Disruption of the Intestinal Mucosal Barrier Induced by High Fructose and Restraint Stress Is Regulated by the Intestinal Microbiota and Microbiota Metabolites

A high-fructose diet aggravated restraint stress-induced changes in the composition of the intestinal microbiome, in which the abundance of A. muciniphila was significantly increased. The high-fructose diet exacerbated restraint stress-induced the changes in the composition of the microbial metabolites, with taurine abundance being downregulated and histamine abundance upregulated. ABSTRACT Environmental (restraint stress) and dietary (high fructose) factors are key triggers for flares of inflammatory bowel disease; however, the mechanisms involved in this phenomenon are not fully elucidated. This study aimed to investigate the mechanisms by which restraint stress and high fructose damage the intestinal mucosal immune barrier. The feces of C57BL/6J mice were subjected to 16S rRNA and untargeted metabolome sequencing, and the intestinal histological structure was analyzed by immunohistochemistry and immunofluorescence staining. The mRNA and protein levels of the intestinal protein were analyzed by reverse transcription-PCR (RT-PCR), Western blotting, and enzyme-linked immunosorbent assay (ELISA). The metabolites of the microbiota were tested in vitro, and Akkermansia muciniphila was used for colonization in vivo. Dietary fructose exacerbated the development of restraint stress, with an extensive change in the composition of the gut microbiota and microbial metabolites. The disturbance of the microbiota composition led to an increase in the abundance of histamine and a decrease in the abundance of taurine, which inhibited the expression of tight junction and MUC2 proteins, destroyed the function of NLRP6, and reduced intestinal autophagy level; this in turn disrupted the function of colonic goblet cells to secrete mucus, leading to defects in the intestinal mucosal barrier, which ultimately codrives colon autoinflammation. However, A. muciniphila supplementation counteracted damage to the intestinal mucosal barrier by high fructose and restraint stress. Therefore, the gut microbiota and microbiota metabolites play an important role in maintaining microenvironment homeostasis of the intestinal mucosal barrier. IMPORTANCE A high-fructose diet aggravated restraint stress-induced changes in the composition of the intestinal microbiome, in which the abundance of A. muciniphila was significantly increased. The high-fructose diet exacerbated restraint stress-induced the changes in the composition of the microbial metabolites, with taurine abundance being downregulated and histamine abundance upregulated. High fructose and restraint stress induced colonic mucosal immune barrier damage, possibly due to changes in the abundance of the microbial metabolites taurine and histamine. Colonization with A. muciniphila stimulated the expression of the NLRP6 inflammasome and activated autophagy in goblet cells, thereby producing more new mucins, which could protect the intestinal mucosal barrier.

softer stools, consistent with increased stool pellet counts (P = 0.011), in five mice within 2 h compared with the control group ( Fig. 1b and c). Blood glucose concentration was elevated significantly in the group receiving high fructose and undergoing restraint stress (H1S group) (P , 0.001) compared to the control (C) group (Fig. 1d). Both serum corticosterone (CORT) and norepinephrine (NE) are markers of stress, and they increased during stress responses (24). The levels of CORT and NE in serum increased significantly after restraint stress treatment ( Fig. 1e and f). At the same time, intestinal permeability increased significantly in the restraint stress (S) (P = 0.013) and H1S groups (P , 0.001) (Fig. 1g). Serum endotoxin levels also changed significantly in the S (P = 0.034) and H1S (P , 0.001) groups compared to the control group (Fig. 1h), which was consistent with changes in intestinal permeability. Similarly, the weight of the liver was significantly increased in the high-fructose (H) group (P = 0.046) and H1S group (P = 0.002) (see Fig. S1A in the supplemental material). The weight of the spleen was significantly increased in the S (P = 0.000) and H1S (P , 0.001) groups (Fig. S1B). Changes in cytokines in serum were detected by Luminex liquid suspension chip technology. In the H1S group, the expression levels of the proinflammatory cytokines interleukin 12 (IL-12) (p70) (P = 0.002), IL-17a (P , 0.001), tumor necrosis factor alpha (TNF-a) (P = 0.006), and IL-6 (P = 0.042) were significantly higher than in the control group; also, the expression levels of IL-12 (p70) (P = 0.016) and IL-17a (P = 0.004) were significantly higher than in the S group ( Fig. 1i and j). The changes in the anti-inflammatory cytokines IL-5 (P = 0.002) and IL-10 (P = 0.000) were significantly lower in the H1S group than the control group (Fig. 1k). Next, we examined the effects of high fructose and restraint stress on the intestinal mucosal immune barrier. The histological score showed that the colon tissue of the H1S group had obvious colon tissue damage compared with the