Akkermansia muciniphila Exerts Strain-Specific Effects on DSS-Induced Ulcerative Colitis in Mice

Akkermansia muciniphila is a commensal bacterium of the gut mucus layer. Although both in vitro and in vivo data have shown that A. muciniphila strains exhibit strain-specific modulation of gut functions, its ability to moderate immunity to ulcerative colitis have not been verified. We selected three isolated human A. muciniphila strains (FSDLZ39M14, FSDLZ36M5 and FSDLZ20M4) and the A. muciniphila type strain ATCC BAA-835 to examine the effects of different A. muciniphila strains on dextran sulfate sodium-induced colitis. All of the A. muciniphila strains were cultured anaerobically in brain heart infusion medium supplemented with 0.25% type II mucin from porcine stomach. To create animal models, colitis was established in C57BL/6 mice which randomly divided into six groups with 10 mice in each group by adding 3% dextran sulfate sodium to drinking water for 7 days. A. muciniphila strains were orally administered to the mice at a dose of 1 × 109 CFU. Only A. muciniphila FSDLZ36M5 exerted significant protection against ulcerative colitis (UC) by increasing the colon length, restoring body weight, decreasing gut permeability and promoting anti-inflammatory cytokine expression. However, the other strains (FSDLZ39M14, ATCC BAA-835 and FSDLZ20M4) failed to provide these effects. Notably, A. muciniphila FSDLZ20M4 showed a tendency to exacerbate inflammation according to several indicators. Gut microbiota sequencing showed that A. muciniphila FSDLZ36M5 supplementation recovered the gut microbiota of mice to a similar state to that of the control group. A comparative genomic analysis demonstrated that the positive effects of A. muciniphila FSDLZ36M5 compared with the FSDLZ20M4 strain may be associated with specific functional genes that are involved in immune defense mechanisms and protein synthesis. Our results verify the efficacy of A. muciniphila in improving UC and provide gene targets for the efficient and rapid screening of A. muciniphila strains with UC-alleviating effects.


INTRODUCTION
Akkermansia muciniphila is a strictly anaerobic gram-negative gut bacterium that uses mucin as its sole energy source (Derrien et al., 2017). This bacterium is a potential candidate for nextgeneration probiotic because it has a positive effect on human metabolic disorders such as diabetes, alcoholic liver disease, obesity and gastrointestinal diseases (Cani and de Vos, 2017;Plovier et al., 2017;Grander et al., 2018;. Increasing numbers of studies are focusing on the relationship between A. muciniphila and intestinal barrier function and host immunity. For example, Png et al. demonstrated that the abundance of A. muciniphila was dramatically decreased in patients with inflammatory bowel disease compared with controls, suggesting that A. muciniphila may be correlated with intestinal mucosal health (Png et al., 2010). Several reports have observed that A. muciniphila is associated with body weight gain, reduction in the expression of pro-inflammatory cytokines and the recovery of gut epithelial barrier function in dextran sulfate sodium (DSS)-induced colitis Zhai et al., 2019). Other studies have shown that supplementation with A. muciniphila can restore the normal thickness of the inner mucus layer and increase the expression of tight-junction (TJ) proteins such as claudins, occludin and zonula occludens-1(ZO-1) in the gut of mice (Everard et al., 2013;Geerlings et al., 2018;Grander et al., 2018). These findings suggest that A. muciniphila plays an important role in protecting the intestinal tract from damage.
Although multiple reports have suggested that probiotics have the potential to alleviate intestinal disease, some studies have demonstrated that the effects of probiotics on intestinal disease are strain specific. For example, Lactobacillus casei ATCC BAA-835 393, DN-114001 and Lbs2 have been demonstrated to alleviate damage in gut disease by restoring histopathological damage, improving the disease activity index (DAI) score, increasing the differentiation of regulatory T (Treg) cells and inhibiting proinflammatory cytokine expression (Zakostelska et al., 2011;Thakur et al., 2016;Xu et al., 2018). However, other reports have revealed that the ingestion of L. casei does not prevent the occurrence of diarrhea or inhibit pathogenic infections (Arena et al., 2017). Moreover, a study found that in vitro A. muciniphila ATCC BAA-835 significantly augmented Treg differentiation, whereas another A. muciniphila strain 139 did not have this effect (Zhai et al., 2019). Another study showed that some murine A. muciniphila strains may aggravate colitis in IL10−/− mice (Seregin et al., 2017). In sum, these findings suggest that different A. muciniphila strains have strain-specific physiological functions.
Studies have shown that confounding factors including the origin, genetic background and physiological characteristics of the strains may influence the functional specificity of probiotics. Different species are able to adjust to specific environments through genome specialization that drive niche-specific adaptations (Frese et al., 2012;Cen et al., 2020;Xiao, 2020). Specific genes in the gut bacteria determine their beneficial function in the intestine. One study conducted a phylogenetic genomic analysis of 100 strains of L. rhamnosus isolated from different sources and found that the differences among the strains were partly related to the differing ecological niches they occupied, resulting in strains with different competitive advantages in different environments that thus exert different probiotic effects (Frese et al., 2012). Furthermore, the intestinal protective effects of probiotic bacterial strains are closely related to their physiological characteristics. A study has claimed that the micro integral membrane proteins of L. plantarum CGMCC 1258 can remedy TJ injury by upregulating the relative expression of TJ proteins such as JAM-1and claudin-1 (Yin et al., 2018). Amuc_1100, a membrane protein secreted by A. muciniphila, can remedy metabolic disorders and enhance the intestinal permeability of mice (Plovier et al., 2017). Extracellular vesicles of A. muciniphila ATCC BAA-835 have been demonstrated to have a positive effect on alleviating DSSinduced colitis in mice (Kang et al., 2013). Furthermore, conjugated linoleic acids and short-chain fatty acids produced by probiotics also improve the gut barrier and regulate gut immune function (Cremon et al., 2018).
Although much evidence has shown that probiotics have a beneficial effect on colitis, the positive effects of A. muciniphila and its ability to moderate immunity to colitis have not been verified. The inter-species or inter-genetic differences of A. muciniphila has been evaluated by several studies, but no definite conclusion has been reached thus far, partly due to the limited number of isolated strains of this species. In fact, most investigations of A. muciniphila have only assessed the type strain ATCC BAA-835. Therefore, we selected A. muciniphila ATCC BAA-835 and three other A. muciniphila strains, FSDLZ36M5, FSDLZ39M14 and FSDLZ20M4, isolated from human feces in our laboratory to investigate the effects of A. muciniphila on DSS-induced ulcerative colitis (UC). The genomic and functional characteristics of the different A. muciniphila strains were used to assess their effects from a genetic perspective.

