Dietary Bacillus licheniformis shapes the foregut microbiota, improving nutrient digestibility and intestinal health in broiler chickens

Bacillus licheniformis is considered a potential alternative to antibiotic growth promoters of animal growth and health. However, the effects of Bacillus licheniformis on the foregut and hindgut microbiota, and their relationships with nutrient digestion and health, in broiler chickens remain unclear. In this study, we aimed to identify the effects of Bacillus licheniformis BCG on intestinal digestion and absorption, tight junctions, inflammation, and the fore- and hind-gut microbiota. We randomly assigned 240 1-day-old male AA broilers into three treatment groups: CT (basal diet), BCG1 (basal diet + 1.0 × 108 CFU/kg B. licheniformis BCG), and BCG2 (basal diet + 1.0 × 109 CFU/kg B. licheniformis BCG). On day 42, the jejunal and ileal chyme and mucosa were subjected to analysis of digestive enzyme activity, nutrient transporters, tight junctions, and signaling molecules associated with inflammation. The ileal and cecal chyme were subjected to microbiota analysis. Compared with the CT group, the B. licheniformis BCG group showed significantly greater jejunal and ileal α-amylase, maltase, and sucrase activity; moreover, the α-amylase activity in the BCG2 group was higher than that in the BCG1 group (P < 0.05). The transcript abundance of FABP-1 and FATP-1 in the BCG2 group was significantly greater than that in the CT and BCG1 groups, and the GLUT-2 and LAT-1 relative mRNA levels were greater in the BCG2 group than the CT group (P < 0.05). Dietary B. licheniformis BCG resulted in significantly higher ileal occludin, and lower IL-8 and TLR-4 mRNA levels than observed in the CT group (P < 0.05). B. licheniformis BCG supplementation significantly decreased bacterial community richness and diversity in the ileum (P < 0.05). Dietary B. licheniformis BCG shaped the ileac microbiota by increasing the prevalence of f_Sphingomonadaceae, Sphingomonas, and Limosilactobacillus, and contributed to nutrient digestion and absorption; moreover, it enhanced the intestinal barrier by increasing the prevalence of f_Lactobacillaceae, Lactobacillus, and Limosilactobacillus. Dietary B. licheniformis BCG decreased microbial community diversity by diminishing Desulfovibrio, Alistipes, Campylobacter, Vibrio, Streptococcus, and Escherichia coli-Shigella levels, and down-regulating inflammatory associated molecule expression. Therefore, dietary B. licheniformis BCG contributed to digestion and absorption of nutrients, enhanced the intestinal physical barrier, and decreased intestinal inflammation in broilers by decreasing microbial diversity and optimizing the microbiota structure.


Introduction
In poultry production, broiler chickens are generally subjected to harsh and stressful conditions, particularly when they are reared at high stocking density, thus resulting in host stress and immune dysfunction (Kridtayopas et al., 2019). This subhealth status decreases the growth potential and causes intestinal dysfunction in broilers, and increases the economic costs of rearing. The intestines not only digest and absorb nutrients, but also are the largest immune organ . Trillions of microorganisms colonize the gastrointestinal tract; the total microbial number is 10 times the number of host somatic cells, and the collective number of genes is 150 times that in the host genome (Collins et al., 2012;Strandwitz, 2018). The intestinal microbiota co-develops with the host and participates in nutrient digestion, improves intestinal development and health, and regulates the body's metabolism and immunity function (Li et al., 2019;Han et al., 2020). Gut microbiota disorders have been associated with subhealth and disease. Broilers show a shift in the gut microbial profile when they experience subclinical forms of necrotic enteritis: the relative abundance of Firmicutes, Lactobacillus, and Bacteroides decreases, thus resulting in low host productivity (Antonissen et al., 2016). Other research has indicated that challenge with Salmonella typhimurium decreased Lactobacillus prevalence, damaged intestinal morphology, and subsequently decreased the growth performance of broilers (Jazi et al., 2019). The overall consensus is that the intestinal microbiota is interlinked with intestinal health and poultry growth.
Probiotics have been demonstrated to be an effective means of promoting animal growth and improving body health in the post-antibiotic era (Mingmongkolchai and Panbangred, 2018). Bacillus licheniformis (B. licheniformis) is a Gram-positive bacterium characterized by resistance to stresses such as high acidity and temperature. Because of these characteristics, B. licheniformis can be used in livestock production practices. Bacillus spores are metabolically dormant under harsh conditions including feed pelleting but subsequently grow in the favorable environment of the gastrointestinal tract after ingestion (Konieczka et al., 2018). B. licheniformis shows growth-promoting effects in poultry (Gadde et al., 2017). Moreover, B. licheniformis benefits broilers by protecting against heat stress and preventing necrotic enteritis (Abdelqader et al., 2020;Xu et al., 2021). These benefits might be attributable to a variety of biologically active substances produced by B. licheniformis, which contribute to feed digestibility, immune system regulation, and enhanced intestinal barrier function (Kim et al., 2004;Zhou et al., 2016;Kan et al., 2021). B. licheniformis improves the intestinal mechanical barrier and decreases intestinal permeability by upregulating the gene expression of mucins and tight junction proteins in laying hens (Wang et al., 2017). Diets containing a mixture of B. licheniformis and B. subtilis have been found to alleviate Escherichia coli-induced enteritis by increasing intestinal epithelial barrier integrity (Yang et al., 2016). The potential underlying mechanism involves regulation of the composition of the intestinal microbiota to restore and maintain intestinal homeostasis (Sanders et al., 2019;Zhao et al., 2022b). For instance, probiotic Bacillus strains have been found to prevent or diminish gut colonization by Chlamydia psittaci, Escherichia coli, Streptococcus, and Salmonella, thus improving intestinal mucosa integrity and gut health (Zuo et al., 2020;Haque et al., 2021). Therefore, probiotic Bacillus appears to prevent disease or stress, and promote growth performance, possibly through an optimized intestinal microbial structure and improved gut health.
Although B. licheniformis has great potential application value in the broiler industry, the efficacy of probiotic Bacillus varies among strains and depends on the exogenous environmental conditions to which animals are exposed (Konieczka et al., 2022). In the present study, we hypothesized that dietary B. licheniformis BCG might alter the ileal and cecal microbiota, and contribute to broiler digestibility and gut health. To this end, we aimed to explore the protective roles of B. licheniformis BCG involving improved nutrient digestion and absorption, a strengthened intestinal barrier, and decreased inflammation, and to understand their relationships with the gut microbiota shifted by B. licheniformis BCG. Our data provided a theoretical basis for application of B. licheniformis BCG in the production of healthy broilers.

