SEBS colonized in ileal mucosa, optimized bacterial composition, improved Se level and activated TLR2 NF-kB1 signaling pathway to regulate β-defensin 1 expression

Both selenium (Se) and probiotic bacillus improve immunity. Beta defensin 1 (BD1), a component of intestinal mucosal immunity, could be up-regulated in dietary selenium enriched bacillus subtilis (SEBS) supplementation. SEBS was supplemented in the culture medium of mouse intestinal crypt cells and the diets of chicks to observe the effects of SEBS on BD1 in the intestine by colonization of bacillus subtilis, its recognition and signaling pathway, bacterial composition optimization, and biological functions of Se.


Background
The low temperature of living condition easily causes to cold stress to animals, led to weak immunity.
Hence, improve body immunity is crucial to prevent the pathogenic infection. Probiotics contain live microorganisms and spores and confer health bene ts to the host when administered in adequate amounts [1]. Certain strains of bacillus subtilis (B. subtilis) are used as probiotic bacteria, contributing a myriad of bene ts to immunity and metabolism [2]. Certain probiotic B. subtilis was administrated orally, survived and colonized in intestinal mucous membrane, optimized the composition of microbiota, which were positively stimulating the intestinal immunity and metabolism. These bene ts assist in overcoming infection stress and clearing pathogens [3]. The relationship between intestinal innate immunity and bacterial composition is close and direct. Beta defensins of the innate immune system have been found in species of bird, human, several domestic animals, and mice [4,5]. Beta defensin 1 (BD1) is expressed and distributed in several organs, such as the intestine, lungs, and bone marrow [6,7]. In the gut, BD1 is formed and secreted by intestinal epithelial cells [8]. BD1 was shown to be involved in the induction of other beta defensins, such as BD2, BD3, and BD4 [9]. The antibacterial ability of BD1 is the strongest among all β-defensins and plays a pivotal role in mucosal immunity, speci cally protecting against infection in the gut [10][11][12].
Probiotic B. subtilis colonized in mucous membranes strengthen the probiotic effect and produce microbial-associated molecular patterns (MAMPs), which could be recognized by toll-like receptors (TLRs) [13], resulting in a signal cascade to mobilize and activate the innate immune system [2]. TLRs are a series of membrane receptors that detect extra-cellular microbes and trigger signaling cascades for their eradication [14,15]. Leukocyte pro-in ammatory cytokines, such as IL-1β, IL-6, and TNF-α, are known as potent inducers and up-regulators of defensins [16,17]. They activate the NF-kB pathway, causing NF-kB entry to the nucleus, inducing the production of in ammatory cytokines, and provoking the rapid activation of innate and adaptive immune system factors, such as BD1. Bacillus was composed mainly by peptidoglycan and proteide, the core was spore [18]. However, there is still unknown that probiotic bacillus as a whole recognized by which TLRs. Selenium (Se) is an important trace element and its bene cial effects are well documented [19]. Suitable doses of Se supplementation can regulate metabolism and antioxidation, especially during bodily stress and Se de cit areas [20]. Since Se confers several bene ts, it is urgent to explore a cheap and convenient organic Se source for supplementation. Meanwhile, Se can help the immune cells to defense the infection caused by pathogen. Still the supplementation of Se could reduce the DNA copy number of pathogens to inhibit pathogenic infectious effects, alleviate the infected cell numbers and stress response [21,22].
Considering the bene cial effects of Se and B. subtilis on immunity and stress [23], it would be valuable to study Se and B. subtilis in combination to determine whether greater effects are induced when administered together. To test this hypothesis, and observe the potential mechanism, we produced Seenriched B. subtilis (SEBS), which combines the virtues of B. subtilis and those of organic Se. The intestinal crypt cells of mice were cultured and treated with SEBS to explore the effects and signaling pathways associated with BD1. Chinese Huainan Partridge chicken is a dual-purpose native breed from South China [2], which was used in our study to observe the regulation of BD1 mechanisms by SEBS in vivo in cold condition (Fig. 1). This study rst discovered that the bacteria of B. subtilis colonized in ileal mucous membrane was recognized by the TLR2 and presentation to MyD88, stimulating and provoking downstream signals of IRAK, TRAF6, IκB, IL-1β, and TNF-α. The signaling pathway is the classic TLR2-MyD88-NF-kB1 pathway. Meanwhile, the in ammatory response of intestinal crypt cells provoking by B. subtilis could be alleviated under the action of Se, while the whole immune status of intestine was still improved with combined use of Se and B. subtilis. The colonization of B. subtilis in ileal mucous membrane optimized the intestinal bacterial composition, which improved the anti-infection ability. This study could provide novel insights regarding combined use of probiotic bacteria and essential microelement Se.

Methods
Preparation and analyses of SEBS B. subtilis was isolated from the ileum of a healthy Chinese Huainan Partridge chicken by our research group, the institute of animal husbandry and veterinary medicine, Anhui Academy of Agricultural Sciences, and stored in the China General Microbiological Culture Collection Center (CGMCC), the strain number is CGMCC 14246. The 16S ribosomal DNA was sequenced and deposited at the National Center for Biotechnology Information (NCBI) of the United States of America (USA), the access number is KT260179. We cultured B. subtilis using a liquid beef extract-peptone medium. The fermentation of selenium-enriched B. subtilis was prepared with sodium selenite supplemented into the culture medium. The morphologic and structural properties of bacteria between B. subtilis and SEBS was monitored using a scanning electron microscope (SEM) and transmission electron microscopy (TEM). The bacteria of B. subtilis and SEBS was concentrated via centrifugation at 3, 000 rpm, and immersed in a 5% glutaraldehyde solution for 24 h [24]. Se concentration in the supernatant and precipitate of B. subtilis and SEBS fermented medium was calculated using atomic absorption spectrometry and the live bacteria was enumerated with yeast extract peptone dextrose medium after term serial dilution.

