Influence of ad Libitum Feeding of Piglets With Bacillus Subtilis Fermented Liquid Feed on Gut Flora, Luminal Contents and Health

Some scholars caution that long-term ad libitum feeding with probiotic fermented food poses potential health risks to baby animals. We conducted a feeding experiment to investigate the influence of ad libitum feeding of pre-and post-weaned piglets with a Bacillus subtilis fermented diet on the gut microbiome, gut metabolomic profiles, bile acid metabolism, proinflammatory cytokines and faecal consistency. Compared with piglets fed a Bacillus subtilis-supplemented pellet diet, piglets fed the Bacillus subtilis fermented liquid diet had lower intestinal bacterial diversity (P > 0.05), higher intestinal fungal diversity (P > 0.05), more Firmicutes (P > 0.05), fewer Bacteroidetes, Actinobacteria and Proteobacteria (P > 0.05), higher concentrations of 3-hydroxypropionic acid (P < 0.05), orotic acid (P < 0.05), interleukin-6 (P < 0.01), lactic acid (P < 0.01), deoxycholic acid (P > 0.05) and lithocholic acid (P < 0.01) and a higher incidence of diarrhoea (P > 0.05). The data show that ad libitum feeding of piglets with a Bacillus subtilis fermented liquid diet during the suckling and early post-weaning periods promotes the growth of lactic acid bacteria, bile salt hydrolase-active bacteria and 7a-dehydroxylase-active bacteria in the intestinal lumen; disturbs the normal production of lactic acid, orotic acid and unconjugated bile acids; and increases circulating interleukin-6 levels and diarrhoea incidence.


Results
Operational Taxonomic Unit (OTU) and Alpha Diversity. The sequence data produced in this experiment have been deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) under accession number SRP060218. Data on OTU and alpha diversity of the microorganism communities in different dietary treatment groups are listed in Table 1 The OTU number, Chao 1 and Shannon values of bacterial communities in the jejunal luminal content of weaned piglets from the Bacillus subtilis fermented liquid diet (BFLD) group were lower (P > 0.05) than those of weaned piglets from the Bacillus subtilis-supplemented commercial pellet diet (BCPD) group. In contrast, the OTU numbers and Chao 1 values of bacterial communities in the colonic luminal content of weaned piglets from the BFLD group were higher (P > 0.05) than those of weaned piglets from the BCPD group, which did not hold true for the Shannon value. These results indicated that the feeding of Bacillus subtilis fermented liquid diet to piglets decreased bacterial richness and diversity in the jejunal luminal content and decreased bacterial diversity in the colonic luminal content. The Chao 1 values of the fungal community were lower (P > 0.05) in the jejunal luminal content but were higher (P > 0.05) in the colonic luminal content of weaned piglets from the BFLD group than those of weaned piglets from the BCPD group. These findings suggest that feeding with a Bacillus subtilis fermented liquid diet increases fungal diversity in the jejunal and colonic luminal contents of weaned piglets compared with feeding with a Bacillus subtilis-supplemented pellet diet.