Materials and methods
The experimental animal protocol for this study was conducted in accordance with the recommendations of "Guidelines on Welfare and Ethical Review for Laboratory Animals" (GB/T 35892-2018), and approved by the Institutional Animal Care and Use Committee of the Institute of Feed Research of Chinese Academy of Agricultural Sciences (FRI-CAAS20210827).

Animals and experimental design
A total of 240 1-day-old male Arbor Acre broilers (body weight, 42.62 ± 0.82 g) were randomly allocated to three groups. Each group consisted of eight replicates (pens) with 10 broilers per pen. Two phase non-medicated basal diets in mashed form were formulated based on the nutrient requirements of the National Research Council (1994); ( Table 1). The three groups included basal diet (CT, n = 8), Frontiers in Microbiology 02 frontiersin.org and basal diet with a dose of 1.0 × 10 8 CFU/kg (BCG1, n = 8) and 1.0 × 10 9 CFU/kg (BCG2, n = 8) B. licheniformis BCG, respectively (Wang et al., 2017;Kan et al., 2021;Xu et al., 2021). All broilers were feed in wire-floored cages in a one-level battery on their respective diets. The study lasted 42 days, during which time broilers had ad libitum to feed and fresh water. Broilers were housed in an environmentally controlled room and temperature was gradually reduced from 35 • C on day 1 to 26 • C at day 21 and then kept roughly constant. A 20 h light-4 h dark cycle was carried out throughout the experimental period. B. licheniformis BCG was isolated from humus soil in the northeast forest area and preserved in the Key Laboratory of Feed Biotechnology of Ministry of Agriculture and Rural Affairs. It presents great biological characteristics in carbohydrate metabolism enzymes and stress tolerance through the whole genome sequencing and in vitro evaluation. B. licheniformis BCG with viable count = 1.08 × 10 10 CFU/g was used and mixed in the basal diet, which was prepared in bacterial mashed form after processed in activation, culture, centrifugation, freeze-drying, and grinding.

Sample collection
At 42 days of age, after fasting overnight, one broiler representing the average weight from each replicate was selected and humanely slaughtered. Jejunum, ileum and cecum segments were divided and fresh ileal and cecal contents were collected for α-amylase and microbiota analysis. Jejunal and ileal mucosa were scraped by autoclaved blade after precooled saline flush for maltase, sucrase and gene expression analysis. All samples were obtained as described previously , and immediately frozen in liquid nitrogen and stored at −80 • C.

Biochemical analysis
Appropriately 100 mg frozen mucosa and chyme of jejunum and ileum, respectively, were taken and mixed with 1 mL cold buffer (pH7.4), containing 10 mM Tris-HCl, 0.1 EDTA-Na 2 and 0.8% (w/v) NaCl, and homogenized using an Ultra-Turrax homogenizer for 30 s. Homogenates were centrifuged at 3,000 × g for 15 min at 4 • C and supernatants transferred to new tubes for protein assay and other measurements. The activities of α-amylase, maltase and sucrase were measured by colorimetry using the commercial kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's protocols. Maltase and sucrase were normalized to tissue protein concentrations, which were measured with a bicinchoninic acid commercial kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's instructions.
Paired-end reads were generated and merged using FLASH software (V1.2.7) 1 . Operational taxonomic units with 97% identity were gathered using Uparse 2 (ver. 7.1). Taxonomic annotations were performed using the Mothur algorithm (70% confidence) in the Silva database 3 . Alpha-diversity was analyzed using Observed_species, Chao1, and Shannon indices. Beta-diversity was visualized using principal coordinate analysis (PCoA) plots based on weighted UniFrac distance. Bacterial biomarkers between groups were displayed using the linear discriminant analysis effect size (LEfSe, linear discriminant analysis (LDA) > 3.5).