control group (P = 0.003), manifested as severe villous epithelial atrophy and crypt epithelial loss ( Fig. 2a and b). Alcian blue-periodic acid-Schiff (AB-PAS) staining showed that goblet cells were blue-purple and scattered among epithelial cells, mainly in the lower part of the villi. In our research, the number of goblet cells in the H1S group was significantly lower than in the S group (P , 0.001) (Fig. 2a and c). MUC2 is secreted primarily by goblet cells and enters the intestinal lumen to form a mucus layer. The integrated optical density (IOD) of MUC2-positive cells in the H1S group was significantly lower than in the S group (P = 0.001) ( Fig. 2a and d). Tff3 is a goblet cell protein that promotes mucosal repair and protection, and the transcription factor Klf3 participates in barrier function. It was consistent with our prediction that the mRNA levels of Muc2 (P = 0.044), Klf3 (P = 0.007), and Tff3 (P , 0.001) were decreased more severely in the H1S group than the S group (Fig. 2e). We further tested the effect of fructose intake and restraint stress on the antioxidant capacity of the intestine by examining five antioxidant parameters, including antioxidant enzymes (glutathione peroxidase [GSH-Px], superoxide dismutase [SOD], and catalase [CAT]), total antioxidant capability (T-AOC), and malondialdehyde (MDA). Consistent with our prediction, the levels of CAT (P , 0.001), GSH-Px (P = 0.000), SOD (P = 0.002), and CAT (P = 0.001) decreased significantly in the H1S group ( Fig. S2A to D). However, the level of MDA, the final product of lipid peroxidation, was significantly increased in the H1S (P = 0.001) group, compared with the control group (Fig. S2E). The above results indicated that high-fructose treatment could aggravate the stress-induced weakening of colon antioxidants. The levels of the proinflammatory cytokines IL-1a (P = 0.017), IL-12 (p70) (P = 0.015), IL-6 (P = 0.011), IL-17a (P = 0.039), RANTES (P = 0.008), and TNF-a (P = 0.030) increased significantly in the H1S group compared with the S group (Fig. 3a). Expression of the anti-inflammatory cytokines IL-5 (P = 0.005) and IL-13 (P = 0.005) decreased significantly in the H1S group compared with the S group (Fig. 3b). Moreover, high fructose and restraint stress activated the colonic NF-k B pathway. The protein levels of p-P65 (P , 0.001) and p-Ik B (P = 0.002) were increased in the H1S group compared with the S group ( Fig. 3c and d). This indicated that the elevated level of intestinal oxidative stress may induce intestinal inflammation. Through analysis of the reverse transcription-PCR (RT-PCR) results for different antimicrobial peptides, we found that the expression levels of Itln1 (P = 0.001) and Ang4 (P , 0.001) were significantly lower in the H1S group than the S group (Fig. 3e). Secretory immunoglobulin A (SIgA) is the first line of defense of the intestinal mucosal barrier; the secretion level of SIgA in the H1S group (P , 0.001) was significantly lower than in the S group (Fig. 3f). When the expression of tight junction proteins in colon tissue was assessed by Western blotting, the expression levels of occludin (P = 0.022), claudin-1 (P = 0.000), and claudin-3 (P = 0.029) in the H1S group were significantly lower than in the S group ( Fig. 3g and h). This indicates that high fructose intake could worsen restraint stress stimulation, aggravate the destruction of intestinal mucosa, reduce the secretion of antimicrobial peptides, and promote the expression of proinflammatory cytokines.
High fructose intake increased colonic apoptosis and reduced autophagy under restraint stress conditions. To further verify the influence of these factors on the decrease of goblet cell number and secretion capacity, the proliferation of colon epithelial cells was detected by immunohistochemistry using Ki-67. The IOD of Ki-67positive cells was significantly lower in the H1S group than in the S group (P = 0.000) ( Fig. 4a and b). TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) staining results showed that the number of apoptotic cells in the H1S group was significantly higher than that in the S group (P , 0.001) (Fig. 4a and c). The expression of antiapoptotic protein Bcl-2, apoptotic protein Bax, caspase-3, and poly(ADP-ribose) polymerase (PARP) was detected by Western blotting. The results showed that the expression of Bcl-2 (P , 0.001) was significantly inhibited, while the expression of Bax (P , 0.001), c-caspase-3 (19 kDa, P , 0.001; 17 kDa, P , 0.001), and cleaved (c)-PARP (P = 0.025) was significantly increased in the H1S group compared with the S group ( Fig. 4d and e). The mRNA levels of autophagy-related proteins in H1S group were significantly decreased (Fig. 5a). Western blotting revealed that the expression levels of ATG7 (P = 0.003) and LC3II/LC3I (P , 0.001) in the H1S group were significantly lower than in the S group, while the expression level of SQSTM1/P62 (P , 0.001) in the H1S group was significantly higher than in the S group ( Fig. 5b and c). In summary, these results indicated that high fructose intake could aggravate apoptosis and enhance the inhibition of autophagy in colonic tissue under restraint stress.