Animal Experiment Design
Sixty adult male specific pathogen-free C57BL/6 mice were randomly divided into six groups with 10 mice in each group: five of the groups were administered DSS to induce UC (UC groups) and the sixth was a control group. Four of the UC groups were supplemented with one of the strains of A. muciniphila (treatment groups) and the fifth received no supplementation (DSS group). All of the mice were fed standard commercial mouse food and kept in a 12-h light/dark cycle environment at 22-24°C under controlled humidity (40-70%). The experimental protocol was approved by the Ethics Committee of Jiangnan University, China (JN. No.20190915c0801101), and was performed in accordance with European Community Directive 2010/63/EU.
Each mouse in the control group was administered 0.2 ml of sterile phosphate-buffered saline (PBS) daily and distilled water during the 7-day experiment. The other five groups were given 3% (w/v) DSS (36-50 kDa, MP Biomedicals Ltd, Santa Ana, USA) in drinking water to induce UC, and those in the treatment groups were given a sterile PBS suspension of A. muciniphila (1 × 10 9 CFU/0.2 ml per mouse per day). After 7 days of treatment, the mice were anesthetized and sacrificed. The feces, blood and colon tissues of the mice were then harvested for analysis. The six groups were as follows.

Assessment of the Severity of UC
During the experiment, the body weight of each mouse was measured daily and the colon length was measured when the mice were sacrificed to determine the severity of the induced UC (Yang et al., 2020).

Gut Permeability
The fluorescein isothiocyanate-conjugated dextran (FITCdextran) assay obtained from Sigma-Aldrich (Saint Louis, MO, USA) was used to determine the gut permeability of the mice. The concentration of FITC-dextran was determined using a previously described method .

Biochemical Analysis of the Colon Tissue
A colon sample from each mouse was accurately weighed (0.1 g) and homogenized in nine volumes of cold PBS. The supernatant was separated by centrifugation at 3000 g (4°C, 5 min). The mucin 2 (MUC2), interleukin (IL)-6, IL-10, IL-1b and tumor necrosis factor alpha (TNF-a) contents were detected using the corresponding kits from SenBeiJia Biological Technology Ltd. (Nanjing, China).