Statistical analysis
Statistical analyses were performed using one-way analysis of variance in SAS 9.4 (SAS Institute, Inc., Cary, NC, USA). Each broiler served as statistical unit. Differences between treatment means for enzyme activity and gene expression were evaluated using Duncan's multiple-range tests. Wilcox test was used for alpha-diversity index. LEfSe, t-test, and Metastat analyses were used to test for significant differences between microbiota relative abundance. Results were represented as mean with standard error of mean (SEM) in the tables and the mean with standard error (SE) in the figures, while P < 0.05 ( * ) and P < 0.01 ( * * ) values were considered statistically and extremely significant, respectively. Bar charts were drafted in Graphpad Prism 7.0 software (GraphPad Software Inc., La Jolla, CA, USA).

Effects of B. licheniformis BCG on jejunal and ileal enzyme activity in broilers
The BCG2 group showed significantly greater jejunal and ileal maltase and sucrase activity, and the BCG1 group showed significantly greater ileal sucrase activity, than the CT group (P < 0.05, Table 2). No differences in these two parameters were observed between the BCG1 and BCG2 groups (P > 0.05). Dietary B. licheniformis BCG resulted in significantly greater α-amylase activity than that in the CT group (P < 0.05), and this activity was higher in the BCG2 group than the BCG1 group (P < 0.05).

Effects of B. licheniformis BCG on nutrient transporter gene mRNA levels in the ileum
FABP-1 and FATP-1 relative mRNA levels in the BCG2 group were significantly higher than those in the CT and BCG1 groups (P < 0.05, Figure 1). The transcript abundance of GLUT-2 and LAT-1 was greater in the BCG2 group than the CT group (P < 0.05), and no significant difference was found between the BCG1 group and the other groups (P > 0.05).

Effects of B. licheniformis BCG on tissue morphology and the mRNA expression of tight junction and inflammatory molecules in the ileum
Histological examination of the ileum indicated that the villi and epithelium in the CT group, as compared with the BCG1 and BCG2 groups, showed damage; however, no clear infiltration of inflammatory cells was observed among groups (Figure 2A). Dietary B. licheniformis BCG significantly up-regulated ileal occludin mRNA levels, which were higher in the BCG2 group than the CT group (P < 0.05, Figure 2B). The transcript abundance of IL-8 and TLR-4 in the BCG1 and BCG2 groups was lower than that in the CT group (P < 0.05, Figure 2C), whereas no significant difference was observed between the BCG1 and BCG2 groups. No significant difference in claudin-1, ZO-1, mucin-2, IL-1β, TNF-α, NF-κB, IL-10, and TGF-β transcript abundance was observed between groups (P > 0.05).  Effects of Bacillus licheniformis BCG on nutrient transporter gene mRNA levels in the ileum. Values are mean ± SE (n = 6). a,b Bars with different letters within the same index indicate a significant difference between groups (P < 0.05).

Effects of B. licheniformis BCG on microbiota diversity in the ileum and cecum
A total of 3,917,657 high quality sequencing reads were generated from 47 broiler gut samples, with an average of 69,233 effective sequences/sample. Alpha diversity analyses indicated varying community richness and diversity among groups in the ileal but not the cecal bacterial communities. Both the BCG1 and BCG2 groups showed significantly lower ileal Observed_species, Chao1, and Shannon indexes than the CT group (P < 0.05, Figures 3A-C). Differences in the microbial structure among groups and niches were evaluated with PCoA analysis based on weighted UniFrac distance. Microbial communities were well separated between the ileal microbiota and counterparts colonizing the cecum, and between groups in the ileum (Figures 3D-F). ANOSIMs also confirmed the structural dissimilarity between the ileum and cecum, and between the CT group and the two BCG groups in the ileum or cecum (R > 0, P < 0.05, Table 3).