Compositions and Relative Abundances of Microorganisms in Jejunal and Colonic Luminal
Contents. The compositions and relative abundances of microorganisms in the jejunal and colonic luminal contents are shown in Tables 2 and 3, respectively. Firmicutes and Ascomycota were the dominant phyla in the jejunal and colonic luminal contents of weaned piglets fed either the Bacillus subtilis fermented liquid diet or the Bacillus subtilis-supplemented commercial pellet diet. Compared with weaned piglets from the BCPD group, weaned piglets from the BFLD group had a higher (P > 0.05) relative abundance of Firmicutes in the jejunal and colonic luminal contents, a lower (P > 0.05) relative abundance of Ascomycota in the jejunal luminal content and a higher (P > 0.05) relative abundance of Ascomycota in the colonic luminal content. The relative abundances of Bacteroidetes, Actinobacteria and Proteobacteria were lower (P > 0.05) in the jejunal and colonic luminal contents of weaned piglets from the BFLD group compared with those from the BCPD group. At the genus level, Lactobacillus and Kazachstania were the dominant genera in jejunal and colonic luminal contents of weaned piglets from both the BFLD and BCPD groups. The relative abundance of Lactobacillus in the jejunal and colonic luminal contents of weaned piglets from the BFLD group was higher (P > 0.05) than that of weaned piglets from the BCPD group. Weaned piglets from the BFLD group had a lower (P > 0.05) relative abundance of Kazachstania in the jejunal luminal content and a higher (P > 0.05) relative abundance of Kazachstania in the colonic luminal content compared with weaned piglets from the BCPD group. The relative abundances of Streptococcus, Clostridium_sensu_stricto, Bacteroides and Flavobacterium in the jejunal luminal content of weaned piglets from the BFLD group were significantly lower (P < 0.01 or P < 0.05) than those of weaned piglets from the  Table 1. Results of OTU, species richness and diversity of microorganism communities.
BCPD group. The relative abundances of Pseudobutyrivibrio, Lachnospiraceae_unclassified, Erysipelotrichaceae_ unclassified, Ruminococcus, Clostridiales_unclassified and Lachnospiraceae_uncultured in the colons of weaned piglets from the BFLD group were significantly higher (P < 0.01 or P < 0.05) than those of weaned piglets from the BCPD group.

Differential Metabolite Levels in Jejunal and Colonic Luminal Contents. All metabolites found
at levels that differed between the two piglet groups are listed in Tables 4 and 5. Thirteen differentially observed metabolites in the jejunal luminal content and eleven differentially observed metabolites in the colonic luminal content were identified. Piglets from the BFLD group had higher (P < 0.05) relative levels of 3-hydroxypropionic acid and orotic acid in their jejunal luminal content and higher (P < 0.05) relative levels of stigmasterol in their colonic luminal content than piglets from the BCPD group. Except for 3-hydroxypropionic acid, orotic acid and stigmasterol, the relative levels of other differentially observed metabolites in the jejunal and colonic luminal contents of weaned piglets from the BFLD group were significantly lower (P < 0.01 or P < 0.05) than those of piglets from the BCPD group.
Serum Cytokines, Intestinal pH and Unconjugated Bile Acids. Piglets from the BFLD group had significantly higher serum interleukin-6 (IL-6) levels (P < 0.01) than piglets from the BCPD group (Table 6). There were no significant differences (P > 0.05) in the levels of serum tumour necrosis factor-alpha (TNF-α ) and IL-1β between the BFLD and BCPD groups. There was no significant difference in the pH values of the jejunal luminal content between the BFLD and BCPD groups (P > 0.05). However, the pH values of the colonic luminal content collected from piglets in the BFLD group were significantly lower than the pH values of the colonic luminal content collected from piglets in the BCPD group (P < 0.05).
The jejunal luminal content collected from the BFLD group had significantly higher lactic acid and lithocholic acid (LCA) concentrations (P < 0.01) and significantly lower cholic acid (CA) concentrations (P < 0.05) than that collected from the BFLD group. There were no significant differences in chenodeoxycholic acid (CDCA)

Phylum level
Genus level and deoxycholic acid (DCA) concentrations in the jejunal luminal content between the BFLD and BCPD groups (P > 0.05). The colonic luminal content collected from the BFLD group had significantly higher lactic acid, chenodeoxycholic acid and lithocholic acid concentrations than that collected from the BCPD group (P < 0.01). The

Phylum level
Genus level concentrations of cholic acid and deoxycholic acid in the colonic luminal content collected from the BFLD group were not significantly higher than those in the colonic content collected from the BCPD group (P > 0.05).
Diarrhoea Incidence. The data in Table 7 show that piglets from the BFLD group had a higher incidence of diarrhoea than piglets from the BCPD group at each experimental time point, but there was no significant difference (P > 0.05) in the incidence of diarrhoea between the BFLD and BCPD groups.

