Impact of high‐grain diet feeding on mucosa‐associated bacterial community and gene expression of tight junction proteins in the small intestine of goats

Abstract The objective of this study was to investigate the impact of a high‐grain (HG) diet on the microbial fermentation, the composition of the mucosa‐associated bacterial microbiota, and the gene expression of tight junction proteins in the small intestine of goats. In the present study, we randomly assigned 10 male goats to either a hay diet (n = 5) or a HG diet (56.5% grain; n = 5) and then examined changes in the bacterial community using Illumina MiSeq sequencing and the expression of tight junction proteins using qRT‐PCR in the mucosa of the small intestine. The results showed that HG diet decreased the luminal pH (p = 0.005) and increased the lipopolysaccharide content (p < 0.001) in the digesta of the ileum, and it increased the concentration of total volatile fatty acids in the digesta of the jejunum (p = 0.015) and ileum (p = 0.007) compared with the hay diet. MiSeq sequencing results indicated that the HG diet increased (FDR = 0.007–0.028) the percentage of the genera Stenotrophomonas, Moraxella, Lactobacillus, and Prevotella in jejunal mucosa but decreased (FDR = 0.016) the abundance of Christensenellaceae R7 group in the ileal mucosa compared with the hay diet. Furthermore, the HG diet caused downregulation of the mRNA expression of claudin‐4, occludin, and ZO‐1 in jejunal and ileal mucosa (p < 0.05). Collectively, our data suggested that the HG diet induced changes in the relative abundance of some mucosa‐associated bacteria, in addition to downregulation of the mRNA expression of tight junction proteins in the small intestine. These findings provide new insights into the adaptation response of the small intestine to HG feeding in ruminants.


| INTRODUC TI ON
To increase the growth rates or milk yields, goats or dairy cattle are often fed high-grain (HG) diet in commercial ruminant production system (Plaizier, Krause, Gozho, & McBride, 2008). Changes in feeding strategies from forage to grains can result in an unhealthy accumulation of organic acids in the rumen (Kleen, Hooijer, Rehage, & Noordhuizen, 2003). Ruminants fed a HG diet may reduce the time spent chewing, thereby decreasing salivary secretion and the neutralization of volatile fatty acids (VFA), leading to a lower ruminal pH, which can depress feed intake and fiber digestion, negatively affect the health of ruminants, as well as alter milk composition and meat quality (Plaizier et al., 2008).
In recent years, nutritionists have paid great attention to HG feeding and gastrointestinal health in ruminants. However, most of the studies mainly focused on how HG feeding affects bacterial community and health of rumen and hindgut in dairy cows and goats (Liu, Xu, Zhu, & Mao, 2014;Ye, Liu, Feng, Zhu, & Mao, 2016). Little information is available about the bacterial community and health of the small intestine in ruminants during HG feeding. The small intestine not only is the functional organ responsible for nutrient digestion and absorption, but also is an important barrier and immune organ in the ruminants.
Its health status has an important role in maintaining animal performance (Branco, Harmon, Bohnert, Larson, & Bauer, 1999). When ruminants are fed a HG diet, excessive undigested starch or VFA flows from the rumen into the small intestine (Moharrery, Larsen, & Weisbjerg, 2014), which may induce imbalance of bacterial communities (Russell & Rychlik, 2001) in the small intestine. This imbalance can finally cause enterotoxaemia, leading to sudden death of animal (Glock & DeGroot, 1998;Russell & Rychlik, 2001). Therefore, the health of ruminants depends on developing nutritional strategies in order to maintain a well-balanced microflora in the small intestine.
Our previous study has showed that HG feeding altered the composition of digesta-associated bacteria in the ileum of goats (Mao, Huo, & Zhu, 2013). However, little information is available on changes in the composition and diversity of the mucosa-associated microbiota and mucosal health in the small intestine of ruminants fed a HG diet.
Mucosa-associated bacteria in the small intestine play a critical role in regulating the barrier function and immune system (Kelly et al., 2016). Thus, maintenance of the well balance of the mucosa-associated microbiota of small intestine in ruminants is extremely important for health and performance of goats.
Here, we hypothesized that feeding goats a HG diet could affect the microbial fermentation, composition and diversity of mucosaassociated bacterial communities, and mucosal health of jejunum and ileum. Therefore, the objective of this study was to investigate the changes in the composition and diversity of jejunal and ileal mucosa-attached microbiota in goats fed HG diets. In addition, we evaluated the impact of HG diets on the mRNA expression of mucosal tight junction proteins of the jejunum and ileum in goats. These findings will be important for better understanding of composition of mucosa-associated microbiota in the small intestine and how HG feeding affects the mucosal health of entire gastrointestine of the ruminants.