Fecal DNA Extraction, Sequencing, and Analysis
The bacterial DNA from mouse feces was isolated using a FastDNA ® Spin kit from MP Biomedicals Ltd. (CA, USA). The sequencing of the intestinal microbiota was performed using a previously described method. A principal component analysis (PCA) was used with the STAMP software to analyze the results. The inter-group differences in the intestinal microbiota composition were determined by an LEfSe analysis.

Comparative Genomic Analysis of A. muciniphila
The genome sequencing and annotation of the A. muciniphila isolates were performed following the method given in previous studies. Maximum likelihood phylogenetic tree using the neighbor-joining method was analyzed and constructed using FastTree (Price et al., 2010). Orthologous genes were generated using Roary with the default parameters (Page et al., 2015). COG annotation was done with the COG database using BLAST (Tatusov et al., 2000). The CAZyme database (Lombard et al., 2014) were used to annotate carbohydrate active enzymes.

Statistical Analysis
Experimental data are presented as the mean ± standard error of the mean. A one-way analysis of variance (ANOVA) was used to analyze the data, followed by Tukey's multiple comparison test to identify statistical significance. P values < 0.05 were regarded as statistically significant. All of the statistical analyses were conducted and visualized with GraphPad Prism (GraphPad Software Inc., San Diego, CA, USA).

Phylogenetic Tree of the A. muciniphila Strains
We selected 37 reported A. muciniphila strains (including the type strain ATCC BAA-835) and 3 A. muciniphila strains from our laboratory to analyze the evolutionary history of A. muciniphila, all of which were isolated from fecal samples from Chinese people. The 40 strains shared 598 homologous genes. We found that FSDLZ36M5, FSDLZ20M4 and FSDLZ39M14 occurred on different branches of the phylogenetic tree ( Figure 1). Therefore, we selected the type strain and these three strains with relatively distant genetic relationships to evaluate their ability to alleviate UC.

Effect of A. muciniphila on Body Weight and Colon Length
The addition of DSS to the drinking water of the mice in the UC groups resulted in a significant drop in body weight and a rapid shortening of the colon relative to that of the control group (Figures 2A, B). After supplementation with A. muciniphila FSDLZ36M5, all indicators showed significant improvement compared with those of the untreated DSS group. However, the other strains did not show any obvious alleviating effects, and A. muciniphila FSDLZ20M4 had a tendency to exacerbate the physiological damage to the gut of the mice.

Effect of A. muciniphila on Gut Permeability
DSS caused significant increases in the concentration of FITCdextran in the serum of the mice in the UC groups compared with the control group (P < 0.001, Figure 3). Treatment with different A. muciniphila had different effects on the increased gut permeability caused by DSS. A reduced gut permeability in mice treated with A. muciniphila ATCC BAA-835, FSDLZ36M5 and FSDLZ39M14 was observed relative to the untreated DSS group, and A. muciniphila FSDLZ36M5 in particular contributed to a marked reduction in the levels of FITC-dextran (P < 0.001, Figure 3), indicating enhanced gut barrier function. However, the oral administration of A. muciniphila FSDLZ20M4 tended to aggravate the damage to the intestinal barrier of the mice.

Effect of A. muciniphila on Inflammatory Cytokine Expression in the Colon
DSS led to the substantial expression of inflammatory cytokines in the colon tissue ( Figure 4). An increase in pro-inflammatory cytokines (TNF-a, IL-1b and IL-6) and a decrease in antiinflammatory cytokines (IL-10) were observed in all of the mice administered with DSS compared with the control group ( Figures 4A-D). After treatment with A. muciniphila FSDLZ36M5, all cytokine indicators significantly improved (P < 0.05). Similar results were found for the three other A. muciniphila strains ( Figure 4B). Furthermore, compared with the control group, the MUC2 content in the DSS group was significantly decreased, but the content was recovered in the treated groups, consistent with the above ( Figure 4E).

Effect of A. muciniphila on the Gut Microbiota Composition
The PCA results revealed that the gut microbiota compositions of the UC groups were significantly different from that of the control group ( Figure 5A). Supplementation with A. muciniphila strains led to structural alterations in the gut microbial communities of all of the treated groups relative to the gut microbiota of the untreated DSS group. Notably, the gut microbiota of the group supplemented with the A. muciniphila FSDLZ36M5 strain was similar to that of the control group, but the other three strains had no prominent recovery effect ( Figure 5A). At the phylum level, in the DSS group, the abundance of Firmicutes and Proteobacteria significantly increased, whereas that of Verrucomicrobia and Actinobacteria decreased to varying degrees compared with the control group ( Figure 5B). At the genus level, the abundance of Bifidobacterium and Lactobacillus was recovered in the A. muciniphila FSDLZ36M5  group ( Figure 5C) to levels greater than those in the DSS group. Meanwhile, a significant drop in the abundance of Bacteroides and Enterobacteriaceae was observed ( Figure 5C). The LEfSe analysis also revealed similar results ( Figures 6A-C).

