB agar plates and cultured at 37°C for 12 h before bacterial counting. Viable colony counts ranged from 30 to 300 per plate.

Biosafety assessment

After a 7-day acclimation period (free access to water and food), mice were randomly divided into four treatment groups (n = 5): low-dose B. velezensis HBXN2020 spores group (L-HBXN2020 group, 10⁷ CFU/mouse), medium-dose B. velezensis HBXN2020 spores group (M-HBXN2020 group, 10⁸ CFU/mouse), high-dose B. velezensis HBXN2020 spores group (H-HBXN2020 group, 10⁹ CFU/mouse), and a control group. During the experimental period (15 days), mice were weighed and orally gavaged with their respective treatments once every two days. On day 15, all mice were euthanized, and blood, heart, liver, spleen, lung, kidney, ileum, cecum, and colon were collected. Blood samples were used for routine blood and biochemistry tests. Major organ tissues (heart, liver, spleen, lung, and kidney) and a 5-mm distal segment of the colon were used for histopathology. The remaining colonic tissues were rapidly frozen in liquid nitrogen and stored at −80°C for cytokine and tight junction protein expression analysis.

Salmonella Typhimurium-induced mouse model of colitis

After 7 days of acclimation (free access to water and food), mice were randomly divided into three treatment groups (n = 8): Control group, STm + PBS group, and STm + HBXN2020 group. The S. Typhimurium-induced colitis was performed as previously described (61), with slight modifications. On the first day of the experiment, all mice in the STm + PBS group and STm + HBXN2020 group were orally inoculated with 200 μL (5×10⁷ CFU/mouse) of STm. On days 1, 3 and 5 following STm infection, each mouse in the STm + HBXN2020 group received 200 μL (1×10⁸ CFU/mouse) of B. velezensis HBXN2020 spores via gavage administration. In contrast, the control group and STm + PBS group received 200 μL of sterile PBS by oral gavage. Fecal samples were collected daily following STm infection and resuspended in sterile PBS. The number of STm in mice feces from both the STm + PBS and STm + HBXN2020 groups was then determined by spreading a serial 10-fold dilution on selective LB agar plates containing 50 µg mL^-1 kanamycin. Throughout the entire experiment, the body weight, stool consistency and fecal occult blood of all mice were monitored daily. As shown in Table S5, disease activity index (DAI) was calculated by the sum of the scores from three parameters (62). On day 7 after STm infection, all mice were euthanized, their ileum, cecum and colon were collected. The length of colon was measured, and a 5-mm distal segment of the colon was fixed in 4% paraformaldehyde for further histopathology. The remaining colon was then rapidly frozen in liquid nitrogen and stored at −80°C for cytokine and tight junction protein expression analysis. Then, the number of STm in the ileum, cecum, and colon was determined by spreading serial 10-fold dilutions on selective LB agar plates.

To investigate the prophylactic efficacy of B. velezensis HBXN2020 in ameliorating STm-induced colitis, another independent experiment was conducted using six-week-old female C57BL/6 mice (SPF). After a seven-day acclimation period, mice were randomly assigned to three groups (n = 8): control group, PBS + STm group and HBXN2020 + STm group. On days 1, 3, 5 and 7, each mouse in the HBXN2020 + STm group received 200 μL (1×10⁸ CFU/mouse) of B. velezensis HBXN2020 spores via gavage administration. Meanwhile, mice in the control group and PBS + STm group received 200 μL of sterile PBS by oral gavage. On day 7, all mice in the PBS + STm group and HBXN2020 + STm group were orally inoculated with 200 μL (5×10⁷ CFU/mouse) of STm. Fecal samples were collected daily following STm infection from both groups and resuspended in sterile PBS. The number of STm in the feces of mice was then determined by spreading serial 10-fold dilutions on selective LB agar plates (50 µg mL^-1 kanamycin). Throughout the entire experiment, the body weight and DAI scores of all mice were monitored daily. On day 12, all mice were euthanized, and the ileum, cecum, and colon were collected. The colon length was measured, and a 5-mm distal segment of the colon was fixed in 4% formalin for sectioning and staining. The remaining colon was stored at −80°C for future analysis. Lastly, the number of STm in the ileum, cecum, and colon was determined using a selective LB agar plate.

