Gut microbiota-derived 5-hydroxyindole-3-acetic acid mediates the anti-colitis effect of Phellinus gilvus in mice

Yougui Li (  liyougui3@126.com ) Zhejiang Academy of Agricultural Sciences https://orcid.org/0000-0002-8322-6034 Yuqing Sun Zhejiang Academy of Agricultural Sciences Shi Zhong Zhejiang Academy of Agricultural Sciences Jinxi Huo Zhejiang Academy of Agricultural Sciences Le Sun Chinese Academy of Medical Sciences & Peking Union Medical College Fang Zhang Chinese Academy of Medical Sciences & Peking Union Medical College Jianxun Zhu Zhejiang Academy of Agricultural Sciences Chongming Wu Chinese Academy of Medical Sciences & Peking Union Medical College


Introduction
In ammatory bowel disease (IBD) with the two main forms of ulcerative colitis (UC) and Crohn's disease (CD) affects approximately 0.5% population in the developed countries and has increased the incidence in developing countries [1]. The typical symptoms of IBD include a urgent diarrhea, intermittent abdominal pain, rectal bleeding, and weight loss, all of which signi cantly depreciate life quality of affected individuals and increase a risk in developing colon cancer [2]. To date, IBD is not medically curable [3] and treatment options currently available for UC include surgical resection of the affected colon, while treatment of all IBDs includes drug, nutritional, and dietetic therapies [3,4]. A number of studies have shown that altered and imbalanced composition of the gut microbiota or microbial dysbiosis contributes to IBD progression; for example, microbial metabolites, like short-chain fatty acids (SCFAs) and indole derivatives, are the key factors in pathogenesis of colitis [5,6]. Butyrate-producing bacterial species, like Faecalibacterium prausnitzii that belongs to the Clostridium IV class, have been previously reported to associate with IBD development [7], while administration of butyrate was able to effectively alleviate symptoms of UC patients [8]. Furthermore, both animal and patient studies showed that downregulation of the aryl hydrocarbon receptor (AhR) promoted colitis development [9], while AhR agonists possessed a potent effect against colitis through increase in Interleukin (IL)-22 [10]. The gut microbiota-derived indole derivatives were the major ligands to activate AhR, which may indicate microbial metabolites as a potential reservoir to develop critical preventive and therapeutic agents against IBD [4]. Since IBD is a group of in ammatory conditions of the colon and small intestine and the cause are also multifactorial [11], further study of IBD novel molecular mechanisms could lead to the effective approaches in IBC treatment.
Traditional Chinese Medicine (TCM), successfully used in treatment of various human diseases for thousands of years, can modulate composition of the gut microbiota by promoting probiotics and decreasing pathogens to ultimately prevent development and progression of colitis and other diseases [12]. Increasing evidence has recently acknowledged the effects of Phellinus gilvus (also named as mulberry Sanghuang, SH) with anti-in ammation [13,14], anti-tumor [15][16][17], and anti-oxidation [14] properties. SH was also reported to modulate growth, immunity, and fecal microbiota level in pigs [18], while other two species, P. linteus and P. igniarius, have been also reported to effectively ameliorate colitis [19,20]. Nevertheless, the effect of P. gilvus on colitis and the behind mechanisms governing the therapeutic effects of Sanghuang is yet unclear.
In this study, we investigated the anti-colitis effects of SH using a mouse model of colitis in vivo and then explored the underlying molecular events. We assessed the anti-colitis activity of SH and con rmed the necessity of gut microbes in SH's therapeutic effect, then identi ed key species (Alistipes onderdonkii) and its active metabolite (5-hydroxyindoleacetic acid (5HIAA)) that may mediate the colitis-ameliorating action of SH, and speculated the potential molecular mechanism (AhR activation) of the active gut bacterial metabolite. Upon completion of our experiments, we expected to provide insightful information regarding gut microbe-derived 5HIAA in control of colitis.