| Ethics statement
The experimental protocol was approved by the Animal Care and Use Committee of Nanjing Agricultural University, Jiangsu Province, China.

| Animals and experimental design
The detail of animal experimental design has been reported in the previous study (Liu et al., 2014;Ye et al., 2016). Ten, 8-month-old male goats (Boer × Yangtze River Delta White) were adapted to a whole hay diet for 14 days and then randomly divided into two different groups.
One group was fed a hay diet (hay, 4% concentrate and 96% hay, n = 5), and the other group was fed a HG diet (75% concentrate and 25% hay, n = 5). The experimental period was 7 weeks. Daily rations (900 g/animal) were divided into two equal portions. One portion was offered at 8:00 hr, and the other was offered at 17:00 hr. Each goat was placed in a separate pen (1.2 × 1.2 m), with free access to water. At the beginning of the feeding trial, there were no significant between-group differences in the body weights of the goats (27 ± 3 kg). The diets were designed according to the nutrient requirements of goats (NY/Y816-2004; Ministry of Agriculture of China, 2004). The metabolic energy in the hay group was 8.32 MJ/ kg dry matter, which permitted a maintenance requirement of 30 kg goat. The metabolic energy in the HG group was 11.56 MJ/kg dry matter, which permitted a growth rate of 200 g/day of goats. Table   S1 presents the nutrient composition of the diets in the study.

| Sample collection
On day 50, all the animals were slaughtered 4 hr after morning feeding. Jejunal and ileal tissue samples were then collected and washed three times using ice-cold phosphate-buffered saline (pH 7.4) within 5 min after slaughter. The jejunal and ileal mucosal samples were scratched using a glass slide and immediately placed in liquid nitrogen. The samples were then stored at −80°C until further processing for RNA extraction and microbial DNA extraction. Digesta of the jejunum and ileum was collected, and the pH was determined.
All the samples of digesta were then homogenized thoroughly and divided into two parts. The first part of the homogenized digesta was mixed thoroughly with a double volume of distilled water and was centrifuged at 2,000 g. The supernatant was then collected and stored at −20°C for determination of the VFA, lactate concentration, and amylase activity. The other part of the homogenized digesta was mixed with an equal volume of endotoxin-free water and was centrifuged at 13,000 g for 40 min. The supernatant was collected using a lipopolysaccharide (LPS)-free filter. The filtrate was transferred to a sterile glass tube and heated at 100°C for 30 min. It was then placed in freezer at a −20°C for later analysis of LPS.

| Measurements of parameters of jejunal and ileal fermentation
The pH of the jejunal and ileal digesta was measured immediately at the time of extraction using a pH meter (HI 9024C; HANNA Instruments, Woonsocket, RI, USA). The VFA and lactate concentrations were determined by methods described previously (Barker & Summerson, 1941;Ye et al., 2016). The LPS content was analyzed using a chromogenic end point Tachypleus amebocyte lysate assay kit (Chinese Horseshoe Crab Reagent Manufactory, Xiamen, China), as reported previously . The amylase activity was detected using an Amylase Activity Assay Kit (Jiancheng Bioengineering Institute, Nanjing, China).