Discussion
Previous studies have reported that the diversity, composition and relative abundance of intestinal flora can be influenced by probiotic administration or dietary patterns 21,22 . The feeding of probiotics and a probiotic fermented diet to animals decreases microbial diversity, and the reduced microbial diversity is often associated with gastrointestinal disorders, including inflammatory bowel disease 23 . In the present study, 25-day feeding with a Bacillus subtilis fermented liquid diet to piglets aged 7 to 31 days decreased the bacterial diversity but increased the fungal diversity of jejunal and colonic luminal contents compared with 25-day feeding with a Bacillus subtilis-supplemented pellet diet; the decreased bacterial diversity of piglets fed a Bacillus subtilis fermented liquid diet resulted in a higher diarrhoea incidence than observed in the piglets fed a Bacillus subtilis-supplemented pellet diet. In addition, Ley et    luminal contents, piglets fed a Bacillus subtilis fermented liquid diet not only had higher (P > 0.05) relative abundances of organisms of the Firmicutes phylum and Lactobacillus genus but also had higher (P > 0.05) relative abundances of organisms of the Ascomycota phylum and Kazachstania genus than did piglets fed a Bacillus subtilis-supplemented pellet diet. These results indicate that a Bacillus subtilis fermented diet has an advantage in promoting the growth of the flora noted above because the Bacillus subtilis fermented liquid diet has ingredients more suitable for the growth of intestinal flora than the Bacillus subtilis-supplemented pellet diet. Orotic acid is often regarded as one of the major oxidative stressors at high concentrations 25 . Additionally, circulating levels of IL-6 and the growth of Coprococcus, Pseudobutyrivibrio and Dorea increase under the action of stressors 26 . Piglets from the BFLD group had significantly higher orotic acid levels in their jejunal luminal content than piglets from the BCPD group. As a result, piglets from the BFLD group had higher circulating IL-6 levels and higher relative abundances of Coprococcus, Pseudobutyrivibrio and Dorea than piglets from the BCPD group.
Elevated levels of circulating IL-6 are often associated with a number of diseases 27 . People with high IL-6 levels have a high risk of systemic mastocytosis 28 , and elevated circulating IL-6 has been proposed as a marker of inflammation linking obesity with insulin resistance and diabetes as well as atherosclerosis 29,30 . High serum levels of IL-6 may also be associated with ankylosing spondylitis in young people, which is characterized by intense joint pain, stiffness, weakness, marasmus and apocleisis 31 .
Studies have demonstrated that diarrhoea can be prevented by the administration of probiotics or probiotic fermented food 32 . However, Li et al. (2012) found that oral administration of Lactobacillus rhamnosus at a high dose to piglets caused diarrhoea 15 . Data in the present study also showed that piglets from the BFLD group had a higher incidence of diarrhoea than piglets from the BCPD group.
Conjugated bile acids have emulsifying and surfactant properties; they are more efficient than unconjugated bile acids in aiding in the emulsification of dietary lipids and preventing small intestinal bacterial overgrowth 33,34 . In normal conditions, the composition of bile acids in the intestine is often in a relative balance; only small amounts of conjugated bile acids are hydrolyzed into primary bile acids (CA and CDCA), and small amounts of primary bile acids are dehydroxylated into secondary bile acids (DCA and LCA) in the small intestine 35 . Approximately 95% of bile acids (conjugated and unconjugated) are reabsorbed by the distal ileum. The small percentage of bile acids remaining reaches the colon, where they are deconjugated and dehydroxylated by bacteria to produce the secondary bile acids (DCA and LCA) 36 . However, overgrowth of bile salt hydrolase-active and 7a-dehydroxylase-active bacteria in the intestine will alter the normal bile acid composition and damage normal enterohepatic circulation.  Table 6. Differences in serum cytokine, pH, lactic acid, total bile acids and unconjugated bile acids between BFLD group and BCPD group.   48 and Peptostreptococcus 49 are 7a-dehydroxylase-active intestinal bacteria. In the present study, compared with piglets from the BCPD group, piglets from the BFLD group had lower (P < 0.01) relative abundances of Clostridium and Bacteroides in their jejunal luminal content. This condition resulted in piglets from the BFLD group having lower CA and CDCA levels in their jejunal luminal content. Lactobacilli are also 7a-dehydroxylase-active bacteria, and the relative abundance of Lactobacilli in the jejunal luminal content of piglets from the BFLD group were higher than those in piglets from the BCPD group. This is the reason why piglets from the BFLD group had higher DCA and LCA levels in their jejunal luminal content than piglets from the BCPD group did. Piglets from the BFLD group also had higher relative abundances of Lactobacillus, Ruminococcaceae, Lachnospiraceae, Erysipelotrichaceae, Ruminococcus and Clostridiales in their colonic luminal contents than piglets from the BCPD group. Therefore, in the BFLD piglets, more conjugated bile acids were hydrolyzed into CA and CDCA, while CA and CDCA were dehydroxylated into DCA and LCA, respectively. As a result, piglets from the BFLD group had higher CA, CDCA, DCA and LCA levels in their colonic luminal content than piglets from the BCPD group.