SH ameliorates DSS-induced colitis
In this study, we rst established a mouse model of colitis (Fig. 1) and orally administrated a low or high dose of SH to the mice. We found that SH signi cantly ameliorated DSS-induced colon colitis in terms of body weight loss, disease activity index (DAI), colon length and colonic histology (Fig. 1a, c-g).
After that, we assessed the role of the gut microbiota in colitis development by established pseudo-germfree mice with daily oral administration of a large dose of antibiotics (ampicillin + nor oxacin, 300 mg/kg) to compare the colitis-mitigating e cacy of SH. We found that administration of antibiotics exhibited a comparable effect on the symptoms of colitis, including improvement of the mouse weight loss, shortened colon length, and colonic histology in DSS + antibiotics vs. DSS-only groups of mice ( Fig. 1b-g). The antibiotics treatment showed a weak effect on improvement of the DAI score (Fig. 1c), expression of the tight junction genes (except Claudin-4 mRNA level) (Fig. 2a, b, c), and goblet cells number (Fig. 2d). However, combination of SH with antibiotics treatment had no any synergic effects on body weight, colon length, colonic histology and tight junction gene transcription in DSS + antibiotics + SH vs. DSS + SH groups of mice ( Fig. 1b-g and Fig. 2a), but had a deteriorated effect on the DAI score, expression of tight junction proteins, and goblet cells number in DSS + antibiotics + SH vs. DSS + SH groups of mice ( Fig. 1c and Fig. 2b, c, d).

Alleviation Of Colon In ammations After Sh Treatment
A previous study reported SH effect on inhibition of in ammation [13]. In this study, we quanti ed level of different cytokines in the colon tissues and sera and found that SH treatment signi cantly reduced level of pro-in ammatory factor mRNAs, including TNFα, IL-1β, IL-6, and IL-17A, and chemokine mRNAs (CXCL-1, CCL2, and MCP-1) but enhanced level of the anti-in ammatory factor mRNAs, such as IL-4, IL-10, and IL-22 in the colon tissues ( Fig. 3a, b, c) or in the sera (Fig. 3d, e). Furthermore, administration of antibiotics also signi cantly decreased level of TNFα mRNA in the colon and improved serum mRNA level of these pro-in ammatory and anti-in ammatory factors except IL-1β. As expected, the combined SH and antibiotics had no additional bene cial effects on regulation of these cytokine expressions (Fig. 3).

Alistipes mimic of SH anti-colitis effect
We next explored the SH effect on the gut microbiota in both conventional and antibiotics-induced pseudo-germ-free mice. Speci cally, we extracted DNA and mRNA samples from the cecal contents and performed both 16S rRNA gene amplicon and 16S rRNA mRNA pyrosequencing to reveal the compositional alterations in total (re ected by gene amplicon sequencing) and alive (re ected by mRNA sequencing) gut bacteria. Our mRNA-based metatranscriptomic analysis showed that SH treatment of mice had a shifted alive gut microbial community vs. the DSS only group, while this shift was disappeared after addition of antibiotics (Fig. 4a, b). The cluster analysis of the main genera revealed that mice treated with SH were clustered with normal animals and departed from DSS only treatment without addition of antibiotics, while the SH + DSS group was clustered with DSS only group after their gut microbes were inhibited by antibiotics (Fig. 4c).
Furthermore, according to the taxon-based analysis, nine genera, including Alistipes, Clostridium IV, and Butyricicoccus, were enriched more than 1.5 folds after SH treatment, whereas co-treatment of mice with antibiotics dramatically decreased or even diminished these bacteria populations (Fig. 4d, e). Our correlation analysis of individual genus and colitis indicators showed that Alistipes and Clostridium IV were signi cantly associated with colon length and the DAI at the top 2 rankings (Fig. 4f). These ndings suggest that SH treatments altered the composition of the intestinal bacterial populations, mainly by elevating the abundance of Alistipes and Clostridium IV, which may be bene cial to inhibiting DSSinduced colitis. These data were reproduced by DNA pyrosequencing (Supplementary Fig. 1), revealing a similar modulation by SH treatment in both total and alive gut microbes.

Alistipes onderdonkii alleviation of DSS-induced colitis
A recent study showed that Alistipes from the gut contains many anti-in ammatory bacteria and its is bene cial in prevention and amelioration of colitis [21,22]. In this study, we found that A. onderdonkii was considerably enriched after SH treatment at the genus and species levels (Fig. 5a, b). Thus, A. onderdonkii may be a key mediator for SH anti-colitis action. To con rm it, we selected three different A. onderdonkii strains to assess their anti-colitis activity in the mouse model of colitis and found that two of these three strains signi cantly improved mouse body weight loss and shortening colon length, DAI index, and colonic histology ( Fig. 5c-g). The adverse effect of DSS on regulation of the serum cytokines was also clearly restored by A. onderdonkii (Fig. 5h), indicating that A. onderdonkii is an effective gut bacterium to possess the anti-colitis effect or mediate the effect of SH. However, the colitis-alleviating effect of A. onderdonkii exhibited obvious species speci city, as oral administration of one A. onderdonkii strain led to more severe colitis ( Supplementary Fig. 2).