Effects of B. licheniformis BCG on ileal and cecal bacterial structures
In the ileum, Proteobacteria, Firmicutes, Bacteroidota, and Campylobacterota were the dominant bacterial phyla, with a relative abundance accounting for > 95% of the total ileal bacterial communities ( Figure 4A). In contrast to the CT-ileal (CT-I) group, the BCG1-I and BCG2-I groups showed an increase in the relative abundance of Proteobacteria and Firmicutes, from 11.79 to 17.44 (16.81%), and from 64.11 to 69.37 (73.94%), respectively (P > 0.05). However, the BCG1-I and BCG2-I groups showed a decrease in the relative abundance of Bacteroidota and Campylobacterota, from 10.25 to 4.19 (4.62%), and from 9.08 to 6.42 (1.98%), respectively (P > 0.05). At the family level, in contrast to the CT group, the BCG1 and BCG2 groups showed an increase in the relative abundance of Peptostreptococcaceae, Sphingomonadaceae, and Lactobacillaceae, from 17.6 to 26.67 (26.92%), from 0.04 to 6.89 (5.39%), and from 12.18 to 23.17 (22.97%), respectively (P > 0.05). The relative abundance of Campylobacteraceae in the BCG1-I and BCG2-I groups was, respectively, 0.68 and 1.44%, and lower than the 7.31% in the CT-I group (P > 0.05, Figure 4B). Diets with BCG1 and BCG2 significantly decreased several low abundance bacteria at the family level (P < 0.05, Figures 4C, D). For the 35 most dominant ileal genera, the BCG1 and BCG2 diet groups showed significantly lower relative abundance of Clostridiales bacterium CHKCI001, Enterococcus, Clostridia_vadinBB60_group, Faecalibacterium, Phascolarctobacterium, Barnesiella, Alistipes, and Ruminococcaceae UCG-005 than that in the CT group (P < 0.05, Figure 4E). The BCG1 diet group showed significantly lower relative abundance of Lactobacillus, Vibrio, Pseudomonas, Bacteroides, Streptococcus, Staphylococcus, and Bacillus than that in the CT group (P < 0.05). The relative abundance of Helicobacter in the BCG2-I group was significantly lower than that in the CT-I and BCG1-I groups (P < 0.05). In contrast to the CT-I group, the BCG1-I and BCG2-I groups showed an increase in the relative abundance of Sphingomonas from 0.03 to 6.34 (11.55%), whereas Campylobacter levels showed a decrease from 7.31 to 0.68 (1.44%) (P > 0.05).
In the cecum, Bacteroidota, Firmicutes, Proteobacteria, Campylobacterota, and Fusobacteriota were the major bacterial phyla, and their relative abundance was seldom affected by BCG treatments (P > 0.05, Figure 5A). On the basis of T-test and LEfSe results, CT-C broilers had a higher relative abundance of f_Erysipelotrichaceae, f_Peptostreptococcaceae, and g_Romboutsia than that in the BCG1-C and BCG2-C groups, and f_Erysipelotrichaceae and f_Peptostreptococcaceae levels in the BCG2-C group were significantly higher than those in the BCG1-C group (P < 0.05, Figures 5B-F). f-Lactobacillaceae and g_Lactobacillus were the dominant bacteria in the BCG2-C group, in contrast to the BCG1-C and CT-C groups, whereas f_Enterococcaceae, f_Campylobacteraceae, f_Tannerellaceae, g_Escherichia-Shigella, g_Campylobacter, and g_Parabacteroides were the dominant bacteria in the BCG1-C group, in contrast to the CT-C and BCG2-C groups. The f_Campylobacteraceae level in the BCG2-C group was significantly lower than that in the BCG1-C group (P < 0.05).

Correlations between ileal microbiota and enzyme activity and nutrient transporters or inflammatory and barrier parameters
A Spearman's correlation analysis was performed to explore the relationships of predominant ileal phyla, families, and genera with the nutrient digestion and absorption, or inflammatory and barrier parameters (Figure 6). Proteobacteria, f_Sphingomonadaceae, and Sphingomonas were significantly positively correlated with FATP-1 expression, whereas Alistipes and Barnesiella were significantly negatively correlated with α-amylase activity (P < 0.05, Figure 6A). Bacteroides and Limosilactobacillus showed a significant positive correlation with SGLT-1 expression and α-amylase activity (P < 0.05). Helicobacter was significantly positively correlated with maltase activity but negatively correlated with SGLT-1 expression (P < 0.05). Firmicutes, f_Lactobacillaceae, Lactobacillus, Fusobacterium, and Limosilactobacillus showed significant positive correlations with ZO-1 expression. f_Lactobacillaceae and Lactobacillus were significantly positively correlated with occludin and mucin-2 expression. f_Lactobacillaceae showed a significantly positive correlation with claudin-1 expression (P < 0.05, Figure 6B). Bacteroidota, Bacteroides, and Phascolarctobacterium were significantly correlated with TLR 4 and IL-10 expression, and Ruminococcaceae UCG-005 was significantly correlated with TNF-α, NF-κB, and IL-8 expression (P < 0.05). Proteobacteria, f_Sphingomonadaceae, and Sphingomonas showed significant negative correlations with IL-1β and NF-κB expression (P < 0.05). Effects of Bacillus licheniformis BCG on alpha-and beta-diversity of the microbiota inhabiting the ileum and cecum in broilers. (A-C) Bacterial richness and diversity, estimated with Observed_species, Chao1 value, and Shannon index. (D-F) Principal coordinate analysis based on weighted UniFrac distances, showing separation in the microbiota between the ileum and cecum, and between groups in the ileum or cecum. Values are mean ± SE (n = 8; BCG1-C, n=7). a,b Boxes with different letters within the same index indicate a significant difference between groups (P < 0.05).