| Mucosal microbial DNA extraction
The DNA of all the mucosal samples was extracted using a QIAamp Fast DNA Stool Mini Kit (Qiagen, Hilden, Germany). The mucosal samples were placed in a centrifuge tube, and 1 ml of InhibitEX buffer was added to each sample and then homogenized by vortexing for 1 min. The resulting liquid was incubated for 5 min at 70°C. The lysis temperature was increased to 95°C, and the cells were then vortexed for 15 s. The liquid was centrifuged for 1 min to produce a pellet of particles. The supernatant was pipetted into a centrifuge tube with Proteinase K (15 μl) and AL buffer (200 μl) and then heated at 70°C for 10 min. The lysate was then placed in 200 μl of ethanol, followed by washing with 500 μl of AW1 and AW2 buffer. Subsequently, the solution was centrifuged to obtain the supernatant. The buffer (200 µl) was then added to the supernatant, which was left at room temperature for 1 min, followed by elution of DNA from the sample.
The concentration of DNA was determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). The extracted DNA samples were stored at −80°C until further processing.
The PCR products were separated using 2% agarose gels, and an

| Sequencing data analyses
The data were analyzed using the Quantitative Insights into Microbial Ecology (QIIME; version 1.8.0; https://qiime.org/) software package (Caporaso et al., 2010). The trimmed sequence reads were distributed to every sample on the basis of barcodes.
High-quality sequences of >250 bp without ambiguous "N" bases and an average base quality score >25 were identified and used in the subsequent analysis. Chimeric sequences were filtered using UCHIME (Edgar, 2010). Using UPARSE (https://drive5.com/ uparse/), operational taxonomic units (OTUs) were clustered based on a threshold of 97% sequence similarity. The OTU sequences were classified at the taxonomic level using the RDP Classifier (Wang, Garrity, Tiedje, & Cole, 2007) and aligned using the Silva reference database (SILVA 128) (DeSantis et al., 2006) and PyNAST (Caporaso et al., 2010). The default index was set by QIIME. The abundance-based coverage estimator (ACE), chao1 estimator, Shannon's diversity index, and Simpson's diversity index were used to calculate the richness and diversity of the bacterial community.
The FastTree tool (Price, Dehal, & Arkin, 2009) was used to build a phylogenetic tree. A principal coordinate analysis (PCoA) (Lozupone & Knight, 2005) of Bray-Curtis distance matrix was performed to measure the strength of the sample cluster. Significant differences between the two groups were calculated based on a Bray-Curtis distance matrix-based analysis of molecular variance (AMOVA) using the MOTHUR program (vs. 1.39.0) (Schloss et al., 2009).

| Mucosal RNA extraction and real-time PCR
The jejunal and ileal mucosal RNA was isolated using TRIzol (Takara Bio, Otsu, Japan), as described by Chomczynski and Sacchi (1987) and quantified using a NanoDrop spectrophotometer (Thermo Scientific Inc., Wilmington, DE, USA). Aliquots of RNA samples were subjected to electrophoresis on a 1.4% agarose-formaldehyde gel to verify integrity. A PrimeScript RT reagent kit (Takara Bio, Otsu, Japan) was used for reverse transcription of RNA. The primers for tight junction proteins (claudin-1, claudin-4, occludin, and zonula occludens-1 [ZO-1]) have been described previously (Liu, Xu, Liu, Zhu, & Mao, 2013) and are listed in Table S2. The primer for the housekeeping gene was designed to recognize and amplify a conserved nucleotide sequence encoding 18S rRNA. The cDNA sequence was distinguished using a basic local alignment search tool (National Center for Biotechnology Information, Bethesda, MD, USA). The primers were designed by the Primer 5.0 (Whitehead Institute, Cambridge, MA, USA). Tight junction proteins, 18S rRNA, were analyzed using the ABI 7300 real-time PCR system (Applied Biosystems, Foster, City, CA, USA), with fluorescence detection of SYBR Green (Takara Bio, Otsu, Japan). The reaction conditions were 95°C for 30 s, 40 cycles at 95°C for 5 s, and then 60°C for 31 s. Then, 1-10 ng of cDNA was added to the 2× SYBR Green mix (Takara Bio), and primers (200 nmol/l of each primer), Rox Reference Dye (50×), and nuclease-free water were added to the mixture to reach a volume of 20 μl. The target gene expression normalized to that of the housekeeping gene was calculated using the 2 −ΔΔCT method.