BFLD group BCPD group p-value
Unconjugated bile acids are less water soluble than conjugated bile acids. Intense elevation of the concentration of unconjugated bile acids has detrimental effects on the intestinal mucosa, including mucosal damage, increased mucosal permeability and potentially colon cancer-promoting effects [50][51][52] . Increased CDCA and DCA levels can inhibit water absorption and induce water and sodium secretion by the colon at concentrations above 3 mmol/L 53 and can disturb the normal microbiota of the gut, leading to diarrhoea and mucosal inflammation in the intestinal contents 54 . Piglets from the BFLD group had higher CDCA and DCA in their colonic luminal contents than piglets from the BCPD group; this is one of the factors contributing to the higher diarrhoea incidence among piglets from the BFLD group.
The Bacillus subtilis fermented liquid diet contains more active components for the growth of lactic acid bacteria and a high lactic acid concentration 55,56 . These active components allowed lactic acid-producing bacteria to grow better in the intestines of piglets from the BFLD group than in the intestines of piglets from the BCPD group. The higher relative abundance of lactic acid bacteria together with the high lactic acid intake resulted in piglets from the BFLD group having higher lactic acid in their jejunal and colonic luminal contents than piglets from the BCPD group. Excessive lactic acid in the intestine often causes lactic acidosis, which can induce diarrhoea 57 . Thus, lactic acidosis is the other factor contributing to the higher diarrhoea incidence among piglets from the BFLD group than among piglets from the BCPD group.
In summary, ad libitum feeding of pre-and post-weaned piglets with a Bacillus subtilis fermented liquid diet decreased intestinal bacterial diversity and increased intestinal fungal diversity, circulating IL-6 levels, intestinal unconjugated bile acid concentrations and diarrhoea incidence. Lactic acidosis, dietary lipid malabsorption and the inducing effect of unconjugated bile salts are the underlying causes for the higher diarrhoea incidence among piglets fed the Bacillus subtilis fermented liquid diet.