Discussion
Bacillus spp. bacteria serve as a potential alternative to antibiotic growth promoters in livestock production, owing to their stress resistance and probiotic characteristics. Previous studies have shown that dietary supplementation with B. licheniformis significantly promoted broiler growth by increasing body weight and decreasing the feed to weight ratio (Liu et al., 2012;Chen and Yu, 2020;Xu et al., 2021). In addition, several studies have indicated that B. licheniformis administration promoted broiler growth under heat stress and Clostridium perfringens challenge conditions (Song et al., 2014;Zhou et al., 2016;Musa et al., 2019). These findings were consistent with those in our previous study indicating that diets with B. licheniformis at a dose of 1.0 × 10 9 CFU/kg significantly increases broiler body weight and average daily gain (Wang et al., 2022). Thus, B. licheniformis improved broiler growth performance under both normal and stress conditions. The mechanism underlying the improvements in nutrient digestion and host health were likely to involve the formation of beneficial metabolites such as extracellular digestive enzymes, organic acids, and antibacterial peptides (Kim et al., 2004;Knap et al., 2010;Xu et al., 2021). Thus, the activity of αamylase, maltase, and sucrase in the jejunum and ileum was further determined, because these enzymes participated in the digestion of nutrients.
In poultry, feed starch is generally degraded by α-amylase into smaller molecular oligomers after initial hydrolyzation by the microbiota, and is finally hydrolyzed into maltose and maltotriose in the small intestine (Dhital et al., 2017). Subsequently, maltose and maltotriose are hydrolyzed into glucose by sucrase, and maltase resides at the surfaces of the intestinal villi (Zhou et al., 2021). B. licheniformis secretes extracellular enzymes such as α-amylase and proteases (Kaewtapee et al., 2017). In the present study, dietary B. licheniformis BCG, compared with the CT diet, significantly increased jejunal and ileal α-amylase, maltase, and sucrase activity, in partial agreement with findings reported by Yang et al. (2021). The α-amylase activity is an important rate-determining factor in starch digestion, because endogenous α-amylase activity is low in broilers (Zhou et al., 2021). The increased α-amylase activity may increase the nutrient digestibility coefficients of starch and organic matter and consequently increase feed metabolizable energy, thus contributing to broiler growth performance (Jiang et al., 2008;Kaczmarek et al., 2014). In addition, nutrient transporters are crucial for nutrient absorption at the brush border membrane in the small intestine. Both FABP-1 and FATP-1 are important for intestinal absorption of lipids and fatty acids, particularly in longchain fatty acid metabolism and intracellular transportation (Coe et al., 1999;Richieri et al., 1999). GLUT-2 is an Na + -independent transporter responsible for the basolateral exit of glucose from the intestinal mucosa into the portal circulation (Wang et al., 2020). Intestinal LAT-1 participates in branched-chain and aromatic amino acid transport in an Na + -independent manner (Broer, 2008). As previously reported, probiotic supplementation enhances expression of some types of nutrient transporters in the small intestines in animals (Faseleh Jahromi et al., 2016;Duan et al., 2018). In agreement with previous findings, the present study indicated that B. licheniformis BCG administration significantly increased FABP-1, FATP-1, GLUT-2, and LAT-1 expression, to a greater extent in the BCG2 group than the BCG1 group. Thus, dietary supplementation with B. licheniformis BCG may aid in starch digestion and the absorption of glucose, amino acids, and fatty acids, thereby resulting in higher growth performance of broilers.
To determine whether B. licheniformis BCG contributed to intestinal health, we assessed the gene expression of TLR4 signaling Effects of Bacillus licheniformis BCG on bacterial composition in the ileum. (A,B) Distribution of ileal bacteria at the phylum and family levels. (C,D) The t-tests were used to assess significant differences at the family level; P < 0.05 indicates a significant difference. (E) Statistical analysis of differences in the relative abundance of the top 35 genera; Metastat was used to test for significant differences; light pink diamonds indicate P < 0.05, and dark pink diamonds indicate P < 0.01 between groups (n = 8).
pathway related molecules, tight junction proteins, and mucin-2. TLR4, as a pathogen-associated molecular pattern, mediates downstream inflammatory signals through the linker protein MyD88, thereby activating NF-kB, which then translocates into the nucleus and elicits pro-inflammatory cytokine secretion and cellular responses of immune-associated cells (Huebener and Schwabe, 2013;Kan et al., 2021). B. licheniformis has been demonstrated to decrease secretion of the pro-inflammatory cytokines IL-8 and IL-6 in vivo (Deng et al., 2012;Roselli et al., 2017). Pretreatment with B. licheniformis has been found to decrease the serum TNF-α and IL-1β levels in an acetaminophen-induced acute liver injury rat model (Neag et al., 2020). Similarly, B. licheniformis also decreases the inflammatory response in an LPS-induced acute inflammation rat model (Deng et al., 2017). Moreover, B. licheniformis treatment markedly counteracts the increase in IL-6, IL-8 and TNF-α inflammatory gene expression induced by enterotoxigenic Escherichia coli F4 in vitro. The immunity homeostasis of HT-29 cells is improved by treatment with B. licheniformis MCC 2514, on the basis of downregulation of IL-1α, IL-6, IL-8, IL-12, and TNF-α, and upregulation of IL-4, IL-10, TGF-2, and TGF-3 (Rohith and Halami, 2021a). These results are consistent with our observation that the ileal TLR 4 and IL-8 expression in the BCG1 and BCG2 groups was significantly lower than that in the CT group. The underlying mechanism was associated with B. licheniformis bacteriostasis. As previously reported, B. licheniformis inhibited the growth of pathogenic bacteria such as Kocuria rhizophila and enterotoxigenic Escherichia coli by adhering to the intestines (Rohith and Halami, 2021b;Li et al., 2022a).
Another key mechanism through which B. licheniformis attenuates inflammation might involve intestinal barrier improvement.