High fructose intake could aggravate the inhibition of NLRP6 inflammasome expression under restraint stress conditions. Next, we investigated the expression level of the NLRP6 inflammasome in the colon. First, RT-PCR results showed that the mRNA levels of Nlrp6 (P , 0.001), Asc (P , 0.001), caspase-1 (P , 0.001), and IL-18 (P = 0.005) were significantly inhibited in the H1S group compared with the control group (Fig. 5d). This decrease in the levels of NLRP6-related protein was further confirmed by Western blotting analysis. Likewise, in the H1S group, the protein expression levels of ASC/TMS1 (P = 0.016) and caspase-1 (P = 0.009) were significantly lower than the S group ( Fig. 5e and f). Moreover, we further verified the correlation between the expression of NLRP6 inflammasome and autophagy. We used LC3B and NLRP6 to perform immunofluorescence double-staining experiments. The results showed that decreased levels of LC3B (P = 0.000) were localized to colonic NLRP6 (P , 0.001) in the H1S group, compared with the control group ( Fig. 5g to i). These findings suggested that High fructose intake promoted gut microbiota disturbances under restraint stress. The effects of dietary fructose and restraint stress on the colonic microflora of mice were evaluated by 16S rRNA gene amplicon sequencing. The number of operational taxonomic units (OTUs) in the S group was significantly increased (P = 0.008), but there was no significant difference among the other groups (Fig. 6a). Analysis of alpha diversity using Shannon and Simpson index, Shannon index (P = 0.001 to 0.003) and Simpson index (P = 0.014 to 0.003) showed that species richness was significantly reduced in the H and H1S groups ( Fig. 6b and c). The Bray-Curtis beta diversity of the four groups of microbial populations and the clustering pattern on principal-coordinate analysis (PCoA) and nonmetric multidimensional scaling (NMDS) plots revealed a distinct clustering of microbiota composition ( Fig. 6d and e). The sample heat map analysis by the Bray-Curtis method showed that stress and high-fructose treatment could significantly affect the difference between the microbiotas of the two groups (Fig. 6f). Specifically, we characterized the effect of restraint stress and high fructose on relative abundance of different bacterial taxa at the phylum and genus levels ( Fig. 6g and h). We used the linear discriminant analysis (LDA) effect size (LEfSe) algorithm to perform LDA to identify operational microbial taxa that were differentially abundant with regard to restraint stress and fructose intake ( Fig. 7a and b). In the H1S group, Clostridium cocleatum, Lachnospiraceae, Muribaculaceae, Odoribacter, and Lactobacillus murinus were among the bacteria that were most extensively reduced, whereas there was increased abundance of Akkermansiaceae and an uncultured bacterium of the Clostridiales vadin BB60 group in the H1S group (Fig. 7c). Together, these results illustrated that fructose intake could profoundly aggravate the effects of restraint stress on the taxonomic composition of the gut microbiota.
High fructose intake induced changes in colonic microbiota metabolites stimulated under restraint stress. We further analyzed the changes in microbiota metabolites after stimulation with high fructose and restraint stress. Analysis of the metabolites using an untargeted metabolomics approach showed that there were 2,575 metabolites in the colon. First, cluster heat map analysis of all colonic microbiota metabolites sequenced in the C, S, H, and H1S groups showed obvious differences among the groups, especially after the addition of high fructose (Fig. 8a). Through correlation analysis between samples, it was found that there was a significant correlation between the samples within a group, while the correlation between the samples between groups was not significant, especially when the H1S group was compared with the other groups (Fig. 8b). A Venn diagram showed that different treatments resulted in different metabolite changes, with an additional 115 metabolites increased in the high-fructose group compared with the control group (Fig. 8c). The beta diversity analysis showed an obvious clustering of microbiota metabolite composition in between C and H1S group ( Fig. 8d and e). The above results were confirmed again by the cluster heat map analysis of C and H1S groups (Fig. 8f). This suggested that the composition of microbiota metabolites was significantly altered after stimulation with high fructose and restraint stress. To determine how changes in gut microbiota metabolites affect host signaling pathways, we classified the annotated results for differential metabolites in the C versus the H1S group according to pathway type in KEGG. The results showed that bile secretion and tryptophan metabolism pathways were significantly enriched (Fig. 8g). Cluster heat map analysis of the 40 metabolites differentially changed in the C and H1S group (Fig. 8h) revealed that the contents of cholic acid, sodium deoxycholate, taurine, and indoleacetaldehyde ( Fig. 8i) were significantly decreased, while the contents of histamine, kynurenic acid, and acetic acid were significantly increased. It could be speculated that high fructose and restraint stress led to changes in metabolites, which in turn affected certain pathways and functions, ultimately leading to intestinal mucosal barrier disruption.  Taurine promoted expression of MUC2, tight junction, NLRP6, and autophagy proteins in HT-29 cells, while histamine inhibited the expression of NLRP6. To further understand the effect of microbiota-modulated metabolites on the intestinal mucosal barrier, we hypothesized that the changes in the abundance of histamine and taurine affect regulation of MUC2, tight junction protein, NLRP6, and autophagy protein expression. Different concentrations of taurine were added to the medium and incubated with HT-29 cell monolayers. Interestingly, taurine supplementation significantly increased the expression of the tight junction protein claudin-1 (P , 0.001) and occludin (P , 0.001) ( Fig. 9a and b) and promoted the expression of NLRP6 (P , 0.001), caspase-1 p20 (P , 0.001), and ASC/TMS1 (P , 0.001) ( Fig. 9a and c) and autophagy proteins ATG16L (P , 0.001), SQSTM1/P62 (P , 0.001), and LC3II/I (P , 0.001) ( Fig. 9a and d) in a dose-dependent manner, compared with the control group. The results of MUC2 immunofluorescence showed that taurine (50 mM and 100 mM) supplementation could promote the expression of MUC2 protein in TH-29 cells ( Fig. 9e and f). The above results indicated that taurine enhanced the integrity of the intestinal mucosal barrier by increasing the expression of tight junction and MUC2 proteins and also significantly increased the expression of NLRP6 and autophagy protein. However, histamine supplementation in HT-29 cells inhibited NLRP6 (P , 0.001), caspase-1 p20 (P , 0.001), and ASC/TMS1 (P , 0.001) expression in a dose-dependent manner, compared with the control group ( Fig. 9g and h). This suggested that high fructose-and restraint stress-induced intestinal mucosal barrier damage may be affected by changes in taurine and histamine levels.