Determination of cytokines and tight junction protein expression in colon tissue

Total RNA was extracted from colon tissues using TRIpure reagent (Aidlab, China), and cDNA was obtained using HiScript^® III RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China). RT-qPCR for each gene was performed in triplicate using qPCR SYBR Green Master Mix (Yeasen Biotechnology, Shanghai, China). The relative expression level of cytokine and tight junction protein genes was calculated using the 2^−ΔΔCt method with β-actin and GAPDH as reference genes. The primer sequences used in the RT-qPCR test are listed in Table S3.

Histopathology analysis

Colon tissue samples (0.5 cm) were fixed in 4% paraformaldehyde for 24 h, and the fixed tissues were then embedded in paraffin and sectioned. The sections were stained with hematoxylin and eosin (H&E) and observed and imaged using an optical microscope (Olympus Optical, Tokyo, Japan). The histopathological score included the degree of inflammatory infiltration, changes in crypt structures, and the presence or absence of ulceration and edema. The scoring criteria were determined as previously described (34).

16S rRNA gene sequencing and analysis

According to the manufacturer’s instructions, colon microbial community genomes were extracted using E.Z-N.A ® Stool DNA Kit (Omega; D4015-01), and quality was detected by 1% agarose gel electrophoresis. The 16S rRNA V3-V4 variable region was amplified by PCR using universal primers 338F (5’-ACTCCTACGGGGGGCAG-3’) and 806R (5’-GACTACHVGGGTWTCTAAT-3’). The PCR products were examined by electrophoresis on 2% agarose gels and then purified with the AxyPrep DNA gel extraction kit (Axygen Biosciences, USA). A sequencing library was constructed using the NEXTFLEX® Rapid DNA-Seq Kit, and sequencing was performed using the Illumina MiSeq platform. Raw reads were quality evaluated and filtered by fastp (version 0.20.0) and merged using FLASH (version 1.2.7). The optimized sequences were clustered into operational taxonomic units (OTUs) based on 97% sequence similarity using UPARSE (version 7.1). The representative sequences of each OTU was classified by RDP classifier (version 2.2; confidence threshold value, 0.7). Alpha diversity was assessed using the ACE, chao, Shannon, and Simpson indices. The β-diversity analysis was performed using Bray-Curtis distances and visualized through principal component analysis (PCA). Linear discriminant analysis (LDA) effect size (LefSe) was used to identify differential microbiota between groups.

Statistical Analysis

Statistical analysis was performed using GraphPad Prism 8.3.0 (GraphPad Software, San Diego, CA, USA) and Excel (Microsoft, Redmond, USA). Data are presented as mean ± standard error of the mean (SEM). Differences between two groups were evaluated using two-tailed unpaired Student’s t-test, and all other comparisons were conducted using one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test. For all analyses, significance differences are denoted as: *, P < 0.05, **, P < 0.01, and ***, P < 0.001.

Data availability

The complete genome sequence of Bacillus velezensis HBXN2020 has been deposited in GenBank and assigned accession numbers: CP119399.1

Conflict of interest

We declare no conflict of interest.

Acknowledgements

This work was supported by grants from the National Program on Key Research Project of China (2021YFD1800300, 2022YFD1800800), “Yingzi Tech & Huazhong Agricultural University Intelligent Research Institute of Food Health” (No. IRIFH202209; IRIFH202301), and the Fundamental Research Funds for the Central Universities (2662016PY004).

Author contributions

Linkang Wang: Conceptualization, Methodology, Data curation, Formal analysis, Writing - original draft. Haiyan Wang, Xinxin Li, Mengyuan Zhu: Methodology, Data curation, Planned and performed the experiments. Dongyang Gao, Dayue Hu, Zhixuan Xiong: Methodology, Writing - Reviewing and Editing. Xiangmin Li: Supervision, Writing-reviewing and editing. Ping Qian: Conceptualization, Funding acquisition, Project administration, Supervision, Writing-Reviewing and Editing. All authors read and approved the submitted version of the paper.

Funding information

This work was supported by grants from the National Program on Key Research Project of China (2021YFD1800300, 2022YFD1800800), “Yingzi Tech & Huazhong Agricultural University Intelligent Research Institute of Food Health” (No. IRIFH202209; IRIFH202301), and the Fundamental Research Funds for the Central Universities (2662016PY004).