Chicks and management
For the chicken origin of B. subtilis and SEBS, homologous feeding of probiotic bacteria could play more bene cial roles than non-homologous. The strain of B. subtilis was isolated from Chinese Huainan Partridge chicken. Hence, Chinese Huainan Partridge chickens were chosen to carry out the animal study.
Animal experiments were approved and performed in accordance with the experimental guidelines of the Institutional Animal Care and Use Committee of China. The experimental protocols in this study, including animal husbandry and slaughter, were approved by the Institute of Animal Science and Welfare of Anhui Province (no. IASWAP2017120649). A total of 500 one-d-old Chinese Huainan Partridge chickens (average body weight, 40.15 g) were randomly allocated to ve groups with ve replicates of 20 each. For the duration of the starter diet, the pen space for each replicate was 2.3 m 2 . The room temperature was maintained at 33-35 ℃ using an electric heating tube in the rst week, then gradually declined to 21 ℃ at the end of the fourth week. At the end of the starter diet period, all chicks were weighed and transferred into larger pens (5.6 m 2 for each replicate); chicks in the previous replicate were also transferred and housed under the same conditions. In the fth week, the temperature was maintained at 18-20 ℃ using infrared warming lights. In weeks six to eight, the number of infrared warming lights was reduced gradually for a room temperature of 6.5 ℃. The chickens were allowed ad libitum access to water and feed throughout the experimental period. The normal immune procedure was implemented throughout the trial.

Feed for each group
Chickens in the control group were fed a basal diet and the four treatment groups were fed the following: basal diet with either inorganic sodium selenite (IS), B. subtilis (BS), selenium enriched B. subtilis (SEBS), and avomycin. Experimental diets were fed in two periods: starter (days 0-28) and nisher (days . The basal diet composition and nutrient analysis results, which did not contain any probiotics or antibiotics, are shown in Supplementary Table 1 (Table S1). All nutrients met or exceeded the nutrient requirements (National Research Council, 2012) [25]. Diet for chickens in the IS group, 1.12 g of sodium selenite (analytically pure) was diluted into 100 mL distilled water, which was blended with 5 kg of feed. Thereafter, the mixed mass feed was added to a blender containing 90 kg of mass feed. The blender was employed for 20 min to ensure uniform mixing of additives. The feed for the avomycin group was prepared using 4 g premixed food containing 10% avomycin, which was blended with 100 kg of feed, to reach a concentration of 4 mg/kg. For bacillus, 50 mL B. subtilis fermentation liquid was measured separately and rst blended with 5 kg of feed, and then with 95 kg of mass feed. The SEBS feed was prepared by blending 1000 mL of SEBS fermentation liquid with 100 kg of feed. After preparing the ve different feedstuffs, the population of B. subtilis was counted using the plate method with a yeast extract peptone dextrose medium. The concentration of Se in all feed types was also measured. The results are listed in Table S2.

Performance and sample collection
Chicks in every replicate of each treatment group were weighed on 0 d and 56 d. Daily feed consumption was accurately recorded. After 56 d, 2 average body weight of chickens in each replicate were selected (n = 5 2), fasted for 12 h, and then the tissue and blood were harvested under general halothane anesthesia.
All blood samples were collected in 5.0 mL sterile heparinized tubes. Blood was centrifuged at 3000 rpm for 10 min to collect the plasma for biochemical assays. Blood samples (1 mL) was removed to measure the Se concentration. Tissues of the jejunum and ileum were removed under aseptic conditions, stored in sterile plastic tubes in boxes packed with ice, and immediately transported to our laboratory for quanti cation of bacteria and other assays.

FISH assay
Two strains of bacteria, B. subtilis and endogenous salmonella [26], residing in the GIT were investigated using uorescence in situ hybridization (FISH). Ileal mucosal samples (0.3 g) were xed by immersion in 10% formaldehyde for 24 h. A 50 µL volume of homogeneous tissue liquid was transferred to poly-Llysine-coated slides, which were then air dried on a sterile benchtop for 3 h. The tissue was incubated with lysozyme at 32 ℃ for 10 min; the slide was then washed with distilled water and immersed in 70% ethanol for 2 min, followed by air drying. Probes with carboxytetramethylrhodamine and 6-carboxyuorescein were designed and conjugated with B. subtilis and salmonella, respectively. The length of the B. subtilis genetic probe was long to ensure speci c integration with samples. Probe sequences of B.
subtilis and salmonella are listed in Table S3. Probes were diluted to 60 nM, denatured at 95 ℃ for 5 min, and maintained at 4 ℃ before use. A probe (12 µL) was added to the tissue, followed by incubation at 46 ℃ for 12 h, and washed with phosphate buffer solution (pH 7.4). The tissue was stained with 4′,6diamidino-2-phenylindole for 5 min, then washed three times with distilled water for 5 min each. After drying, the slides were mounted with uoromount-GTM (Abcam, Cambridge, UK) and observed with a uorescence microscope (BX53; Olympus, Tokyo, Japan).
qRT-PCR for number of B.subtilis After fermentation in beef extract-peptone medium, a tenfold dilution series of B. subtilis KT260179 was planned. Colonies of B. subtilis were counted using the plate method under a microscope to obtain samples of 1 × 10 4 , 10 5 , and 10 6 . Total RNA in each dilution was extracted using the RNA Extraction Kit (Invitrogen, Carlsbad, CA, USA). Reverse transcription was performed using a GoScript Reverse System (Invitrogen). First-strand cDNA was synthesized by incubating a reaction mixture containing 11 µL RNA and 1 µL RNase-free dH 2 O at 70 ℃ for 3 min, followed by 0 ℃ for 5 min. A dNTP mixture (1 µL; 10 mmol/L), 4 µL GoScript 5X reaction buffer, 1 µL GoScript reverse transcriptase, 1.5 µL Mg 2+ (25 mM), and 0.5 µL RNase inhibitor were combined in a total volume of 20 µL and incubated at a 37 ℃ in a water bath. Primers were designed according to the 16S rRNA of B. subtilis KT260179, which are described in Table S4. Ampli cation was performed in a 20 µL mixture containing 10 µL of 2 × qPCR SYBR Premix Ex-Taq, 2 µL template cDNA, 0.5 µL each primer (10 µmol/L), and 7 µL PCR-grade water. The cycling protocol was as follows: 95 ℃ for 30 s, followed by 40 cycles of 95 ℃ for 5 s and 60 ℃ for 30 s, and one cycle for melting curve analysis, consisting of 95 ℃ for 60 s, 65 ℃ for 60 s, and 95 ℃ for 1 s. The ampli cation curve was generated based on the dilution of the standard curve of B. subtilis KT260179. The standard curve of B. subtilis KT260179 was described according to the results of quantitative realtime polymerase chain reaction (qRT-PCR).
Samples (0.2 g) of mucous membrane from the distal segment of the ileum were prepared to extract total RNA and qRT-PCR was carried out as described above to evaluate colonization of B. subtilis.