5-hydroxyindole-3-acetic Acid (5hiaa) As A Sh-enriched Gut Microbial Metabolite
To further elucidate the mechanisms of SH action to prevent colitis, we carried out a metabolomics analysis of mouse cecal materials to identify the key metabolites of the gut microbiota that associated SH anti-colitis effect. We found that SH treatment signi cantly changed the metabolomics of mouse cecal materials but such a change was substantially deteriorated by antibiotics treatment (Fig. 6a, b).
Speci cally, DSS treatment signi cantly reduced level of six metabolites, i.e., adenine, 2-isopropylmalic acid, 2(1H)-quinolinone, luminchrome, 5-hydroxyindole-3-acetic acid, and equol, whereas SH treatment markedly restored them (Fig. 6c). Our correlation analysis of individual metabolite and pathological indexes of colitis, including the DAI and colon length, revealed that 18 metabolites were associated with colitis remission, among which quinolinone and 5HIAA were the top 2 compounds (Fig. 6d). 5HIAA is a known gut microbial metabolite and a previous study demonstrated that the gut microbes-derived indole-3-acetic acid (IAA), a closely related compound to 5HIAA, possessed an anti-colitis activity [23]. In this study, we found that the IAA level was rather low, but 5HIAA was high; thus, we speculated that 5HIAA may be the effective anti-colitis metabolite. Our correlation analysis of 5HIAA and the gut bacteria showed that 5HIAA level was associated with the abundance of Alistipes and Clostridium IV, suggesting that they could be potential 5HIAA producers (Fig. 6e). In addition, we further quanti ed the level of 5HIAA in the spent culture supernatants of A. onderdonkii using the high-performance liquid chromatography (HPLC), and found that the 5HIAA concentration in A. onderdonkii spent culture medium was 33.5 µg/mL high, indicating that A. onderdonkii is a key producer of 5HIAA ( Supplementary Fig. 3).

5hiaa Suppresses Colitis
To further con rm the effect of 5HIAA on suppression of colitis, we treated the mouse model of colitis with an equal dose of 5HIAA or IAA. Our data showed that oral administration of 5HIAA signi cantly improved mouse body weight loss and shortening colon length ( Fig. 7a, b, c), while the DAI index and colonic pathological damage were also remarkably improved after 5HIAA treatment (Fig. 7d, e). Compared with IAA, 5HIAA was more potent (Fig. 7a-e). 5HIAA also e ciently increased expression of colonic tight junction genes, including occludin, claudin-2, -3, and − 4 mRNAs ( Fig. 7f-h).
We also found the similar anti-in ammatory effect of 5HIAA on colitis as aryl hydrocarbon receptor (AhR) activator indole-3-acetic acid (IAA). A previous study has reported that the indole derivatives ameliorated colitis through activation of the aryl hydrocarbon receptor (AhR) and suppression of pro-in ammatory cytokines/chemokines such as TNF-α, IFN-γ, MCP-1, and IL-17 but promotion of anti-in ammatory factors such as IL-10 and IL-22 5,23 . In this study, 5HIAA had a similar anti-in ammatory effect with IAA on colitis. We found that both 5HIAA and IAA were able to remarkably reduce expression of TNF-α, IL-1β, IL-6, IL-17A, CXCL-1, CCL2, and MCP-1, but increase transcriptions of IL-4, IL-10 and IL-22 in the colon tissues ( Fig. 8a, b, c). Consistently, the dysregulated serum levels of pro-in ammatory and anti-in ammatory factors in DSS-induced colitis mice were signi cantly reversed by 5HIAA and IAA treatment (Fig. 8d, e).
To assessed the effects of SH and 5HIAA on activation of the AhR pathway genes, we found that SH or 5HIAA treatment did not signi cantly induce level of AhR mRNA, but activated the transcription of its downstream genes, including CYP1A1, CYP1A2 and COX2 mRNAs in the mouse abdominal macrophages ( Supplementary Fig. 4), suggesting an activating response of the AhR pathway after administration of SH and 5HIAA to the macrophages. Similarly, SH and 5HIAA treatment was able to slightly enhance level of AhR mRNA in the mouse colons ( Fig. 9a, b) but signi cantly induce transcription of AhR downstream genes CYP1A2 and CYP1B1 mRNAs ( Fig. 9c-f).