The integrated intestinal barrier plays an important role in maintaining epithelial cell function, because the epithelium is permeable to feed-associated antigens, and luminal pathogens and their toxins (Awad et al., 2017). The transmembrane proteins The t-tests were used to assess significant differences at the family level; P < 0.05 indicates a significant difference. (E) LEfSe analysis of differences in taxa enrichment in microbial communities between groups; bacterial taxa with a logarithmic LDA score > 3.5 were biomarker taxa. (F) Cladogram showing bacteria with significant differences between groups (n = 8; BCG1-C, n=7). claudins, occludins, and zonula occludens involved in tight junctions are responsible for connecting epithelial cells and regulating paracellular and intracellular permeability (Aijaz et al., 2006). The regular permeability and integrity of tight junctions are generally negatively affected by stress factors, such as heat stress and pathogen invasion (Song et al., 2014;Musa et al., 2019;Li et al., 2022a). Under Clostridium perfringens challenge, dietary B. licheniformis significantly increases the gene expression of claudin-1 and ZO-1 in the duodenum in broilers at post challenge days 7 and 21 (Musa et al., 2019). In laying hens, B. licheniformis administration enhances the intestinal mechanical barrier by upregulating the gene expression of mucin-2 and tight junction proteins (Wang et al., 2017). In in vitro experiments, B. licheniformis PF9 application has been found to reverse the decrease in ZO-1 and occludin expression in the cell membrane after challenge with enterotoxigenic Escherichia coli . In agreement with previous studies, the present study indicated that a diet with a high dose of B. licheniformis BCG resulted in significantly higher occludin expression than that in the CT group. Thus, B. licheniformis BCG alleviates ileal inflammation partly by enhancing the physical barrier. The improved barrier function resulting from B. licheniformis BCG might be attributable to its biologically active substances including bacteriocin and antibacterial peptides, which suppress pathogenic bacterial colonization and balance the intestinal microbiota Kan et al., 2021).
The gut microbiota provides a broad range of functions for hosts, for example, the digestion of complex dietary nutrients, defense against pathogens, enhancement of the intestinal barrier, and promotion of immune maturation (Koh et al., 2016). Owing to differences in histology and function, the broiler intestinal tract is generally divided into the fore-and hindgut. Previous studies in broilers have focused primarily on the hindgut microbiota and their functions, whereas studies on the microbiota in the foregut and their interactions with the host have been limited (Oakley et al., 2014;Huang et al., 2018). In the current study, we analyzed ileal and cecal microbial diversity and composition, and their effects under B. licheniformis BCG treatment. The PCoA revealed a clear separation between the ileum and cecum compartments, thus indicating a large difference between them, in agreement with findings from a previous study (van der Wielen et al., 2002). In addition, piglets fed Bacillus species probiotics show diminished microbial richness in feces (Kaewtapee et al., 2017;Poulsen et al., 2018). These findings are partially consistent with those in the current (A) Correlations between the ileal microbiota and enzyme activity and nutrient transporters. (B) Correlations between the ileal microbiota and inflammatory and barrier parameters. Red represents a positive correlation, and blue represents a negative correlation. *P < 0.05 and **P < 0.01 indicate significant and extremely significant correlations. study indicating diminished bacterial richness and diversity in the ileum but not cecum in broilers in response to B. licheniformis BCG. Our findings indicated that dietary supplementation with B. licheniformis BCG shifted the ileal bacterial community structure in broilers. Theoretically, gut bacterial diversification is a gradual process that increases with age, and high diversity is considered a sign of gut bacterial maturity (Micah et al., 2007). Premature formation of an adult-type bacterial community negatively affects host gut immunity (Nylund et al., 2013). Dietary supplementation with B. licheniformis BCG significantly decreased ileal inflammation in the present study, a result partially attributed to the decrease in bacterial diversity.
Bacterial composition is also closely associated with host physiology, including digestion, metabolism, and immunity. In the present study, considerable variations in Proteobacteria, Firmicutes, Campylobacterota, and Bacteroidota were found between the ileum and cecum. Oxygen-sensitive Bacteroidota markedly increased from the ileum to the cecum, because it is adapted to a low oxygen environment. However, oxygen-tolerant Proteobacteria and Campylobacterota, such as Enterococcaceae and Campylobacteraceae, decreased from the ileum to the cecum. Our results are consistent with previous observations that spatial shifts in bacterial composition depend on microenvironment change (Huang et al., 2018;Li et al., 2022b). Firmicutes include Peptostreptococcaceae, Clostridiaceae, and Lactobacillaceae, which are positively associated with energy intake and have been found to provide an additional 628 kJ of energy when their abundance increases by 20% (Hildebrandt et al., 2009;Jumpertz et al., 2011). This causal relationship has also been found in piglet models (Mach et al., 2015;Li et al., 2022a). Firmicutes fermentation is a more suitable energy source than Bacteroidota because it produces more short chain fatty acids, thus enhancing efficient heat absorption (Krajmalnik-Brown et al., 2012;Pan et al., 2022). We observed that a diet with B. licheniformis increased ileal Firmicutes levels but decreased Bacteroidota levels, a finding partially explained by our previous results indicating that B. licheniformis administration promoted broiler growth (Wang et al., 2022). In recent studies, Sphingomonas involved in the f _Sphingomonadaceae has been observed as an abundant bacterium in chicken intestines (Chen et al., 2018;Zhang et al., 2022b). The abundance of Sphingomonas is significantly positively correlated with fat catabolism in the liver, serum, and muscle Zhang et al., 2022b). In addition, Sphingomonas participates in lipid metabolism across the entire process of chicken embryonic development (Akinyemi et al., 2020). Diets with B. licheniformis resulted in significantly greater ileal f _Sphingomonadaceae and Sphingomonas levels than those in the CT group; these levels were significantly positively correlated with FATP-1 expression but significantly negatively correlated with IL-1β and NF-κB expression. These findings suggested that the shifts in the prevalence of f _Sphingomonadaceae and Sphingomonas after B. licheniformis BCG treatment contributed to the intestinal absorption and oxidative decomposition of lipids and fatty acids, and alleviation of intestinal inflammation.