Plate method
The distal segment ileum samples were cut open and washed with sterilized physiological saline (pH 7.0). Samples of ileal mucous membrane were scratched from the ileum using a slide. The ileal mucous membrane and cecal lumen contents (0.4 g each) were prepared and 10-fold dilutions were prepared with sterilized saline. The bacterial composition of the ileal mucous membrane and cecal lumen contents in all groups was determined using the plate method. Salmonella Shigella agar was used for salmonella and MRS agar was used for lactic acid bacteria [3,27]. The assays were repeated three times. After the cecal content dilution had been heated (80 ℃) for 15 min to inactivate bacteria, bacillus remained. The total number of bacillus was counted using beef extract peptone medium. Salmonella and bacillus were cultured at 37 ℃ aerobic conditions, total lactic acid bacteria were cultured under anaerobic conditions. Gut bacterial 16 s rDNA sequence and analysis Samples (0.25 g) of the ileal mucous membrane were prepared (n = 50). Microbial DNA was extracted from these samples using the E.Z.N.A.® soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to the manufacturer's protocol. The nal DNA concentration and puri cation were determined using a NanoDrop 2000 UV-vis spectrophotometer (Thermo Scienti c, Waltham, MA, USA), and DNA quality was determined using 1% agarose gel electrophoresis. The V3-V4 hypervariable regions of the bacteria 16S rRNA gene were ampli ed with primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) using a thermocycler PCR system (GeneAmp 9700, Applied biosystems, Foster City, CA, USA). PCR was conducted as follows: Diversity metrics were calculated using the core-diversity plugin within QIIME2. Feature level alpha diversity indices and operational taxonomic units (OTUs) were used to estimate the microbial diversity within an individual sample. Beta diversity distance measurements were performed with weighted UniFrac to investigate the structural variation in the microbial communities across samples, and then visualized via principal coordinate analysis (PCoA). Co-occurrence analysis was performed by calculating Spearman's rank correlations between predominant bacterial species and the network plot. Additionally, the potential KEGG Ortholog (KO) functional pro les of microbial communities were predicted using PICRUSt.

Immunohistochemical assay
Intestinal sections, 2 mm in length from ileum of chick, were collected and immediately placed in 10% buffered formalin overnight at room temperature. Para n-embedded tissues were cut into 4 µm slices.
The slides containing para n-embedded ileum were rst depara nized using 100% xylene and rehydrated by decreasing ethanol percentages. Subsequently, the slides were then incubated in a bath (containing 10 mmol/L sodium citrate) at 90 ℃ for 60 min for antigen retrieval and were next transferred to a bath of 0.20% sodium borohydride to block the endogenous peroxides. After washing three times with PBS, the slides were blocked in blocking buffer containing 0.025% Triton X-100, 2% FBS, and 5% BSA at 25 ℃ for 60 min. Slides were further incubated in monoclonal-proliferating cell nuclear antigen antibody (1:500 dilution; BIOSS, Beijing, China) overnight at 4 ℃ in a humidi ed chamber. Slides were then incubated with the biotin-conjugated goat anti-chick immunoglobulins (1:250 dilution; BIOSS) at 25 ℃ for 90 min after the primary antibody was rinsed off. Thereafter, the slides were further treated with 1 mL of 3,3-diaminobenzidine tetrahydrochloride (ComWin Biotech, Beijing, China) to visualize the antigenic structures. Finally, slides were examined by an observer blinded to the treatment group. Computerized densitometric analyses of BD1 tissue expression in the images were performed using Image-Pro Plus version 5.0 software (Media Cybernetics Inc., Bethesda, MD, USA). Micro histological structures of intestines were visualized under a light microscope, and villous length and crypt depth of each crypt and villous of the tissue were enumerated using AxioVision version 4.6 software.
Elisa assay for BD1 level and activity of GPX Samples of ileal mucous membrane (n = 10) were prepared with homogenization and ten times dilution. The ileal mucous membrane samples were investigated for types of glutathione peroxidase (GPX), namely cytoplasmic (GPX1) and gastrointestinal (GPX4), and BD1 [28]. Whole blood samples from chicks (1 mL) were harvested for plasma GPX (GPX2) measurement. This experiment was performed using an ELISA kit (Invitrogen).
The Se concentration assay A ZEEnit 700 P atomic absorption spectrometer (Analytik Jena, Germany) was employed to determine the selenium levels in samples of whole blood, heart, and liver. All measurements were performed using hydride generation atomic absorption spectrometry (HG-AAS) [28]. Samples of whole blood (1 mL), liver (0.3 g) of chicks, and the supernatant and precipitate of fermented medium (3 mL) of B. subtilis and SEBS were placed in beakers and digested by adding 10 mL of a nitric acid-perchloric acid (HNO 3 -HClO 4 ) mixture. The cooled mixture was made up to 10 mL with 5% hydrochloric acid solution. Eight replicates were used for each group.