| Statistical analyses
The statistical calculations were analyzed using the SPSS software package (SPSS vs. 20, SPSS Inc., Chicago, IL, USA). Fermentation parameters, bacterial diversity and richness data, and mRNA expression of tight junction proteins were analyzed by an independent t test. The relative abundance of microbial phyla, genera, and OTUs was determined using the nonparametric Kruskal-Wallis test. All p-values of the bacterial community obtained by the nonparametric Kruskal-Wallis test were corrected for the false discovery rate (FDR). A value of p < 0.05 was considered significant.

| Average richness and diversity of the jejunal and ileal mucosal microbiome
In this study, we detected a total of 360,439 reads in jejunal mucosal

| Composition of the jejunal and ileal mucosal microbiota
At the phylum level, the results revealed 19 phyla in jejunal mucosal samples and 20 phyla in ileal mucosal samples. In jejunal mucosal samples, most of the sequences belonged to the phyla Proteobacteria (59.33%) and Firmicutes (38.83%), followed by the phyla Actinobacteria (1.19%), Tenericutes (0.30%), and Bacteroidetes (0.24%) ( Values are means ± SEM, n = 5. The data were analyzed by an independent t test. by the phyla Actinobacteria (2.51%) and Bacteroidetes (0.33%) (Table 4). Bacteria belonging to these four phyla accounted for more than 99% of the sequences (Table 4).
At the genus level, the examination of jejunal and ileal muco-

| Changes in the abundance of jejunal and ileal mucosal bacterial phyla, genera, and OTUS
At the phylum level, in jejunal mucosa, the abundance of the phyla Bacteroidetes (FDR = 0.009), Synergistetes (FDR = 0.034), and unclassified Bacteria (FDR = 0.009) was greater in the HG group compared with the hay group (Table 3). However, the results revealed no significant difference between-group in the percentage of any bacterial phyla in ileal mucosa (FDR > 0.05) (Table 4).
At the genus level, in jejunal mucosa, the HG diet increased the abun-  (Table 5). In ileal mucosa, the HG diet reduced the relative abundance of the Christensenellaceae R7 group (FDR = 0.010), as compared with the hay diet (Table 6).
At the OTU level, of the top 30 OTUs detected in jejunal mucosa (Table S3), the relative abundance of five OTUs was higher in the HG group (FDR < 0.05), and the relative abundance of one OTU was reduced in ileum mucosa (FDR < 0.05) (Table S4).  Values are means ± SEM, n = 5. The data were analyzed by an independent t test.

| Relative mRNA expression of tight junction proteins in the jejunum and ileum
F I G U R E 1 Bray-Curtis distance matrix principal coordinate analysis (PCoA) of mucosa-attached bacterial microbiota in the jejunum (a) and ileum (b) based on the operational taxonomic unit data