Intestinal colonization with A. muciniphila attenuated colonic barrier damage from high-fructose and restraint stress stimuli. Next, we further explored the effect of A. muciniphila colonization on intestinal mucosal immunity (Fig. 10a). First, we determined changes in fecal flora abundance; the results showed that the abundance of Bacteroides, Firmicutes, Lactobacillus and A. muciniphila decreased significantly in the antibiotic treatment (ABX) group (P , 0.001) and the H1S group receiving antibiotic treatment (ABX1HS group) (P , 0.001) ( Fig. S3A to D). In the A. muciniphila colonization (AKK) group (P , 0.001) and AKK1HS group (P , 0.001), the abundance of Bacteroides, Firmicutes, and Lactobacillus decreased significantly, while the abundance of A. muciniphila increased significantly (P , 0.001) ( Fig. S3E to H), indicating that the antibiotic treatment and A. muciniphila colonization were successful.
Animals exposed to high fructose and restraint stress did not experience significant body weight (P = 0.122) changes after A. muciniphila colonization. In contrast, the antibiotic-treated mice significantly lost weight (P = 0.002) after being stimulated by high fructose and restraint stress (Fig. 10b). Interestingly, there were no changes in the livers and spleens of the antibiotic-treated and Akkermansia-colonized mice ( Fig. S1C and D). Intestinal permeability was significantly lower in the AKK1HS group (P = 0.023) than the ABX1HS group (Fig. 10c). Serum endotoxin levels were consistent with gut permeability results showing that the endotoxin level in the AKK1HS group (P = 0.022) were significantly lower than in the ABX1HS group (Fig. 10d). AB-PAS staining showed that the number of colonic goblet cells was significantly increased after A. muciniphila colonization, and the staining of mucin in crypts was significantly deepened ( Fig. 10e and g). Similarly, immunohistochemical results of MUC2 showed that mucin secretion capacity in the AKK1HS group was significantly higher than in the ABX1HS group (P = 0.009) ( Fig. 10f and g). The expression of colonic tight junction protein was detected by Western blotting. The results showed that the expression of occludin (P = 0.047), claudin-1 (P = 0.000), and claudin-3 (P = 0.001) in the AKK1HS group was significantly higher than in the ABX1HS group ( Fig. 11a and b). These data showed that high fructose combined with restraint stress stimulation after colonization with A. muciniphila reduced the damage to the intestinal barrier. Intestinal colonization with A. muciniphila reduced the level of apoptosis and restored the level of autophagy and the expression of NLRP6 induced by high fructose and restraint stress. The expression of apoptosis, autophagy, and NLRP6related proteins in the colon was detected by Western blotting. We found that levels of the apoptosis protein c-PARP (P = 0.034), Fas (P = 0.001), Bax (P = 0.003), and c-caspase-3 (19 kDa, P = 0.016; 17 kDa, P = 0.001) in the AKK1HS group were significantly lower than in the ABX1HS group ( Fig. 11c and d). The expression of autophagy protein ATG7 (P = 0.007) and the ATG5-ATG12 complex (P = 0.005) increased significantly while the expression of SQSTM1/P62 (P , 0.001) decreased in the AKK1HS group compared with the ABX1HS group ( Fig. 11e and f). The expression of the inflammasome 6related protein NLRP6 (P , 0.001), caspase-1 (P = 0.017), ASC/TMS1 (P = 0.001), and IL-18 (P = 0.002) in the AKK1HS group was significantly higher than in the ABX1HS group ( Fig. 11g and h). These data demonstrated that A. muciniphila colonization reversed the abnormal levels of colonic apoptosis, autophagy, and NLRP6 induced by high fructose and restraint stress.

DISCUSSION
A number of epidemiological, observational, prospective, and retrospective casecontrol studies have examined the impact of diet as a risk factor for the development of IBD and IBS (25,26). Chronic intake of fructose has been shown to be associated with a loss of tight junction protein in the intestine as well as an increased translocation of bacterial endotoxins from the intestine to the liver (27). It has also been found that intake of high fructose promotes colitis pathogenesis (28). Although it has been predicted that a high-fructose diet is a key contributor to the development of IBD, the exact mechanism remains unclear. This study using an animal model provides compelling evidence that a fructose-rich diet predisposes individuals to the pathogenesis of inflammatory bowel diseases.
In the present study, using a murine restraint model to simulate the occurrence of intestinal inflammation, we showed that stress could alter intestinal host-commensal homeostasis and induce mucosal damage and microbial dysbiosis in the gut. These data are consistent with prior publications on stress-induced disorders of the intestinal microbiota and immune disorders (3,29,30). However, the intake of fructose could aggravate the occurrence and development of these symptoms. Moreover, the present findings showed that high-fructose intake combined with restraint stress could enhance oxidative stress and proinflammatory reactions. Substantial evidence supports the coupling of increased oxidative stress with chronic intestinal inflammation (31).