Cell Studies
Preparation of mouse primary intestinal crypt cells Mice are frequently used as an animal model to study mechanisms of substances, such as microorganisms and drugs. Mouse crypt cells, other than chicken, were prepared to unveil the target molecular signal (TLRs-NF-kB-BD1) via SEBS supplementation in vivo, which was more suitably adapted to other animals in clinical applications.
C57BL/6 mice (4 weeks old, speci c-pathogen-free) were purchased from Anhui Provincial Hospital Experimental Teaching Center. The mice were sacri ced to obtain the ileum and placed into a sterilized glass plate (9 cm diameter) containing PBS (0 ℃). Fat and connective tissue were removed from the ileum. The intestine was opened longitudinally, and contents were washed with PBS. This process was repeated until the supernatant appeared clear. The ileum was cut, washed with PBS twice, and the supernatant was removed. Small pieces of intestine were shaking incubated with 100 U/mL collagenase (Sigma, USA) buffer (20 µg/mL neutral protease , 32 mM HEPES, 127.5 mM NaCl, 3.15 mM KCl, 0.7 mM Na 2 HPO 4 , and 3 mM CaCl 2 , pH 7.60, 37 ℃) for 90 min. The cell suspension was ltered through mesh (grid size about 150 and 75 µm) and further diluted with serum-free medium. After sedimentation for 10 min at 4 ℃, the supernatant was discarded. The ileal crypt cells were centrifuged (3, 000 rpm), then resuspended, and washed twice with the same serum-free medium at 4 ℃. Lastly, the crypt cells were resuspended in DMEM/Ham F12 medium (Sigma, USA), containing 2% fetal bovine serum (FBS), 24 mM Thereafter, all cells from different groups in step one and two were collected for RNA extraction. Cytokines of MyD88, IRAK, TRAF6, IκB, BD1, IL-1β, and TNF-α were measured using qRT-PCR.
qRT-PCR for cytokine expression Quantitative real-time PCRs were carried out using SuperReal PreMix Plus (SYBR Green, FP205) on an ABI 7900HT Fast Real-time PCR System. Relative expression levels of target genes were quantitatively normalized against the expression of GAPDH using the ΔΔCT method. First-strand cDNA was synthesized by incubating a reaction mixture containing 11 µL RNA and 1 µL RNase-free dH 2 O at 70 ℃ for 3 min, followed by 0 ℃ for 5 min. The cDNAs were also used for PCR, and the PCR products were detected on 2% agarose gel. A dNTP mixture (1 µL; 10 mmol/l), 4 µL GoScript 5X reaction buffer, 1 µL GoScript reverse transcriptase, 1.5 µL Mg 2+ (25 mM), and 0.5 µL RNase inhibitor were combined in a total volume of 20 µL and incubated at a 37 ℃ in a water bath. Primers for MyD88, IRAK, TRAF6, IκB, BD1, IL-1β, and TNF-α were designed according to the mouse RNA genes submitted to NCBI. All PCR primers used in this study are described in Table S4.
Chicken intestinal mucosa cell collection and qRT-PCR Chicken mucosal tissues, collected from the distal segment of the ileum, were washed with ice-cold PBS to remove intestinal contents, and longitudinally cut into small specimens. Crypt cells were isolated using a PBS buffer containing 1 mM EDTA, 1 mM dithiothreitol, and 5% fetal bovine serum, with shaking at 37 ℃ for 10 min. This supernatant was ltered and combined with RPMI 1640 and centrifuged at 1,500 rpm to isolate crypt cells.
Samples of crypt cells were prepared to extract total RNA to evaluate the level of the target TLRs-BD1 post-SEBS supplementation in vivo. Relative expression levels of target genes were quantitatively normalized against the expression of GAPDH using the ΔΔCT method. Primers for TLR2, IL-1β, IL-6, TNFα, nuclear factor kappa B1 (NF-kB1), BD1, and IFN-γ were designed according to the chicken RNA genes submitted to NCBI. All PCR primers used in this study are described in Table S4.

Statistical analyses
Body weight, Se concentration, qRT-PCR, western blotting and DNA sequencing data were subjected to one-way ANOVA using the GLM procedure of SPSS, with signi cance reported at P < 0.05. Means were further separated using Duncan's multiple range test. All data were statistically processed as repeated measures to determine interaction of Se and B. subtilis. A P-value of less than 0.05 was considered statistically signi cant.

Results
Analysis of SEBS SEBS were harvested after fermentation of B. subtilis in medium containing sodium selenite. The medium appeared pale pink in color. The concentration of ionic Se in fermentation supernatant and precipitate of SEBS was measured using HG-AAS, and concentrations reached 1.77 µg/mL and 48.13 µg/mL, respectively. In precipitate, Se existed primarily as Se protein (valence 2-) and nano particles Se (valence 0) in the cell of B. subtilis. The live cell of B. subtilis and SEBS both reached 9.2 ⋅ 10 8 CFU/mL. The appearance and internal structure did not vary between B. subtilis and SEBS, this was proved using SEM and TEM (Fig. 2a, 2b).

Impact on growth performance and mortality
To compare the probiotic effect of B. subtilis and SEBS, the measurements regarding body weight of chicks and the results of growth performance and mortality of Chinese Huainan Partridge chickens are shown in Fig. 3. The nal body weight of chicks with SEBS supplementation was signi cantly increased (P < 0.01) compared to those of the control and inorganic Se supplemented groups, indicating a body weight increase of 220 g (Fig. 3a). In suitable living temperature and nutrition, morbidity of chicks was not found in 0-5 weeks. Between 6-8 weeks for the cold condition, died chicks could be found. The mortality of chicks with B. subtilis, SEBS, and avomycin supplementation was signi cantly reduced ( Fig. 3b, P < 0.01). The result with SEBS exposure appears lowest, with a decrease of 6.1 compared to controls (P < 0.01).