On the basis of the analysis of the abundance of bacteria in the present study, Lactobacillus, Bacteroides, Alistipes, Escherichia-Shigella, Desulfovibrio, Streptococcus, Ruminococcaceae_UCG-005, Fusobacterium, and Campylobacter predominated in broilers. In general, the presence of Lactobacillus is considered beneficial for intestinal health and animal health, owing to their immunomodulation, pathogen inhibition and bacteriocin production ability (Kaewtapee et al., 2017;Zhao et al., 2022b). In the current study, broilers in the BCG2 group had higher Lactobacillus levels in both the ileum and cecum than those in the CT or BCG1 group, in agreement with findings from previous studies (Hung et al., 2019;Zhao et al., 2022b). An increase in Lactobacillus might have resulted from B. licheniformis BCG supplementation, which creates a suitable environment for the colonization of Lactobacillus anaerobic bacteria by consumption of oxygen in the gut (Fazelnia et al., 2021). In previous studies, Lactobacillus administration has been found to contribute to nutrient absorption by significantly increasing the expression of sugar transporter genes, including GLUT-2, GLUT-5, SGLT1, and SGLT4 (Faseleh Jahromi et al., 2016). Lactobacillus enhances occludin levels and suppresses Escherichia coli invasion in intestinal epithelial cells (Resta-Lenert and Barrett, 2003). Moreover, Lactobacillus reverses LPS-induced disruption in tight junction proteins, such as occludin, claudin-1, and ZO-1 (Zhou et al., 2014;Roselli et al., 2017). These findings were consistent with our observations indicating that f_Lactobacillaceae and Lactobacillus were significantly positively correlated with the expression of ZO-1, claudin-1, mucin-2, and occludin, and the occludin expression in the BCG2 group was significantly higher than that in the CT group. In addition, Limosilactobacillus spp. and Lactobacillus spp. have been found to significantly decrease the abundance of Helicobacter, one of the most common pathogens globally associated with gastritis and cancer, by 90% and 83%, respectively, in an infected mouse model (Zhao et al., 2022a). Moreover, Limosilactobacillus spp. administration significantly increased the prevalence of Lactobacillus spp. but decreased the abundance of Desulfovibrio. These relationships among Limosilactobacillus, Lactobacillus, Helicobacter, and Desulfovibrio were consistent with the microbial structure shift induced by B. licheniformis BCG administration. Limosilactobacillus showed a significantly positive correlation with ZO-1 expression and α-amylase activity. These findings indicated that the increase in f_Lactobacillaceae, Lactobacillus, and Limosilactobacillus levels attributed to B. licheniformis BCG treatment improved intestinal barrier function through interaction with epithelial cells in broilers.
Desulfovibrionaceae is a family of opportunistic pathogens such as Desulfovibrio, which is a major sulfate-reducing bacterium that is ubiquitous in human intestines (Ichiishi et al., 2010). These sulfate-reducing bacteria destroy intestinal epithelial cells through generating large amounts of hydrogen sulfide (Zhang-Sun et al., 2015). In agreement with the current findings, B. licheniformis DSM5749 administration sustains intestinal health in laying hens by decreasing the Desulfovibrio level (Pan et al., 2022). Alistipes has been isolated from the appendicular, abdominal, perirectal, and brain abscesses, thus indicating a potential opportunistic pathogenic role in humans (Parker et al., 2020). A high relative abundance of Alistipes is strongly associated with gut disorders and host diseases, such as liver fibrosis and non-alcoholic steatohepatitis, which is inhibited by Bacillus licheniformis H2 administration (Parker et al., 2020;Zhao et al., 2022b). Campylobacter is well known as a major cause of acute bacterial enteritis in humans. Poultry is considered a major reservoir of Campylobacter and generally colonized by Campylobacter at the age of 2 weeks; therefore, Campylobacter may be as a principal vehicle of transmission to humans (Evans and Sayers, 2000;Lamb-Rosteski et al., 2008). Furthermore, Vibrio, a potential pathogen, should be harmful for host health (Richards et al., 2016). Streptococcus and Escherichia coli are pathogens that metabolize proteins in the small intestine (Ma et al., 2017). The above bacteria decreased after B. licheniformis BCG treatment, particularly in the BCG2 group. Dietary supplementation with B. licheniformis BCG significantly decreased the expression of pro-inflammatory molecules, and thus might decrease the risk of bacterial enteritis in broilers and its associated postinfectious sequelae in humans. However, in the present study, the relative abundance of Bacteroides, Prevotellaceae_UCG-003, and Prevotellaceae_NK3B31_group and Ruminococcaceae_UCG-005 were lower in the BCG1 and BCG2 groups than the CT group. As reported, Prevotella can metabolize plant cell walls and produce short chain fatty acids that benefit intestinal immunity homeostasis (Ramayo-Caldas et al., 2016). Prevotella and Prevotella_9 show significantly negative correlations with levels of the pro-inflammatory cytokines IL-6 and IL-12 in the jejunum . Bacteroides shows a significantly negative correlation with serum IL-6 but a positive correlation with IL-4 . The abundance of Ruminococcaceae, beneficial bacteria that are more commonly found in healthy people, is correlated with the production of short chain fatty acids (Koh et al., 2016;Feng et al., 2022). These reports are contrary to those from a present study reporting that bacteria whose abundance decreased after B. licheniformis BCG administration are positively correlated with the pro-inflammatory molecules IL-8, TLR-4, NF-κB, or TNF-α expression; therefore, these findings must be verified in further study.