The colonic mucus layer is an early barrier that pathogenic microorganisms must transit to enter the intestinal lumen (32). When this barrier is breached, it increases susceptibility to pathogens and the risk of inflammatory bowel disease (12,22). We observed that the normal histological composition of colon tissues was destroyed and goblet cell number and mucin secretion capacity were significantly reduced in the H1S group. The mechanism by which host immunity is involved in the microbial interface encompasses IgA and mucous secretion as well as antimicrobial-peptide (AMP) production (33). We found that the expression of antimicrobial peptides and SIgA was significantly reduced in the H1S group. Tight junction protein expression affects the integrity and permeability of intestinal mucosa (34). Therefore, in line with our expectations, excessive fructose ingestion caused weakening of tight junction protein expression, which worsened the stress-induced barrier deterioration. As demonstrated by Todoric et al. (27) and Do et al. (35), excessive fructose intake leads to downregulation of tight-junction proteins, intestinal barrier deterioration, and endotoxemia. In this study, we further analyzed the mechanism of feeding high levels of fructose to mice stimulated by restraint stress affecting intestinal mucosal injury. Previous studies have found that intestinal goblet cells secrete mucus that requires autophagy (36). Several studies have reported changes in granular structure in secretory IEC with autophagy defects, such as goblet cells (37,38). This indicates that autophagy is necessary for goblet cells to maintain normal secretory function. NLRP6 is a key regulator of colon homeostasis (16,39). NLRP6 is expressed primarily in intestinal epithelial cells, including goblet cells, where it has been found to be essential for mucosal selfrenewal, proliferation, and mucus secretion (16,39). Moreover, studies have found that NLRP6-deficient mice are susceptible to Citrobacter infection as a result of altered secretion of goblet cell mucus particles due to impaired autophagy (39). A study by Xiao et al. found that NLRP6 is involved in inflammation and brain damage after cerebral hemorrhage by activating autophagy (40). These results suggested that the NLRP6 inflammasome plays an important role in the regulation of autophagy activation. In our study, colon NLRP6 and autophagy-related proteins were significantly downregulated, and the number of goblet cells and the ability to secrete MUC2 were significantly reduced, which damaged the intestinal mucosal barrier, in the H1S group. These results suggested that NLRP6 expression deficiency may affect the normal autophagy function of goblet cells, leading to the weakened mucin secretion ability. However, the exact mechanisms by which NLRP6 expression regulates autophagy remain an enigma.
Intriguingly, another study reported that normal expression of NLRP6 could maintain intestinal microenvironment homeostasis (15). Therefore, microbial composition was analyzed further in this study. The proportion of pathogenic bacteria of the Clostridiales vadin BB60 group, which is a Gram-positive intestinal pathogen that can cause colitis and diarrhea upon treatment with antibiotics (41), was increased in the H1S group. The abundance of Clostridium cocleatum decreased significantly in our study; according to reports of patients with IBS, the level of Clostridium cocleatum decreases (42). The abundances of Muribaculaceae and Odoribacter, which are butyrate-producing bacteria (43,44), were significantly reduced in the H1S groups; butyric acid can accelerate the repair of intestinal epithelial cell injury and maintain intestinal homeostasis (45). The relative abundances of Lachnospiraceae and Lactobacillus were reduced in the H1S group. Lachnospiraceae has an anti-inflammatory effect, while its reduced content suggests that it is associated with susceptibility to colitis (46). Lactobacillus could induce the expression of anti-inflammatory cytokines and prevent the colonization of pathogens into the intestinal epithelium, thus maintaining intestinal homeostasis (47).
The abundance of A. muciniphila, a member of the Verrucomicrobia, increased significantly through high-fructose feeding. Similar results were observed with dietary simple sugars, which increased A. muciniphila and promoted colitis in mice (28). In addition, the diversity and richness of intestinal flora in NLRP6-deficient C57BL/6 mice were significantly reduced, but the abundance of A. muciniphila was significantly increased (48). Unexpectedly, by colonizing mice with A. muciniphila, we found that stimulation with high fructose and restraint stress did not damage the intestinal barrier, and changes in intestinal permeability and serum endotoxin levels were mainly absent. An explanation for the counterintuitive lack of a direct relationship between serum endotoxin levels and the abundance of Gram-negative bacteria (Akkermansia muciniphila), suggested by Everard et al., is that Akkermansia muciniphila regulates barrier function at different levels of the gut (21). The colonization by A. muciniphila increased the number of mucus-producing goblet cells in the colon, consistent with previous reports (49). At the same time, the expression of tight junction protein was significantly increased. These data were consistent with A. muciniphila colonization

Intestinal Microbiota and Microbiota Metabolites
Microbiology Spectrum increasing goblet cell numbers and upregulating tight junction protein expression in obese mice and mice with alcoholic fatty liver (21,50). Recent studies have shown that extracellular vesicles from A. muciniphila MucT (AmEVs) reduce intestinal permeability by regulating tight junctions in mice (51). Notably, Akkermansia-colonized mice were able to restore intestinal autophagy levels and NLRP6 expression and reduce the occurrence of intestinal apoptosis after stimulation with high fructose and restraint stress. However, whether this was directly regulated by A. muciniphila or through its specific microbiota metabolites remains unclear. Metabolites are thought to be key mediators of host-microbiome communication (52). Here, we observed a significant decrease in cholic acid, indoleacetaldehyde, and taurine in the colons of the H1S group. Studies have found that cholic acid can be produced to indirectly stimulate host production of antimicrobial peptides, such as cathelicidin and angiogenin I (53). Indoleacetaldehyde is a tryptophan metabolite that play a critical role in maintaining the homeostasis of the gut and systematic immunity (54). Recent studies have shown that upregulation of taurine in the gut increases intestinal epithelial integrity by strengthening tight junctions to reduce intestinal leakage and inflammation (55). Other independent studies have also shown that the addition of taurine to drinking water reduces dextran sulfate sodium-induced colitis (56). We found that in vitro, taurine could promote the expression of tight junction protein and the mucin MUC2, confirming that taurine can enhance the integrity of the intestinal mucosal barrier. In addition, taurine was found to be a microbe-dependent positive inflammasome modulator that enhanced NLRP6 inflammasome function (15). Recently, taurine was found to alleviate inflammation caused by Streptococcus uberis by activating autophagy in mammary epithelial cells (57). Interestingly, these observations are consistent with our in vitro results that taurine treatment of HT-29 cells promoted NLRP6 and autophagy protein expression. The abundance of kynurenic acid increased in the H1S group, and it has been found to bind to the aryl hydrocarbon receptor (AhR) to produce the proinflammatory cytokine IL-6 (58). Moreover, the highest increase in histamine levels was in the colons of the H1S group. The microbial metabolite histamine has been found to inhibit NLRP6 expression (15,59), which is consistent with our in vitro results. Therefore, these results suggested that microbe-host interactions may directly affect intestinal mucosal barrier development through microbiota metabolites. However, the exact mechanism by which taurine depletion and histamine increase lead to these dysfunctions and the related signaling pathways remain largely unknown. In conclusion, the intestinal microbiota and microbiota metabolites play an important role in maintaining homeostasis of the intestinal mucosal barrier microenvironment. At the same time, this study provides novel avenues to explore and develop therapeutic options for intestinal diseases in the future.