Colonization and levels of B. subtilis and SEBS in intestine
To explore the colonization of B. subtilis in intestinal mucous membranes, we designed an in vivo study and employed FISH and qRT-PCR. B. subtilis is indicated by green spots detected in the distal segment of the ileum, indicated by the FISH assay (Fig. 4a), with both B. subtilis and SEBS supplementation. The qRT-PCR assays agree with the FISH results. The standard curve of B. subtilis is based on the tenfold dilution series of fermented culture (Fig. 5b). B. subtilis growth increased while being monitored in the distal segment of the ileum (Fig. 4c, P < 0.01). However, salmonella decreased with B. subtilis and avomycin supplementation (Fig. 4d). In contrast, B. subtilis supplementation increases the number of total lactic acid bacteria and bacillus in ileal mucous membranes (Fig. 4e).

SEBS optimized ileal microbiota
High-throughput sequencing of all samples produced a total of 602,704 clean tags, which were identi ed as a total of 551 operational taxonomic units (OTUs) in all samples (Fig. 5a). This sequencing depth closely re ects the total microbial species richness. Chicks with SEBS supplementation represented 432 OTUs, the most abundant gure (Fig. 5b). All ve groups are represented by 234 OTUs, contributing 42,47% of the total proportion (Fig. 5c). Bacterial compositon in pulum show a few differences with SEBS and BS supplementation (R = 0.2274). Similarities of the weighted UniFrac-based principal coordinate analysis (PCoA) indicated the mainly factors dominated 67.68% variations (Fig. 6b), which in uenced the composition of mcirobiota. Results of bacterial community structure indicate that the numbers of three speci c phyla of bacteria, namely Bacteroidetes, Actinobacteria, and Epsilonbacteraeota, signi cantly increased with B. subtilis and SEBS supplementation compared with controls, with Actinobacteria higher than IS and avomycin groups ( Figure S1, S2). Regarding the genus, the bacterial diversity of all supplementary groups was more abundant than controls ( Figure S3, Fig. 6a). Six genera represent 87.28% of the total proportion, Candidatus arthromitus, Romboutsia, Escherichia-Shigella, Enterococcus, Gallibacterium, and Tyzzerellawere, with Candidatus arthromitus representing 35.54% in controls ( Figure  S3). Further, Alistipes, Helicobacter, Ruminococcaceae, and Ruminococcus genera were exhibited in supplementary groups, and Lactobacillus and Bacteroides were found in B. subtilis and SEBS groups. Similarities of PCoA showed that two main factors in uenced the bacterial cluster at the genus level, indicating a ratio of 61.09% (Fig. 6b, R = 0.2545). In the PCoA of bacterial OTUs, the SEBS group samples clustered together more than any other group. Results from controls were most scattered. These suggest that the bacterial communities were most stable and optimal with SEBS supplementation.

SEBC increases organic selenium levels and GPX
The Se levels in the plasma and breast meat were measured, with the results shown in Figure S4. The results indicate that the IS group displays signi cantly more Se in the plasma and breast meat than the control and BS supplementation groups (P < 0.01). Chickens in the SEBS group exhibited the highest Se levels in both tissues (P < 0.01). Types of GPX, cytoplasmic (GPX1), plasma (GPX2), and gastrointestinal (GPX4), observed in chicks were primarily antioxidative factors, containing selenium as a co-factor. The activities of GPX1, GPX2, and GPX3 were measured using ELISA assays. IS and SEBS improved the activities of GPX1, GPX2, and GPX3 compared with those without Se supplementation.

SEBS improves immunity and metabolism
The relationship between the microbiota of ileal mucous membranes and chicken body function was manifested through KEGG pathway classi cation and one-way ANOVA analysis. Results (Fig. 7a and 7b) indicate that more microbiota were found to promote the genetic expression of amino acids, carbohydrates, co-enzymes, lipid transport and metabolism (P < 0.01), energy production and conversion (P < 0.01), signal transduction mechanisms, and defense mechanisms (P < 0.05). Meanwhile, primary infectious diseases caused by pathogenic bacteria was analyzed (Fig. 7c), which showed that all four supplements strengthened defenses against bacterial invasion of chicken epithelial cells, compared to controls (P < 0.01). With B. subtilis, SEBS, and avomycin supplementation, body defenses regarding bio lm formation against Vibrio cholerae improved signi cantly (P < 0.01). Further, defense against Salmonella infection and Pertussis improved in two B. subtilis and avomycin supplementary groups (Fig. 8). Moreover, chicks receiving SEBS exhibited enhanced defense against pathogenic Escherichia coli infection and Shigellosis (P < 0.01).
Chicken immunity was regulated by intestinal bacterial composition. To unveil the relationship, we used the Spearman's correlation analysis on species of bacteria and environmental factors, BD1 levels and anti-infection, results (Fig. 8) indicated that the abundance of species of uncultured Candidatus arthromitus were negatively correlated with BD1 concentration and anti infection in ileal mucous membranes (P < 0.01), and included Romboutsia regarding anti-infection analysis. However, abundant species of unclassi ed Lachnospiraceae, uncultured Anaerosporobacter, uncultured Ruminococcaceae_UCG-014, uncultured Peptococcus, Lactobacillus salivarius, and unclassi ed Butyricicoccus were positively correlated with two environmental factors (P < 0.01). The diversity of species in all groups is shown in Figure S5 and text 1. Species of unclassi ed Lachnospiraceae, uncultured Anaerosporobacter, Ruminococcaceae_UCG-014, Peptococcus, Lactobacillus salivarius, and unclassi ed Butyricicoccus were all increased in SEBS-supplemented chicks (Fig. S4). Lachnospiraceaer, Ruminococcaceae_UCG-014, Peptococcus, Lactobacillus salivarius, and Butyricicoccus aid in digestion and nutrient absorption. Speci cally, species of Lactobacillus salivarius, Butyricicoccus, and Anaerosporobacter are associated with immunity and BD1 concentration in ileal mucous membranes.

