Dietary Supplementation With Bacillus subtilis Direct-Fed Microbials Alters Chicken Intestinal Metabolite Levels

Direct-fed microbials (DFMs) are dietary supplements containing live microorganisms which confer a performance and health benefit to the host, but the mechanisms are unclear. Here, a metabolomics approach was used to identify changes in intestinal metabolite levels in chickens fed an unsupplemented diet or a diet supplemented with B. subtilis strain 1781 or strain 747. Body weight gains of chickens fed the B. subtilis-supplemented diets were increased up to 5.6% in the B. subtilis 1781 group and 7.6% in the B. subtilis 747 group compared with chickens fed the unsupplemented diet. Compared with unsupplemented controls, the levels of 83 metabolites were altered (p < 0.05) (25 increased, 58 decreased) in chickens given the B. subtilis 1781-supplemented diet, while 50 were altered (p < 0.05) (12 increased, 38 decreased) with the B. subtilis 747-supplemented diet. Twenty-two metabolites were altered (p < 0.05) (18 increased, 4 decreased) in the B. subtilis 1781 vs. B. subtilis 747 groups. A random forest analysis of the B. subtilis 1781 vs. control groups gave a predictive accuracy of 87.5%, while that of the B. subtilis 747 vs. control groups was 62.5%. A random forest analysis of the B. subtilis 1781 vs. B. subtilis 747 groups gave a predictive accuracy of 75.0%. Changes in the levels of these intestinal biochemicals provided a distinctive biochemical signature unique to each B. subtilis-supplemented group, and were characterized by alterations in the levels of dipeptides (alanylleucine, glutaminylleucine, phenylalanylalanine, valylglutamine), nucleosides (N1-methyladenosine, N6-methyladenosine, guanine, 2-deoxyguanosine), fatty acids (sebacate, valerylglycine, linoleoylcholine), and carbohydrates (fructose). These results provide the foundation for future studies to identify biochemicals that might be used to improve poultry growth performance in the absence of antibiotic growth promoters.


INTRODUCTION
Microbial resistance to antibiotics is a serious and growing public health problem worldwide with multiple causes (1,2). One likely contributory factor is the widespread use of antibiotics as growth promoters in commercial livestock and poultry production. Dietary antibiotic growth promoters (AGPs) have been used in the food animal industry for more than 60 years to increase feed efficiency and improve growth performance (3,4). Increasing evidence, however, suggests that AGP use in food animal production leads to the development of antibiotic resistance among the endogenous gut commensal microbiota with the potential for transfer to the human population (5)(6)(7)(8)(9). Consequently, there is an unmet need to elucidate the molecular and cellular interactions between the intestinal microbiota and host that might be modulated by other means to promote food animal growth in the absence of AGPs. Among the substances that have been tested as alternatives to AGPs are probiotics and direct-fed microbials (DFMs) (10)(11)(12).
DFMs are viable, naturally occurring microbial cultures, including bacteria belonging to the genera Lactobacillus, Enterococcus, Streptococcus, Bifidobacterium, and Propionibacterium, which when fed to the host animal, generate a beneficial health response through their ability to modulate the diversity and composition of the gut microbiota. Various species of Bacillus also have been tested as DFMs on the basis of their ability to inhibit pathogens by producing antimicrobials (13)(14)(15). Bacillus has the added advantage as a DFM of producing spores that are resistant to low pH, bile salts, and other harsh conditions of the gastric environment (16)(17)(18)(19). In poultry, multiple strains of B. subtilis have been used as DFMs to promote growth, immunity, and overall gut health (11,12). In humans, both observational studies and randomized controlled trials have documented the health-promoting effects of probiotic Bacillus spp., particularly when used to treat intestinal disorders (20).
Our prior report demonstrated that dietary supplementation of healthy broiler chickens with B. subtilis strain 1781 as a DFM increased body weight gains, compared with chickens fed an unsupplemented diet (21). Further, following infection with Eimeria maxima, oral supplementation of chickens with B. subtilis reduced the clinical signs of experimental avian coccidiosis, compared with unsupplemented and infected controls (22). Independently, we also reported that the AGPs, virginiamycin and bacitracin methylene disalicylate, altered the chicken intestinal metabolome and speculated that their growth enhancing effects were due to the selection of gut microbial species capable of producing metabolites that promoted more efficient energy utilization (23). Therefore, the current study was undertaken to characterize the metabolic alterations in the chicken gut following dietary supplementation with B. subtilis DFMs with the goal of identifying potential chemical compounds that might be directly used to improve poultry growth performance without the use of AGPs.