Materials and Methods
Animal Treatment Protocol. Twelve lactating sows (Large White x Landrace, artificially inseminated with semen from one Duroc boar) with similar body conditions and suckling piglets were randomly assigned to one experimental group and one control group (6 lactating sows + 56 suckling piglets vs 6 lactating sows + 54 suckling piglets) at the 7 th day after farrowing. There was no significant difference (P > 0.05) in the average body weight (2.79 ± 0.19 kg/piglet vs 2.85 ± 0.27 kg/piglet) of suckling piglets at 7 days of age between the experimental and control groups. All lactating sows were fed the same commercial lactation diet (7.5 kg/d). Piglets in the experimental group and control group had free access to a Bacillus subtilis fermented liquid diet (live Bacillus subtilis: 12.75 × 10 8 CFU/g) or a Bacillus subtilis-supplemented commercial pellet diet (live Bacillus subtilis: 2.80 × 10 8 CFU/g), respectively, from 7 to 31 days of age. All suckling piglets were weaned at 21 days of age. The Bacillus subtilis fermented liquid diet was produced using a previously described method 46 .
A total of six piglets (each with a body weight closest to the average weight of the litter) in each dietary treatment group (3 males and 3 females) were slaughtered in the morning at 32 days of age according to the protocol approved by the Animal Ethics Committee of Jiangxi Agricultural University. Sample Collection. Before slaughter, blood was collected with a 10 mL fresh tube from the jugular vein.
Serum was separated by centrifugation after blood clotting and stored at − 20 °C for the analysis of serum cytokines.
After slaughter, the segments of the jejunum and colon were quickly excised. The contents of the jejunum and colon were separately collected with 10 mL fresh tubes and immediately stored at − 80 °C for the analysis of pH, lactic acid, unconjugated bile acids, microbial composition and differential metabolites. Enzyme-linked Immunosorbent Assay. The concentrations of interleukin − 1β , interleukin − 6 and tumour necrosis factor-alpha were determined in three replicates for each sample using enzyme-linked immunosorbent assay kits (R&D Systems, Nanjing Jiancheng Bioengineering Institute).
Concentrations of D-/L-lactic acid, cholic acid, chenodeoxycholic acid, deoxycholic acid and lithocholic acid in the intestinal contents were determined in three replicates for each sample using enzyme-linked immunosorbent assay kits (R&D Systems, Shanghai Enzyme-linked Biotechnology Co., Ltd.).
Scientific RepoRts | 7:44553 | DOI: 10.1038/srep44553 Microbiome and Metabolomics Analysis. Genomic DNA of each sample was extracted using the E.Z.N.A. Soil DNA kit (OMEGA, USA), and six genomic DNA preparations for each treatment group were pooled into three DNA mixtures prior to polymerase chain reaction (PCR).
Bacterial genomic DNA was amplified with primers covering the V1-V3 region of the 16 S rRNA bacterial gene; the bar-coded primers 27 F and 533 R containing A and B sequencing adaptors (454 Life Sciences) were used. The forward primer (B-27F) was 5′ -CCTATCCCCTGTGTGCCTTGGCAGTCGACTAGAGTTTGATCCTGGCTCAG -3′ ; the sequence of the B adaptor is shown in italics and is underlined. The reverse primer (A-533R) was 5′ -CCATCTCATCCCTGCGTGTCTCCGACGACTNNNNNNNNNNTTACCGCGGCTGCTGGCAC-3′ ; the sequence of the A adaptor is shown in italics and underlined, and the Ns represent a ten-base sample specific barcode sequence 58 .
Fungal genomic DNA was amplified using the forward primer (A-ITS1) and reverse primer (B-ITS4). The forward primer (A-ITS1) was 5′ -CCATCTCATCCCTGCGTGTCTCCGACGACT NNNNNNNNNNTCCGTAGGTGAACCTGCGG-3′ ; the sequence of adaptor A is shown in italics and underlined, and the Ns represent a ten-base sample specific barcode sequence. The reverse primer (B-ITS4) was 5′ -CCTATCCCCTGTGTGCCTTGGCAGTCGACTTCCTCCGCTTATTGATATGC-3′ , and the sequence of adaptor B is shown in italics and underlined.
The protocols for PCR, pyrosequencing, sequence processing and bioinformatic analyses were previously described 56,59 . Differential metabolites were determined using gas chromatography time of flight mass spectrometry 55 . Diarrhoea Incidence Calculation. Faecal consistency was visually examined at the same time each morning by the same person during experimental periods. A piglet was considered to have diarrhoea when it developed a pasty or watery faecal consistency. Diarrhoea incidence was defined as the percentage of animals with diarrhoea on a specific day 60 .
Statistical Analysis. Data analysis was performed with SPSS software (version 13.0). One-way analysis of variance (ANOVA) was used to evaluate significant differences between means with a significance level at α = 0.01. Tukey's test was used to perform comparisons between means. Data were presented as the means ± SEM.