Improvement of intestinal microbiota and levels of BD1
BD1 of the innate immune system have been found in chicken intestine. BD1 was formed and secreted by small intestinal crypt cells (Fig. 9a), the protein of BD1 was transported from the intestinal crypt to villus and released from villous brush border into lumen indicated by immunohistochemical assays. A greater concentration of BD1 was secreted from the crypt of ileal mucous membranes of chicks with dietary Se, B. subtilis, SEBS, and avomycin supplementation (P < 0.01, Fig. 9b). Chicks supplemented with IS exhibited improved secretion of BD1 (P < 0.01), with signi cant differences observed in ileum levels compared with controls (P < 0.01). The levels of BD1 with B. subtilis, SEBS, and avomycin were signi cantly higher than that of the control, while BD1 levels were highest with SEBS inclusion (P < 0.01, Fig. 9c).

Potential signaling pathway of SEBC
To explore the potential signaling pathway of BD1 expression regulated by SEBS, mouse primary intestinal crypt cells were cultured. The duration of proliferation of primary crypt cells reached 80% cell fusion was 48 h (Fig. 10a, 10b). Crypt cells treated with B. subtilis, Se-BS, and SEBS showed improved expression of BD1 than those of control and IS cells (Na 2 SeO 3 ) (Fig. 10c, P < 0.01). B. subtilis exhibited the most robust effect on BD1 expression and several cytokines, with or without an antagonistic role.
With the action of Se, expression of all measured cytokines and pro-in ammatory factors, IL-1β and TNFα, reduced signi cantly compared to B. subtilis use alone. SEBS played a weaker role in activating the signals of IκB, IL-1β, and TNF-α than Se-BS and B. subtilis.
The mouse crypt cell results indicate that the potential signaling pathway of B. subtilis and SEBS may originate from the recognition of TLR2 and presentation to MyD88, stimulating and provoking downstream signals of IRAK, TRAF6, IκB, IL-1β, and TNF-α. Experiments conducted in mouse crypt cells indicate that the recognized receptor was TLR2, and the signaling pathway is the classic TLR2-MyD88-NF-kB pathway.
The signaling molecules in the chick ileum were further inspected using qRT-PCR. The mRNA expression of cytokines BD1, TLR2, IL-1β, TNF-α, IL-6, IFN-γ, and NF-kB1 were monitored (Fig. S7), with increased expression of all cytokines observed in in SEBS, BS, and avomycin groups, compared to controls (P < 0.01). In the IS-supplemented group, the expression of BD1 and IFN-γ were signi cantly improved compared with control groups (P < 0.01). The expression of cytokines IL-1β, TNF-α, IL-6, and IFN-γ in the B. subtilis group was higher than those observed in the SEBS group (P < 0.01). Cytokines IL-1β, TNF-α, IL-6, and IFN-γ were also pro-in ammatory factors, which easily induce in ammation. With the action of Se, the expression of three pro-in ammatory factors was signi cantly reduced.