| D ISCUSS I ON
Over the past few decades, research on the effect of a HG diet on VFA profiles in the rumen and hindgut of ruminants has been widely explored Metzler-Zebeli et al., 2013). However, little information is available on changes of VFA pattern in the small intestine during HG feeding. In the present study, we found that the HG diet increased the TVFA, propionate, and butyrate concentration in digesta of the jejunum and ileum. These results were also consistent with the findings of our previous study , which showed that, compared with 0% corn diet feeding, 50% corn diet feeding increased the levels of acetate, butyrate, and total VFA in the ileum digesta of goats. Our previous study has showed that HG feeding increased the concentrations of acetate, propionate, butyrate, and total VFAs in the rumen fluid of goats (Zhang, Ye, Liu, & Mao, 2017). Therefore, the higher VFA in small intestine of HG-fed goats may be due to the facts that the excessive VFA escaped from the rumen into the small intestine of goats. In the current study, due to the insufficient small intestinal digesta and large amount of water in digesta, we are not able to measure the amount of starch in small intestinal digesta. Therefore, we could not demonstrate directly that there were more rumen bypass and starch inflow into the small intestine of HG-fed goats. However, we found that the HG diet significantly increased the activity of alpha-amylase in the jejunum and ileum. In the small intestine, amylase breaks down starch into glucose, which is absorbed by the animal body (Nozière, Steinberg, Silberberg, & Morgavi, 2014). Swanson et al. (2000) confirmed that high starch and high energy diets can increase the activity of alpha-amylase and alpha-amylase protein in lambs. Thus, we inferred that there may be more rumen bypass and starch inflow into the small intestine in HG-fed goats, which resulted in higher VFA concentrations.
Mucosa-associated bacteria in the gastrointestine play a vital role in regulating the barrier function and immune system (Kelly et al., 2016). Previous researchers used the high-throughput sequencing method to study the diversity and composition of epithelium-as- producing several pro-inflammatory cytokines (Ialenti et al., 2006).
It is speculated that Halomonas may play a role in facilitating the gut immune development. However, this hypothesis needs to be documented in future study through in vitro cultivation approach. The common distribution of this genus in jejunum and ileum indicates its specific importance in immune function of small intestine. Our results showed that Bacillus is dominant in jejunal mucosa but not in TA B L E 3 Effects of high-grain (HG) feeding on average relative abundance of jejunal mucosa-attached bacterial microbiota (% of total sequence) at the phylum level in goats (n = 5)  (Puri, Beg, & Gupta, 2002) to facilitate starch and protein degradation. Therefore, the higher relative abundance of Bacillus in the jejunum may be linked with its specific function in digestion of starch and protein.
Previous studies showed that a long-term HG diet appeared to disrupt the balance of the ruminal microbiota and affect the rumen health of dairy cattle and goats (Hua et al., 2017;Wetzels, et al., 2017). In the present study, the HG diet increased the richness of bacterial communities of jejunal mucosa, but it seemed to have no apparent effect on the richness of bacterial communities of ileal mucosa, with no significant between-group difference observed.
These results indicate that the HG diet had a regional effect on the richness of the bacterial microbiota of the small intestine. These different effects of gut site may be due to the functional heterogeneity of jejunum and ileum in goats. Furthermore, the different changes in VFA patterns in the jejunum and ileum may also somehow affect the richness of the bacteria.
In addition, the current study revealed that the HG diet increased the abundance of Prevotella, Lactobacillus, Stenotrophomonas, and Moraxella in jejunal mucosa. Of these genera, Prevotella was the most abundant genus in the phylum Bacteroidetes of jejunal mucosal microbiota. A previous study reported that members of Prevotella utilized various substrates and that they played an essential role in plant starch and protein digestion (Avgustin, Wallace, & Flint, 1997). Research also showed that this bacterial genus exhibited a remarkable degree of phenotypic diversity, enabling it to occupy various ecological niches within the rumen (Avgustin et al., 1997).
Our results shed further light on the presence of Prevotella in the jejunal mucosa, especially in animals fed a HG diet. Additionally, Respondek, Goachet, and Julliand (2008) showed that the population of Lactobacillus, as the main lactate producer, was increased in the hindgut after a large intake of starch in horses. Thus, the higher proportion of Lactobacillus in jejunal mucosal microbiota may be associated with increased starch content in HG group. Jenkins, Waite, Mansfield, Kim, and Pluske (2015) reported that species belonging to the Christensenellaceae family were common in the rumen and that these species played an essential role in maintaining gastrointestinal structure and function. We also detected a higher proportion of the TA B L E 4 Effects of high-grain (HG) feeding on average relative abundance of ileal mucosa-attached bacterial microbiota (% of total sequence) at the phylum level in goats (n = 5) enteric pathogens pass the mucosal barrier, intestinal inflammation occurs (Pedron & Sansonetti, 2008). Our findings revealed that HG feeding increased the percentage of Stenotrophomonas and Moraxella in the jejunal mucosa. Of these two genera, Stenotrophomonas spp.