Animals
Eighty-four, day-old, male Ross 708 broiler chickens (Longenecker's Hatchery, Elizabethtown, PA) were randomly housed in starter brooder cages (Petersime, Zulte, Belgium) and provided with starter feed.  (24). Each treatment group contained 4 cages with 7 chickens/cage. Cage dimensions were 0.65 × 0.75 m for a total of 0.4875 m 2 , resulting in 14.4 chickens/m 2 . All chickens were housed in the same room and provided adlibitum access to water and feed throughout the study. Feed additions were weighed and recorded daily, and feeders were shaken once per day. The chickens and feed were weighed at 14 and 21 days of age for computation of growth performance. Dead chickens were removed and weighed to calculate mortality and adjust growth performance data. At 21 days of age, 8 chickens/group were euthanized by cervical dislocation and the intestinal ileum harvested. Intestinal ileal contents were collected aseptically by gently finger-stripping the ileal segment, immediately placed on dry ice, and stored at −80 • C. Global metabolomic profiling of the intestinal contents was performed by mass spectrometry (MS) (Metabolon, Durham, NC) as described (23). Raw data was extracted and processed using the DiscoveryHD4 global metabolomics platform. Compounds were identified by comparison to library entries of purified standards or recurrent unknown entities based on retention index, accurate mass match to the library ±10 ppm, and MS/MS forward and reverse scores between experimental data and authentic standards. MS/MS scores were based on comparison of the ions present in the experimental spectrum to the ions present in the library spectrum.

Statistical Analysis
Each cage was considered the experimental unit. The type of experimental diet was considered the treatment factor, and each cage was considered as a blocking factor. Data were analyzed using a mixed model methodology (PROC MIXED, SAS Institute, Cary NC). For growth performance, mean ± SEM values were calculated for initial body weight (IBW) at 14 days, final body weight (FBW) at 21 days, body weight gain (BWG) between 14 and 21 days, feed intake (FI) between 14 and 21 days, and feed efficiency (FE = BWG/FI) between 14 and 21 days. Differences between means were compared using the 2tailed Student's t-test with p ≤ 0.05 considered significantly different. All data analysis were conducted with the same way of Gadde et al. (23). Briefly, ANOVA was used to identify

B. subtilis DFMs Increase Chicken Growth Performance
Chickens were fed from 14 to 21 days post-hatch with a basal diet, or a basal diet supplemented with 1.5 × 10 5 colony forming units (CFU)/g feed of B. subtilis strain 1781 or B. subtilis strain 787. Body weights at 14 and 21 days of age, as well as feed intake and feed efficiency between 14 and 21 days, were identical between the 3 groups ( Table 2). However, body weight gains between 14 and 21 days were greater for chickens fed with either of the Bacillus-supplemented diets, compared with unsupplemented controls.