Discussion
In order to explore the combined effect of Se and B. subtilis, we rst cultured the SEBS, then analyzed its morphological and biochemical characteristics. These characteristics did not exhibit any differences compared to B. subtilis after being bio-transformed with inorganic Se. Selenomethionine was proved as the primary ionic form of selenium in bacteria after fermentation [31]. The ionic form of Se in the supernatant and precipitate of fermented SEBS medium constituted nano-selenium in red particles [32].
Hence, the composition of Se in the supernatant and precipitate of fermented SEBS medium was mainly as selenomethionine and nano-selenium in red particles, which confers a pale pink color to the medium.
The valences of Se changed from 4 + into 2-and 0. Then, Chinese Huainan Partridge chickens were supplemented with SECB, to unveil the effect on intestinal innate immune expression of BD1 and its potential mechanism.
Chicks exhibited higher nal body weights in the chicken study with B. subtilis and SEBS supplementation. The promoting effect on growth was also shown in the nal body weight of the B. subtilis, SEBS, and avomycin groups. This result was conformed with previous study [33,34]. Both Se and B. subtilis modulate the growth performance of chickens [35,36]. However, we did not observe any signi cant effects of IS supplementation. In our study, we supplemented basal feedstuff with a dose of 0.5 µg/g Se in an inorganic form, which had no positive effects on Chinese Huainan Partridge chickens; this is in accordance with our previous studies [2,3,37].
Exploring the colonization of bacteria in vivo must be more eloquent than in cells in vitro. B. subtilis fed to chicks can colonize in the ileal mucous membranes, the bacterial number was ascertained using FISH and qRT-PCR assays. The bacteria used nutrients in the intestine, ingested by the chick, for propagation, while conferring health bene ts to the chick indicating a reciprocal relationship [38,39]. With bacterial growth, metabolites of B. subtilis include antimicrobial substances, digestive enzymes, such as protease, lipase, and amylase, which play important roles in maintaining health and breaking down feedstuff for nutrient absorption [40]. Chickens receiving SEBS exhibited a higher nal body weight and greater feed utilization e ciency than control chickens, compared with the group receiving B. subtilis, suggesting that this treatment was more e cient in regulating growth for the biological roles of Se than B. subtilis alone. SEBS combined the merits of Se and B. subtilis to improve growth performance of chicks.
One bene t of B. subtilis is represented by improving the development of intestinal stem cells, such as promoting the development of intestinal stem cells, increasing the villous height of the small intestine, and decreasing crypt depth [33,41]. Accompanied with the colonization of B. subtilis, immunity was also improved through such reciprocal pathways [39]. The results of immunohistochemical assays showed that BD1 is secreted in the small intestinal crypt and stimulated with B. subtilis and SEBS colonization, which improved innate immunity to defend against infection by endogenous pathogenic or conditioned pathogenic bacteria [6,15]. The colonization of B. subtilis in the ileal mucous membrane enriched bacterial diversity, which improved the phylum and genus of bacterial composition. Genera of lactobacillus, Peptococcus, Butyricicoccus, and Ruminococcaceae_UCG-014 serve probiotic functions for the body, and the proportion of these bacteria increased in chicks receiving B. subtilis and SEBS supplementation [42][43][44]. Meanwhile, the ration of conditioned pathogens or pathogens, such as Escherichia-Shigella, Vibrio cholerae, Salmonella, and Pertussis bacilli, signi cantly declined. Improved immunity can be indirectly attributed to these results. Abundant species of unclassi ed Lachnospiraceae, uncultured Anaerosporobacter, uncultured Ruminococcaceae_UCG-014, uncultured Peptococcus, Lactobacillus salivarius, and unclassi ed Butyricicoccus can be detected in B. subtilis and SEBS groups. The species of Lachnospiraceaer, Ruminococcaceae_UCG-014, Peptococcus, Lactobacillus salivarius, Butyricicoccus, and Anaerosporobacter improved overall health [45][46][47][48]. KEGG Ortholog (KO) functional pro les of microbial communities indicated that metabolism and immunity were improved by these species of bacteria. Body immunity was improved by optimized intestinal microbiota. BD1 as an important component of innate immune, to determine the mechanism of BD1 expression regulated by SEBS is urgent. We rst explored the potential signaling pathway of B. subtilis and SEBS on mouse ileal crypt cells, then chicken crypt cells were employed. The innate immune system initiates a response to microorganisms via pattern recognition receptors, which have been extensively studied to explain the regulatory effects of immunobiotic bacteria in both immune and intestinal epithelial cells [49]. TLRs play a critical role in the recognition of pathogens by the innate immune system [50]. Individual TLRs recognize distinct components of microorganisms. The TLRs primarily responsible for recognition of gram-positive bacteria are TLR 2, 4, and 6 [13]. Results of mouse ileal crypt cells treated with agonists Pam3, LPS, and FSL-1 showed varying in uences on target signal molecular expression. No differences were observed regarding the function of agonists LPS and FSL-1, compared with non-supplementation. However, with the action of Pam3, the classic TLR2 signaling pathway was stimulated under the in uence of B. subtilis and SEBS, which indicates that B. subtilis was recognized by TLR2, as opposed to TLR4 and 6. In further, the mouse crypt cells pretreated with anti-TLR2 protein, the down-stream signaling pathway was inhibited. These results suggested that the TLR2 was the recognized receptor of B. subtilis. TLR2 is an important PRR that plays an essential role in immune regulation by probiotics. It has a broad recognition pro le that includes products from gram-positive bacteria, including lipoteichoic acid, peptidoglycan, lipoproteins [50,51]. B. subtilis and SEBS as gram-positive bacteria adhered to ileal mucous membrane as an integrated whole, which were recognized by TLR2 and presented into downstream. The signal was presented downstream to MyD88, IRAK, TRAF6, IκB, BD1, IL-1β, TNF-α, and IL-6 [14,52], which resulted in nuclear localization and stimulated nuclear factor kappa B1-induced expression of BD1 protein. The TLRs-NF-kB1 pathway represents the classic signaling pathway in cells, which was studied and discovered in molecular biology research [52][53][54]. Our results proved that the same TLR2-MyD88-NF-kB1 pathway recognized SEBS to evoke BD1 expression. In addition, whether TLR2 exerts its actions by forming heterodimers with other TLRs namely TLR1 or TLR6 [55,56] and cooperating with TLR2 coreceptors remains to be investigated.
B. subtilis and SEBS inoculated in mouse crypt cells stimulated cellular immune in ammation, leading to increased expression of pro-in ammatory factors, such as IL-1β, TNF-α, and IL-6. Supplementary dose of 0.125 µg/mL Se in culture medium could downregulate the in ammation caused by B. subtilis infection [57]. The same results could be found in chicken ileum, with decreased expression of IL-1β, TNF-α, IL-6, and IFN-γ. While BD1 levels in vivo were higher in SEBS than that of B. subtilis, which was attributed to selenium functionality. In mouse crypt cells, the combined use of Na 2 SeO 3 and B. subtilis could reduce in ammation, this reduction was lower than that of SECB, which suggested that the bene ts were magni ed with combined use of Se bio-conversion by B. subtilis as SECB.
The supplementation of Se also enhanced the bacterial diversity of the ileum. SEBS, due to the effects of B. subtilis and Se, exerted a more robust effect. Previous studies have reported that Se supplementation can increase Se tissue content, although the results varied [58]. In our study, the Se contents in the plasma and liver increased signi cantly in the IS and SEBS groups, which indicated that the dietary supplement containing Se exerted a signi cant in uence on the Se level; this is in accordance with previous reports [37,57]. The effect of SEBS was greater than that of IS in this experiment, suggesting that Se availability was greater with SEBS supplementation. Meanwhile, the activities of cytoplasmic, plasma, and gastrointestinal GPX in chicks were enhanced with Se supplementation, which is helpful to overcome the stress caused by cold temperatures and other stressors [28]. Se supplementation enriched the bacterial diversity compared to controls, which enhanced nutrient metabolism and immunity, as indicated by results of KEGG function classi cation. Further, the expression of cytokines in mouse and chicken intestinal cells indicated that TNF-α, IFN-γ and BD1 expression increased. Selenium improved the body immunity and alleviated the in ammatory response meanwhile [59]. Both the qPCR results of mouse and chicken crypt cells suggested that the pro-in ammatory factors IL-1β, TNF-α, IL-6, and IFN-γ declined in combined use of B. subtilis and Se than single use of B. subtilis. And the inhibition of TLR2 led to decreased expression of alleviated roles. All of these advised that Se was partly through recognition of TLRs and presentation to the signaling pathway TLR2-NF-kB1 to affect the expression of cytokines as well as seleno-protein [22,60,61]. Se absorbed and deposited in organism as compound substances such as seleno-protein alleviate the bodily in ammatory response through TLRs recognized may be in uenced by the recognition of pro-in ammatory substances combined such as lipopolysaccharide or pathogen by certain TLRs. Ionic Se was transformed into organic Se and nano Se when culture with B. subtilis. Se in body of B. subtilis as SECB mainly as seleno-protein bounded with peptidoglycan, spore was recognized by TLR2 of intestinal crypt cells and activate the immune response.