are common mucosal pathogens in patients with chronic obstructive pulmonary disease (Murphy, Brauer, Aebi, & Sethi, 2005). The relatively higher abundance of these two genera in the HG group indicates that HG feeding may have a negative effect on the jejunal health of goats. It is very interesting that we found that it has higher percentage of Stenotrophomonas and Moraxella in jejunum, but these two genera were not seen in ileum samples. Many previous studies have demonstrated that Stenotrophomonas (Lukas, Simunek, Mrazek, & Kopecny, 2010) and Moraxella (Camboim et al., 2012) were also identified in the rumen of cattle and goats. Therefore, we inferred that these pathogens in jejunum may origin from the rumen and when they entering into the ileum, the changed chemical environment may be not fit for their survival in the ileum.
The mucosal barrier of the small intestine of the host provides defense against the entry of harmful pathogens and toxins from the luminal digesta. Changes in the composition of jejunal and ileal mucosal microbiota and the chemical environment may affect the TA B L E 5 Effects of high-grain (HG) feeding on average relative abundance of genus level (% of total sequences) in the jejunal mucosa (n = 5) a integrity of the mucosal barrier of the small intestine (Ulluwishewa et al., 2011). Compared with ruminal epithelium, which has multicellular layers, the intestinal epithelium of ruminants consists only of a single layer of epithelial cells , making it more susceptible to be damaged by an abnormal chemical environment or bacteria. In the present study, the HG diet reduced the relative expression of claudin-4, occludin, and ZO-1 in the ileum, as shown by the real-time PCR results. The HG diet also lowered the expression of claudin-1, claudin-4, occludin, and ZO-1 genes in jejunal mucosa.
These results suggested that the HG diet may have damaged the mucosal barrier of the small intestine in goats. Our data are somewhat consistent with those of an earlier study (Liu et al., 2013), which found that HG feeding damaged the ruminal epithelium and downregulated the expression of epithelial tight junction proteins in goats. Small intestine is mainly responsible for protein digestion and absorption in ruminants. Previous study has shown that mucosa-associated bacteria play vital roles in nutrients transport (Kleessen & Blaut, 2005). Therefore, the changes in bacterial composition may TA B L E 6 Effects of high-grain (HG) feeding on average relative abundance of genus level (% of total sequences) in the ileal mucosa (n = 5) a disturb the amino acid absorption, which may somehow affect the synthesis of proteins related to tight junctions and immune function and finally resulted in dysfunction of barrier and immune in the small intestine. Presumably, a lower pH and higher level of LPS in digesta may also contribute to the damage of the mucosal barrier of the ileum. Moreover, previous studies showed that the pathogenic bacteria Stenotrophomonas and Moraxella might induce the release of inflammatory cytokines and result in mucosal damage in colitis (Elinav et al., 2011;Saleh & Trinchieri, 2011). Overall, the findings of the present study indicate that, in the small intestine of animals fed a HG diet, impaired epithelial barrier function may be related to changes in the composition of mucosal-attached microbiota.
Furthermore, a combination of alterations in the chemical environment and increases in LPS levels in response to a HG diet may have adverse effects on small intestinal health.
In summary, our research revealed increased concentrations of TVFA and butyrate in jejunal and ileal digesta, as well as decreased ileal pH levels, in goats fed a HG diet, as compared with goats fed a hay diet. Our findings also showed that feeding goats a HG diet increased the proportion of some potential pathogens (Stenotrophomonas and Moraxella) in jejunal mucosa. In ileal mucosa, the HG diet increased the relative abundance of the Christensenellaceae R7 group. The HG diet also decreased mRNA expression of tight junction proteins in the jejunum and ileum.
Overall, our findings indicate that feeding goats a HG diet changes the VFA patterns, alters the relative abundance of some specific bacteria associated with mucosa, and results in downregulation of genes related to tight junction proteins in the small intestine of goats.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

AUTH O R S CO NTR I B UTI O N
J. L., C. X., and D. S. carried out all the experiments. J. L., C. X., and S.
M. analyzed the data. S. M., J. L., and W. Z. were responsible for study design. J. L. and S. M. were responsible for writing the manuscript.

DATA ACCE SS I B I LIT Y
The sequencing data for all the samples analyzed in this study were deposited in the Sequence Read Archive (SRA) under accession SRP157077.

Junhua Liu
http://orcid.org/0000-0002-1028-5586 F I G U R E 2 The relative expression of tight junction protein genes in the jejunum (a) and ileum (b) of goats fed hay diet and high-grain (HG) diet (means ± SD, n = 5). *0.01 < p < 0.05, **0.001 < p < 0.01, ***p < 0.001. NS, not significant. The 18S rRNA (a housekeeping gene) mRNA level was used to normalize the relative amount of each studied mRNA, and the 2 −ΔΔCT method was used to analyze the data. The data were analyzed by an independent t test.