DISCUSSION
We previously reported that dietary supplementation of broiler chickens with B. subtilis-based DFMs increased body weight gains, compared with unsupplemented controls (21). Further, chickens fed a B. subtilis-containing diet had increased intestinal villus height and crypt depth, augmented mitogen-and antigen-induced spleen cell proliferation, greater macrophage phagocytosis of Salmonella bacteria, decreased serum levels of α-1-acid glycoprotein and nitric oxide, altered expression of cytokine genes (interferon-γ, interleukin-1β, CXCL2) in intestinal lymphocytes, and a modified CD4 + /CD8 + ratio of peripheral blood lymphocytes (24)(25)(26)(27)(28)  gains, and also demonstrated increased Lactobacillus spp. and reduced Escherichia coli levels in the intestinal contents of DFMsupplemented birds, compared with controls. Increased growth performance, nutrient utilization, and intestinal villus/crypt morphometry, as well as altered composition of gut microbiota, also were observed in chickens orally administered with Bacillus DFMs, compared with controls (32)(33)(34)(35).
In addition to their effects on healthy animals, Bacillus DFMs also have been shown to ameliorate the deleterious outcomes associated with enteric bacterial infections of poultry. B. subtilis dietary supplementation reversed the effects of Clostridium perfringens infection on decreased growth performance and alterations in the intestinal microbiota, compared with unsupplemented controls (24). In a C. perfringens/Eimeria maxima coinfection model of avian necrotic enteritis, feeding of B. subtilis reduced gut pathology and animal mortality, and restored alterations in the intestinal microbiome, compared with controls (22). Other studies have confirmed the ability of Bacillus DFMs to reduce C. perfringens colonization of the intestinal mucosa and improve animal growth performance in avian necrotic enteritis (36)(37)(38)(39)(40)(41). In general, while the beneficial effect of Bacillus DFMs on body weight gains in healthy, uninfected chickens is generally moderate (<10%), a more pronounced outcome of the dietary supplements has been observed when weight loss is aggravated as a consequence of pathogenic infection (42).
Given the reported effects of DFMs on modulating the chicken gut microbiome (10)(11)(12), and the ability of intestinal bacteria to synthesize vitamins and nutrients beneficial to the host (43), it is somewhat surprising that few reports have examined the ability of DFMs to alter the avian metabolome (35,44). Cao et al. (35) reported that dietary supplementation of chickens with B. amyloliquefaciens altered the levels of gut metabolites related to amino acid and glyceride metabolism. More specifically, increased levels of 4-aminobutyric acid, gentiobiose, quinic acid, 3,7,12-trihydroxycoprostane, N-ethylglycine, glycine, N-acetyl-D-galactosamine, and 5hydroxyindole-3-acetic acid, and reduced the levels of diglycerol and N-acetyl-β-D-mannosamine, were seen in the DFMsupplemented chickens, compared with unsupplemented controls. In another study, Wang et al. (44) observed that chickens fed a diet containing the yeast Kluyveromyces marxianus had altered serum levels of 60 biochemical metabolites (39 increased, 21 decreased), compared with controls. Biochemicals involved in carbohydrate and amino acid, primarily glutamate and glutamine, metabolism were identified as the most relevant.
In humans, probiotics have been used to treat a variety of diseases and disorders, and in some cases, metabolic profiling has begun to unravel their mechanisms of action. For example, in a double-blind, randomized, placebo-controlled clinical trial of a fermented milk probiotic provided to patients with irritable bowel syndrome, decreased selected clinical symptoms was correlated with increased serum glucose and tyrosine levels (45). In women with mastitis, consumption of a Lactobacillus probiotic decreased staphylococcal/streptococcal bacterial load in breast milk and breast pain, while increasing urinary levels of creatinine, hippuric acid (a glycine conjugate of benzoic acid), and trimethylamine-N-oxide (a choline metabolite) (46). Oral administration of a Bifidobacterium animalis probiotic to patients with atopic dermatitis improved dermatology-specific quality-of-life scores, while increasing fecal kynurenic acid and decreasing 3-hydroxyproprionic acid levels, compared with placebo controls (47). In a study of chronic kidney disease patients orally administered with a probiotic containing Streptococcus thermophilus, Lactobacillus acidophilus, and Bifidobacterium longum, serum levels of metabolites related to carbohydrate, choline, and energy metabolism differentiated individuals with increased vs. decreased blood urea nitrogen levels (48). Per os administration of a probiotic mixture containing 4 strains of lactobacilli, 3 strains of bifidobacteria, and 1 strain of Streptococcus thermophilus to infants with colic reduced clinical symptoms, while increasing fecal levels of alanine, leucine, isoleucine, acetate, and pyruvate and decreasing the levels of uracil and propylene glycol, compared with placebo controls (49). Finally, a similar relationship between reduction of disease scores and alterations in metabolic profiles induced by administration of bacterial probiotics has been reported in animal models of human diseases, including fecal metabolites in mice with chemical-induced colits (50) and serum metabolites in rats with depression-related behavioral changes (51).
The alteration of metabolites observed in the current study suggest that these small molecules (gut metabolites) might play a role in the recognition of pathogen-associated molecular patterns (PAMPs) through influencing the host immune response to maintain homeostasis against intestinal diseases and inflammation. In this way, these responses may ultimately lead to improved growth performance (52-55). Shortchain fatty acids (SCFAs), such as acetate, n-propionate, and nbutyrate, are known to play a role in the recognition of PAMPs (52), however, alternations in SCFA levels in the gut were not found in the current study. SCFAs can be sensed by G-protein coupled receptors (GPRs) which are expressed on many cells, such as epithelial cells, neutrophils, and macrophages (52). GPR pathways are associated with cytokines and tight junction protein expression (56,57). Lipid metabolites in the current study may also be sensed by GPRs, which implies that lipid metabolites can affect the regulation of immunity or inflammation in the gut.
In summary, to the best of our knowledge, this is the first report to demonstrate that dietary supplementation with B. subtilis has profound effects on the levels of a wide variety of chemical metabolites in the chicken gut, particularly those related to amino acids, nucleosides, fatty acids, and carbohydrates. Compared with unsupplemented controls, these altered metabolite levels provide a biochemical signature unique to each B. subtilis supplementation group. Our result suggest that altered metabolites can be used to maintain gut homeostasis within epithelial or immune cells, which might account for their affect on overall gut health as well as chicken growth. Through future in vitro and/or in vivo studies, identification of the altered metabolites that confer properties of AGPs would suggest their potential use as antibiotic alternatives.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/supplementary material.

ETHICS STATEMENT
The animal study was reviewed and approved by the Beltsville Agricultural Research Center Institutional Animal Care and Use Committee and performed in accordance with the guidelines and regulations.

AUTHOR CONTRIBUTIONS
IP and HL designed the research and conducted research. IP, EL, and HL analyzed data. IP, NZ, AS, TR, EL, and HL had responsibility for content. All authors read and approved the final manuscript.