Conclusion
In this study, we cultured selenium enriched B. subtilis (SEBS). Thereafter, SEBS was added to the diet of chicks to determine its effect on growth and immunity. A representative graphical model is shown in Fig. 1. Our study rst reported the colonization of probiotic bacteria B. subtilis in the bodily intestine using FISH and qRT-PCR. Thereafter, we observed that the composition of intestinal microbiota and immunity were improved under the interplay of colonization. BD1 as an important innate immune component, the potential signaling pathway of SEBS was proposed to be TLR2-MyD88 NF-kB1. Se binding to the body of B. subtilis can reduce the in ammatory response of intestinal crypt cells and promote the immunity and anti oxidation. Combined use of Se and B. subtilis as SEBS induced further improvements compared to those observed when administered alone. Our research provided a new avenue in use of probiotics and essential micro-elements.

Declarations
Ethics approval and consent to participate The experimental protocols in this study including animal husbandry and slaughter were approved by the Institution of Animal Science and Welfare of Anhui Province (no. IASWAP2017110649).

Consent for publication
Not applicable Availability of data and materials The data supporting the conclusions of this article are included within the article.

Competing interests
All of authors declare no con icts of interest.  The graphical conclusion of the SECB on expression of BD1. The morphologic and structural properties of bacteria between B. subtilis and SECB. Fig. 2a The morphologic of B. subtilis and SEBS (×10000 magni cation). Fig. 2b The structure of B. subtilis and SEBS (×10000 magni cation).

Figure 3
Growth performance and mortality. Fig. 3a Growth performance of different groups. Fig. 3b Mortality of different groups. Data was statistically processed as repeated measurements. The different superscript capital letters in the same color of column in the same experimental duration mean signi cant difference at 0.01 levels (P<0.01).

Figure 4
Colonization of measured bacteria of different groups with FISH and qRT-PCR assay. Fig. 4a Colonization of B. subtilis in the distal ileum measured with FISH. Fig. 4b The standard curve of B. subtilis. Fig. 4c Number of Colonized B. subtilis. Fig. 4d Colonization of endogenous Salmonella in the distal ileum of different groups. Fig. 4e Number of native bacteria in the ileal mucous membrane. Values are means, with standard errors are represented by vertical bars. Mean value was signi cantly different from that of the control group by one-way ANOVA followed by Tukey's multiple comparison tests: the different superscript capital letters in the same color of column in the same experimental duration mean signi cant difference at 0.01 levels (P<0.01).
Page 29/37 Figure 5 The composition of bacterial OTU in ileal mucous membrane (Venn gures). Fig. 5a The size of OTU in each groups. Fig. 5b The number of OTU in each groups. Fig. 6c Number of elements shared by 5, 4, 3, 2 or 1 groups.

Figure 6
The composition of bacteria in ileal mucous membrane. Fig. 6a The circos gure of relationship between bacterial composition (genus level). In circos gure, left half circle is the bacterial components in different treatment. Colored belts in outer layer are treatments, and colored belts in inner layer are bacterial species. The length of the belts is the proportion of bacterial abundances. The right half circle means the proportion of certain bacterial species contained in treatments. Colored belts in outer layer are bacterial species, and colored belts in inner layer are treatments. The length of the belts is the proportion of bacterial abundances. Fig. 6b PCoA analysis on bacterial composition (genus level). Fig. 6c One-way ANOVA analysis on bacterial composition (genus level). * mean signi cant difference at 0.05 levels (P<0.05), **mean signi cant difference at 0.01 levels (P<0.01), *** mean signi cant difference at 0.001 levels (P < 0.001).

Figure 7
Function classi cation based on bacterial composition. Fig. 7a KEGG pathway function classi cation. STAMP software was applied to detect the differentially abundant Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways among groups with false discovery rate correction. Fig. 7b One-way of ANOVA analysis between bacterial composition and function. * mean signi cant difference at 0.05 levels (P<0.05), **mean signi cant difference at 0.01 levels (P<0.01). Fig. 7c The relative abundance of pathogen in One-way of ANOVA analysis of KEGG pathway function classi cation caused by pathogenic bacteria. The different superscript capital letters in the same color of column in the same experimental duration mean signi cant difference at 0.01 levels (P<0.01).

Figure 9
Development of ileal micro-structure of groups. Fig. 9a Immunohistochemical of BD1 in ileal of groups.   Mouse intestinal crypt cells cultured for 24h, stained with 0.4% trypan blue dye (×400 magni cation). Fig.  10b Mouse intestinal crypt cells cultured for 48h, stained with 0.4% trypan blue dye(×400 magni cation). Fig. 10c The expression of cytokines crypt cells. mRNA: glyceraldehyde phosphate dehydrogenase (GAPDH) internal control was used for statistical comparison. Values are means, with standard errors are represented by vertical bars. Mean value was signi cantly different from that of the control group by oneway ANOVA followed by Tukey's multiple comparison tests: the different superscript capital letters in the same color of column in the same experimental duration mean signi cant difference at 0.01 levels (P<0.01).

Figure 11
The expression of cytokines in crypt cells pre-treated with anti-TLR2 protein. mRNA: glyceraldehyde phosphate dehydrogenase (GAPDH) internal control was used for statistical comparison. Values are means, with standard errors are represented by vertical bars. Mean value was signi cantly different from that of the control group by one-way ANOVA followed by Tukey's multiple comparison tests: the different superscript capital letters in the same color of column in the same experimental duration mean signi cant difference at 0.01 levels (P<0.01).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.