Comparative Genomic Analysis of the Selected A. muciniphila Strains
Although there was no significant difference in the cluster of orthologous groups (COGs) and carbohydrate-active enzyme (CAZy) levels between FSDLZ36M5 and FSDLZ20M4 strains ( Figures 7A, B), there were some specific genes in the FSDLZ36M5 strain (Table 1).

DISCUSSION
This study aimed to verify whether the UC-alleviating effects of A. muciniphila are strain specific. Although the evidence of functional differences between different strains of this species is abundant, the factors that contribute to the specificity of these strains remain unclear. The genetic background of the bacterial strains is considered to have an important influence on the functional specificity of probiotics (Ribbera et al., 2013;Luongo et al., 2017;. A previous report revealed that although the genomes of A. muciniphila ATCC BAA-835 and A. muciniphila 139 are highly similar, they contain strain-specific genes that result in a differing ability to relieve gut inflammation and restore intestinal flora in chronic colitis (Zhai et al., 2019). It was also reported that L. kefiranofaciens ZW3 showed a stronger ability to protect the gut barrier than other investigated strains of the same species due to the presence of four specific genes (pgm, ugp, uge and pgi), which were identified to encode enzymes that regulate carbon flux and in turn affect exopolysaccharides yield, thus exerting a positive effect on the host gut microbiota (Xing et al., 2017). Furthermore, a study showed that Bifidobacterium infantis EVC001 had a remarkable fitness advantage over other tested strains due to the presence of an H5 gene cluster associated with the ability to metabolize human milk oligosaccharides and to colonize the infant gut, resulting in the strain having a beneficial function in the gut of infants (Duar et al., 2020). We therefore selected four A. muciniphila strains with relatively different genetic positions on the phylogenetic tree as the test strains to evaluate their ability to alleviate UC. Our results verified that the therapeutic effect of A. muciniphila strains on UC was strain specific. Among the four A. muciniphila strains used in the animal study, A. muciniphila FSDLZ36M5 showed a positive effect on UC symptoms, but the other tested strains failed to show such effects. Two indicators, body weight and colon length, were used to determine the severity of UC. Most previous studies have found increased gut permeability and a thinner colonic mucus layer in mouse and human patients with UC, indicating impairment of the gut barrier that exacerbates UC (Petersson et al., 2011;Dicksved et al., 2012). A. muciniphila is considered to be a mucindegrading bacteria that has the ability to continuously refresh   and reshape the gut mucus layer, thus improving the integrity of the gut epithelial barrier and maintaining intestinal homeostasis (Derrien et al., 2008;Everard et al., 2013). In vivo, we observed a decreased concentration of FITC-dextran in the blood and increased MUC2 content in the colon tissue of mice due to A. muciniphila supplementation, which indicated decreased permeability and restored gut barrier function, in accordance with the results of previous reports .
Furthermore, several studies have reported that UC may be linked with an overreaction of the immune system to the gut microbiota Zhang et al., 2017). Hence, the gut microbial composition of each group in this study was analyzed to investigate the effect of each strain of A. muciniphila on the gut microbiome. Our findings revealed an altered gut microbiota in the mice treated with DSS, with a change in the relative abundance of specific bacterial taxa. After the administration of A. muciniphila FSDLZ36M5, we found changes in the fecal bacteria, such as increased Lactobacillus abundance and decreased Enterobacteriaceae abundance. Specific species in the Lactobacillus genus have been found to protect the gut from inflammatory damage and augment the gut barrier (Simeoli et al., 2015;Zhang et al., 2018;Li et al., 2020;Panpetch et al., 2020). Certain Enterobacteriaceae species are enriched in mice and in human patients with inflammatory bowel disease, and Escherichia species have also been found in ileal biopsies of patients with UC (Meng et al., 2018). Notably, augmented Escherichia abundance may aggravate disease severity by increasing gut permeability. We observed similar findings. In previous studies, most of the regulatory effects of A. muciniphila on UC were mainly found with the type strain ATCC BAA-835, with some studies showing that gavage of A. muciniphila strain ATCC BAA-835 effectively ameliorated the adverse effects of DSS-induced colitis . However, in our study, ATCC BAA-835 did not seem to have a significant therapeutic effect. This result is consistent with that of a previous study by Kang et al. (2013). We speculate that the number of strains, the phase of colitis in mice and sex of the mice may influence the effects of the A. muciniphila strain ATCC BAA-835. In our experiment, the bacterial dose in the treated groups was 1 × 10 9 CFU/0.2 ml, whereas in previous studies, the dose was 1 × 10 9 CFU/ml, which may explain the differences in the UC-alleviating effects in this study . A previous study also reported that different doses of probiotics had significantly different effects on reducing the colonization of Salmonella enterica serovar Typhimurium in the intestine, with high doses resulting in a significant decrease in S. enterica serovar Typhimurium counts in the intestine relative to low doses, thus protecting gut barrier function (Haghighi et al., 2008). Further, the pathological characteristics of DSS-induced acute UC and chronic UC are not the same, so the performance of bacterial strains in these two models will also be different (Persě 2012). Moreover, sex of the mouse models has been revealed to affect the pathogenicity of DSS-induced colitis (Wagnerova et al., 2017). Due to the different colonization ability of intestinal microorganisms in the gastrointestinal tract, their beneficial effects could be affected. Finally, the potential genetic background of A. muciniphila isolates can be explained by comparative genomic analysis. Nineteen strain-specific COG categories were detected in the strain that showed significant protective effects against colitis (A. muciniphila FSDLZ36M5) but not in the strain that showed negative effects (A. muciniphila FSDLZ20M4; Table 1).