Conclusion. High-fructose consumption exacerbated alterations in the composition of the intestinal microbiome and microflora metabolites caused by restraint stress, leading to an increase in intestinal permeability and a decrease in NLRP6 and autophagy protein expression, destroying the function of colonic goblet cells, increasing intestinal inflammation, and leading to intestinal mucosal barrier defects. However, these phenomena were reversed when mice were colonized with A. muciniphila, and a possible mechanism is shown Fig. 12. These results show that the intestinal microbiome and microbiota metabolites play an important role in maintaining intestinal mucosal barrier microenvironment homeostasis and that A. muciniphila may be able to be used in the treatment of intestinal diseases.

MATERIALS AND METHODS
Animal treatments. A total of 80 female C57BL/6J mice (5 to 6 weeks old) were purchased from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China), and housed under specific-pathogenfree (SPF) conventional conditions (14 h daylight per cycle). After 1 week of adaptation, the animals were divided for two types of animal experiments. In experiment I (Fig. 1a), 40 mice were randomly assigned to four groups: the control group (C; n = 10), the restraint stress group (S; n = 10), the high-fructose group (H; n = 10), and the group receiving high fructose and undergoing restraint stress (H1S; n = 10). Mice were individually placed into ventilated transparent 50-mL plastic centrifuge tubes to limit their movements for 6 h (from 10:00 to 16:00) for 14 consecutive days in the S and H1S groups. Then, 20% fructose was added to the drinking water of the mice in the H and H1S groups. During the restraint stress period, the mice in the control group and the high-fructose group were not given food and water.
In experiment II (Fig. 8a), 40 mice were randomly assigned to four groups: the antibiotic treatment group (ABX; n = 10), the A. muciniphila colonization group (AKK; n = 10), the group receiving antibiotic treatment and high fructose and undergoing restraint stress (ABX1HS; n = 10), and the group receiving A. muciniphila colonization and high fructose and undergoing restraint stress (AKK1HS; n = 10). Antibiotic treatment was given in the drinking water. The antibiotics were formulated as 1 g/L ampicillin, 100 mg/L gentamicin, 0.5 g/L neomycin, 0.5 g/L vancomycin, and 10 mg/L erythromycin, given continuously in drinking water for 14 days. All antibiotics were obtained from Solarbio Science & Technology Intestinal Microbiota and Microbiota Metabolites Microbiology Spectrum Co., Ltd. (Beijing, China). After antibiotic treatment ended, in the AKK and AKK1HS groups, mice were treated with A. muciniphila by oral gavage at a dose of 1 Â 10 8 CFU/0.2 mL, suspended in sterile anaerobic phosphate-buffered saline (PBS). A. muciniphila was given via intragastric administration before constraint stress. High-fructose and restraint stress treatments were the same as in experiment 1. The control group was orally administered an equal volume of sterile anaerobic PBS containing similar final concentrations of glycerol. At the end of the treatment, the mice were euthanized for cervical dislocation, followed by decapitation to immediately collect trunk blood, intestinal tissues, and colonic content. Intestinal segments were fixed in 4% paraformaldehyde. Tissue samples were rapidly frozen and stored in liquid nitrogen for molecular analyses.

FIG 12
The disruption of the intestinal mucosal barrier induced by high fructose and restraint stress was regulated by intestinal microbiota and microbiota metabolites. Dietary fructose exacerbated the development of restraint stress, a strong change in the composition of the gut microbiota and microbial metabolites, decreasing taurine abundance, increasing histamine abundance, increasing intestinal permeability, decreasing expression levels of tight junction proteins, and limitation of the expression of NLRP6 inflammasomes, thereby affecting the normal autophagy level and disrupting the function of colonic goblet cells to secrete mucus, leading to the destruction of the intestinal mucosal immune barrier, activating the NF-k B signaling pathway and inducing an intestinal inflammatory response. However, A. muciniphila supplementation counteracted damage to the intestinal mucosal barrier caused by high fructose and restraint stress. The colonization of A. muciniphila could increase the expression of colonic tight junction protein, restore the function of NLRP6 inflammasomes, activate autophagy, and promote the ability of goblet cells to secrete mucin, thereby protecting the intestinal mucosal barrier and reducing the occurrence of intestinal inflammation.

Intestinal Microbiota and Microbiota Metabolites Microbiology Spectrum
Ethics approval and consent to participate. All animal experiments were approved by the Institutional Animal Care and Use Committee of the China Agricultural University, Beijing, China, under permit no. AW11011202-2-1 (Beijing, China). In this study, all experimental methods were performed following the China Agricultural University of Health Guide for the Care and Use of Laboratory Animals.