Conclusion
The present study indicated that dietary B. licheniformis BCG significantly increased jejunal and ileal α-amylase, maltase, and sucrase activity; up-regulated ileal occludin mRNA levels; and decreased the transcript abundance of IL-8 and TLR-4 in the ileum in broilers. Diets with B. licheniformis BCG significantly decreased bacterial community richness and diversity in the ileum but not the cecum. Dietary B. licheniformis BCG shaped the ileac microbiota; increased the prevalence of f_Sphingomonadaceae, Sphingomonas, and Limosilactobacillus; contributed to nutrient digestion and absorption; increased the prevalence of f_Lactobacillaceae, Lactobacillus, and Limosilactobacillus; and enhanced intestinal barrier function. In addition, dietary B. licheniformis BCG decreased microbial community diversity; decreased the abundance of Desulfovibrio, Alistipes Campylobacter, Vibrio, Streptococcus, and Escherichia coli-Shigella; and down-regulated expression of inflammatory molecules. Therefore, diets with B. licheniformis BCG contributed to broiler digestion and absorption of nutrients; enhanced the intestinal physical barrier; and decreased intestinal inflammation by decreasing microbial diversity and optimizing the microbiota structure. Our data provided a theoretical basis for B. licheniformis BCG application in broilers.

Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: NCBI -PRJNA648691.

Ethics statement
The animal study was reviewed and approved by the Institutional Animal Care and Use Committee of the Institute of Feed Research of Chinese Academy of Agricultural Sciences (FRI-CAAS20210827).

Author contributions
YH, JW, and PY designed the experiments. XX, DL, HZ, and YH conducted experiments and collected samples. JW, XX, and HC performed sample analysis. YH analyzed the data. YH, PY, and KM wrote and revised the manuscript. All authors contributed to the article and approved the submitted version.