A B
FIGURE 7 | Clusters of orthologous groups (COGs) repertoires (A) and carbohydrate-active enzyme (CAZy) families (B) of A. muciniphila FSDLZ36M5 and A. muciniphila FSDLZ20M4. Among these genes, res, cpt, and hsdR have been reported to be involved in immune defense mechanisms that help bacterial strains develop self-defense in harsh environments, thus giving them a colonization advantage (Kholodii et al., 1995;Ferri et al., 2010;Morash et al., 2010). A previous study found that certain adverse conditions can induce the expression of several key stress and resistance genes/proteins in bacteria, including chloramphenicol phosphotransferase encoded by cpt, which can help bacteria colonize and develop resistance in adverse ecological environments (Hassan et al., 2019). It has been reported that the adaptability of probiotics to environmental conditions in the gastrointestinal tract is important to their residence time and survival rate (Saulnier et al., 2011). This property is consistent with probiotic performance in DSS-induced colitis. The gene metE detected in A. muciniphila FSDLZ36M5 is responsible for methionine formation, an essential amino acid that supports protein synthesis and the regulation of the gut mucosal immune response and barrier function (Bauchart-Thevret et al., 2009). This specific gene is also present in Lactobacillus helveticus MTCC 5463, and can protect against oxidative stress induced by macrophages and prevent intestinal mucosal cell damage (Senan et al., 2015). In agreement with this finding, supplementation with A. muciniphila FSDLZ36M5 in this study significantly increased the MUC2 content and decreased the expression of inflammatory cytokines compared with the other test strains (P < 0.05, Figure 4D). Our results suggest that the integration of key phenotypic and genotypic characteristics of strains may be an effective method to screen for strains that are adapted to specific functions that could be clinically exploited. In addition, it has been reported selective colonization ability of human fecal microbes in different mouse gut environments.

CONCLUSIONS
Overall, our study confirms that the beneficial effects of A. muciniphila in DSS-induced UC are strain specific. Among the four A. muciniphila strains used in the animal study, A. muciniphila FSDLZ36M5 showed a positive effect on UC symptoms, but the other tested strains failed to show such effects. Firstly, on physiological indicators including colon length and body weight, A. muciniphila FSDLZ36M5 showed significant improvement compared with those of the untreated DSS group. Besides, In vivo, we observed a decreased concentration of FITC-dextran in the blood and increased MUC2 content in the colon tissue of mice due to A. muciniphila FSDLZ36M5 supplementation, which indicated decreased permeability and restored gut barrier function. Moreover, after treatment with A. muciniphila FSDLZ36M5, all cytokine indicators (TNF-a, IL-1b and IL-6) significantly improved. Furthermore, notably, the gut microbiota of the group supplemented with the A. muciniphila FSDLZ36M5 strain was similar to that of the control group. We have added related contents in the Discussion section. In addition, comparative genome analysis detected a series of specific genes related to immune defense mechanisms and methionine transport and metabolism that have been reported to regulate the mucosal immune response and protect the intestinal epithelial barrier, and thus play a positive role in alleviating UC.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi. nlm.nih.gov/, PRJNA714373.

ETHICS STATEMENT
The animal study was reviewed and approved by the