Endotoxin levels. Endotoxin levels in mouse blood were determined using an ELISA-based method (CK-E20316; Laibo Terui Technology Development Co., Ltd., Beijing, China) per the manufacturer's instructions. The concentrations of endotoxin were expressed in nanograms per milliliter of plasma.
Blood glucose. Blood glucose levels in the tail blood were measured by a Roche blood glucose meter (GC14906201; Accu-Chek Active; Roche, Germany) before sacrifice.
Measurement of serum CORT and NE. The total CORT and NE concentrations were detected by using a mouse CORT and NE radioimmunoassay kit (HY-114; HY-169; Laibo Terui Technology Development Co., Ltd., Beijing, China). All tests were performed according to the kit's instructions.
Intestinal permeability. An in vivo permeability test was performed using the fluorescein isothiocyanate (FITC)-labeled dextran method to evaluate the barrier function. Food and water were withdrawn overnight, and mice were subjected to gavage with 60 mg FITC-labeled dextran per 100 g body weight (46944; Sigma-Aldrich Co., Ltd., Shanghai, China). Serum was collected 5 h after FITC-dextran-4 (FD-4) gavage, and the fluorescence intensity of each sample was measured (excitation, 492 nm; emission, 525 nm). FITC-dextran was diluted in phosphate-buffered saline (PBS) to form a standard curve. FITCdextran concentration in serum was calculated using a standard curve.
SIgA measurement. Intestinal tissue and normal saline were ground in a ratio of 1:9, and then the supernatant was taken by centrifugation. SIgA concentrations were measured using competitive ELISA kits (CK-E20416; Laibo Terui Technology Development Co., Ltd., Beijing, China). The concentrations of SIgA were expressed in nanograms per milliliter. All assays were performed according to the kit manufacturer's instructions.
Quantitative real-time PCR. Total RNA was extracted with TRIzol reagent (CW0580; CoWin Biotech Co., Inc., Beijing, China). The tissue samples were ground to powder form with liquid nitrogen, transferred to an RNase-free centrifuge tube containing TRIzol, and mixed by shock. Samples were let stand at room temperature for 5 min to allow the RNA to be fully released. Chloroform was added, and mixtures were shaken vigorously and left at room temperature. Samples were then centrifuged, and the supernatant was removed. The same volume of isopropyl alcohol was added, and mixtures were left at room temperature. Mixtures were centrifuged; the supernatant was discarded, and the precipitate was kept. Ethanol (75%) was added and mixed by rotation or reverse mixing; mixtures were then centrifuged, and the supernatant was discarded. The precipitate was left at room temperature and dried with ethanol, and an appropriate amount of RNase-free water was added to dissolve RNA. Mixtures were again let sit at room temperature until the RNA was completely dissolved. The concentration and purity of RNA were measured with a NanoPhotometer (P330; Implen, Germany). Then, cDNA was synthesized with HiScript QRTsupermix for quantitative qPCR (1gDNA wiper) (R312-02; Vazyme Biotech Co., Ltd., Naijing China). Real-time qPCR was performed with SYBR green master mix (Q141-02; Vazyme Biotech Co., Ltd., Naijing China). Changes in fluorescence were monitored on a OneStep Plus instrument (Applied Biosystems, USA). The primers used are shown in Table 1.
Microbial DNA extraction and full-length 16S rRNA gene sequencing. The bacterial genomic DNA was isolated from frozen colon contents according to the manufacturer's instructions by using a PowerSoil DNA isolation kit (MoBio, Shanghai, China). The purity and concentration of the obtained DNA were determined by means of a Nanodrop 1000 instrument (Thermo Fisher Scientific, Wilmington, DE, USA). The preparation of the amplified library uses PCR to amplify the full length of the 16S rRNA. All amplicon libraries were sequenced using a PacBio SMRT platform (Pacific Biosciences, Menlo Park, CA, USA). The bioinformatics analysis of this study was performed with the aid of the BMK Cloud platform (Biomarker Technologies Co., Ltd., Beijing, China).
The UCHIME algorithm (28) (V4.2) was used to detect and remove the chimeric sequence to obtain clean reads. USEARCH (V10.0) was used to cluster sequences with a similarity of 97% into the same operational taxonomic unit (OTU), and the OTUs with abundances of ,0.005% were filtered (24). The abundances of OTUs were normalized with the serial number standard corresponding to the sample with the least sequence, and alpha diversity and beta diversity were further analyzed based on the normalized output data. QIIME (V1.8.0), mothur (V1.3.0), and R software (V3.1) were used for alpha diversity analysis, including sparseness, Shannon curve, and Shannon and Simpson calculations. Beta diversity was calculated by QIIME (V1.8.0) using weighted and unweighted UniFrac distance matrices, including PCA and NMDS heat maps (25). LEfSe analysis was performed to find biomarkers with statistical differences between groups (26). Simply put, LEfSe analysis was done with an LDA threshold of .4, using nonparametric factor Kruskal-Wallis (KW) and rank tests and then using the (unpaired) Wilcoxon rank sum test to identify the most diverse taxa.
Metabolomic analysis. Fecal samples were added to the extract (volume ratio of methanolic acetonitrile to water, 2:2:1; internal standard concentration, 2 mg/L), vortexed, and mixed for 30 s; then, porcelain beads were added, and the mixtures were treated with a 45-Hz grinding instrument for 10 min and subjected to ultrasonication for 10 min (ice water bath). After standing at 220°C for 1 h, the samples were centrifuged at 12,000 rpm for 15 min at 4°C; the supernatant was carefully removed into an Eppendorf micro test tubes (EP) tube, and the extract was dried in a vacuum concentrator. The extract was added to the dried metabolites (acetonitrile-to-water volume ratio, 1:1), redissolved, vortexed for 30 s, ultrasonicated in an ice water bath for 10 min, and centrifuged at 12,000 rpm for 15 min at 4°C. Ten microliters of each sample was taken for detection in the machine. A hybrid quadrupole-Orbitrap mass spectrometer (Q Exactive; Thermo Scientific, Waltham, MA) was coupled to an ultrahigh-performance liquid chromatography (UHPLC) system (Dionex UltiMate 3000; Thermo Scientific) to perform untargeted metabolomics analysis. An Acquity UPLC HSS T3 1.8um 2.1-by 100-mm column from Waters Corporation (Milford, MA, USA) was used for chromatographic separation with 0.1% formic acid in water (solvent A) and with 0.1% formic acid in ethanol (solvent B). MS1 and MS1-dependent MS2 spectra were acquired at a resolution of m/z 37,500. Data were analyzed using Progenesis QI software (Waters), the Human Metabolome Database (HMDB), and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database for metabolite identification; at the same time, the theoretical debris identification was carried out, and the mass number deviation was within 100 ppm.
Measurement of bacterial DNA by real-time PCR. Mouse feces DNA was isolated using a stool DNA kit (D4015; Omega Bio-tek, Norcross, GA) following the manufacturer's instructions. 16S rRNA gene PCRs were monitored on a OneStep Plus instrument (Applied Biosystems, USA), and real-time qPCR was performed with SYBR green master mix (Q141-02; Vazyme Biotech Co., Ltd., Naijing China) according to the manufacturer's directions. Primers specific to 16S rRNA were used as an endogenous control to normalize loading between samples. The relative amount of 16S rRNA genes in each sample was estimated using the DDC T method. Primer sequences were obtained from the Primer Bank primer pairs listed in Table 2.
Luminex liquid suspension chip detection to test the cytokines. Luminex liquid suspension chip detection was carried out by Wayen Biotechnologies (Shanghai, China). The Bio Plex Pro human cytokine group I 23-plex panel was used in accordance with the manufacturer's instructions. The tissue lysate sample was centrifuged to remove the supernatant, and the protein concentration was determined by the BCA method. In brief, the tissue lysate samples and serum samples were incubated in 96-well plates with microbeads embedded for 1 h and then incubated with detection antibody for 30 min. Subsequently, the values were read using the Bio-Plex MAGPIX system (Bio-Rad).
Cell treatment. HT-29 (FS-0269, ATCC) cells were cultured in the Dulbecco's modified Eagle's medium (DMEM) (Vivacell) supplemented with 10% fetal bovine serum (FBS) and penicillin-streptomycin (100 ng/mL) at 37°C in a humidified atmosphere with 5% carbon dioxide. The cells were cultured in 96well culture plates (5 Â 10 5 cells/mL) and 12-well culture plates (5 Â 10 4 cells/mL). Cells were allowed to attach overnight and were then washed twice with PBS and subsequently treated for 24 h in serum-free medium supplemented with different concentration gradients of taurine (HY-B0351; MCE, New Jersey, USA) and histamine (HY-B1204; MCE, New Jersey, USA). Taurine and histamine were dissolved to 1 mM with sterile water and then diluted with medium in a concentration gradient. Finally, the cells were lysed with the lysate, and the total protein was extracted for Western blot analysis.
HT-29 cells immunofluorescence staining. Immunofluorescence was used for detecting MUC2 in HT-29 cells. Cells cultured in 48-well plates were fixed with 4% paraformaldehyde for 30 min. Then, membranes were broken with 0.1% Triton X-100 for 10 min and incubated with 5% bovine serum albumin (BSA) at 37°C for 30 min. Rabbit anti-MUC2 (1:1,000; ab272692; Abcam, Fremont, CA, USA) was added to 48-well cell culture plates, which were incubated overnight at 4°C. Subsequently, HT-29 cells were washed on a decolorization shaker with PBS and then incubated with goat anti-rabbit Alexa Fluor 594 (1:300; ab150080; Abcam, Fremont, CA, USA) at room temperature for 60 min. The nucleus was stained with DAPI (C0065; Solarbio, China) and washed three times with PBS. Cells were observed under a fluorescence microscope, and images were collected. The positively stained puncta were counted using ImageJ software (National Institutes of Health, Bethesda, MD, USA).
TUNEL staining assay. The intestines were washed with PBS, fixed with 4% paraformaldehyde, and embedded in paraffin. The tissue was then sectioned (5 mm), and the sections were dewaxed to water. The formalin-fixed colon sections were subjected to TUNEL staining, using a TUNEL assay kit (T2190; Solarbio, China). Nuclei were stained, and images were acquired using a fluorescence microscope (Nikon Instruments, Inc., Melville, New York). The positively stained puncta were counted using ImageJ software (National Institutes of Health, Bethesda, MD, USA).
Statistical analysis. Data are expressed as means and standard errors. Data analysis was performed with GraphPad Prism 8.0 (GraphPad Software, San Diego, CA). Two-way analysis of variance (ANOVA) with Tukey's post hoc test for differences between groups was used for statistical analysis and multiple comparisons. A P value of ,0.05 was considered statistically significant.
Data availability. All 16S rRNA gene sequencing read data have been deposited in the NCBI Sequence Read Archive (SRA) repository under accession number PRJNA827692. The remaining data are in the supplemental materials.