Discussion
A great number of previous studies have demonstrated that the gut microbiota or modulation of the gut microbiota associated with development or control of colitis [24][25][26][27][28], although the precise knowledge regarding the speci c bacterial species and the active metabolites that can be used for the clinical management of colitis remains to be de ned. In the current study, we not only demonstrated the key role of the gut microbes and metabolites in mediation of SH anti-colitis effect in a mouse model of colitis, but also identi ed A. onderdonkii as a novel bene cial microorganism that possessed an anti-colitis activity in vivo. We further found that 5HIAA, an AhR activator and produced by A. onderdonkii, is an active metabolite with a more potent anti-colitis effect than indole-3-acetic acid (IAA). Our results revealed that the gut microbiota-indole derivatives-AhR signaling could be used as a novel target in development of next-generational anti-colitis drugs.
Accumulating evidence acknowledged that SH extracts possessed medical usage against in ammation or cancer; for example, the whole-genome sequencing of Phellinus gilvus (mulberry Sanghuang) reveals the unique medicinal values [15], SH impacted growth, immunity, and fecal microbiota in pigs [18].
Moreover, SH possessed antioxidant and anti-in ammatory activities in vitro [14] by prevention of intraperitoneal adhesions and abscesses in a rat peritonitis model [29], while SH extracts suppressed pulmonary in ammation induced by lipopolysaccharide in rats [13]. SH was also able to inhibit melanoma cell growth in mice [16] and induce murine B16-F10 cell arrest at the G0/G1 phase of cell cycle as well as apoptosis [17]. In the current study, we found that SH had an anti-colitis effect in a mouse model of colitis, further con rming the SH anti-in ammation activity. Furthermore, we found that the SH anti-colitis activity was substantially abolished when concomitant with antibiotics. Although antibiotics administration is a common treatment option for colitis and antibiotics alone exhibit potent effect on various colon pathological indexes, such as colon length, histopathology, and cytokines expression, but it showed a quite weak effect on the DAI score, tight junction factors (except Claudin-4 mRNA level) and goblet cell number. Thus, elimination of bacteria alone using antibiotics cannot promote restoration of damaged intestinal epithelial barrier. However, oral SH administration was able to signi cantly improve the gut barrier by enhancing expression of the tight junction genes and restoring mucin-producing goblet cells. These SH effects were associated with presence of the gut microbiota, whereas these bene cial SH effects were substantially diminished after addition of antibiotics that eliminated the gut microbes.Although germ-free animals are still needed to prove the necessity of gut bacteria in the anticolitis action of SH, these results suggest that SH may exert its therapeutic action on colitis via, at least partially, modulation of the gut microbiota.
Indeed, the disruption of and imbalanced composition of the gut microbiota, also called gut dysbiosis, has been recently recognized as a key pathogenic factor in colitis development [30]. In IBD, the gut microbiota was associated with expansion of pro-in ammatory bacteria like Enterobacteriaceae and Fusobacteriaceae and reduction of anti-in ammatory species such as butyrate-producers within the Lachnospiraceae family [31,32]. In the current study, we identi ed four genera, namely Clostridium IV, Alistipes, Butyricicoccus, and Lachnospiraceae_incertae_sedis, to be positively associate with colitis alleviation in mice, all of which are known as SCFA-producing bacteria and have been broadly reported as bene cial microbes in colitis management [33,34]. As a genus in the family of Rikenellaceae, Alistipes, a relatively novel genus of bacteria, are highly relevant in dysbiosis and disease because Alistipes contain enzymes that can degrade cellulose, while loss of cellulose changed the microbial metabolome, gut immune response, gene expression in intestinal epithelial cells and the mice had an increased sensitivity to colitis induction [22,35], while the gut dysbiosis plays a role in determination of the compositional Alistipes abundance in the feces and associated with correlate with human health [22]. Previous studies also showed that colitis was associated with increase in Alistipes abundance [36] and A. negoldii could activate the IL-6/STAT3 signaling to promote colitis-associated cancer development in IL-10 −/− mice [37]. However, more recent studies recognized Alistipes as SCFA-producing anti-in ammatory bacteria and associated with colitis alleviation [21,38]. Dziarski et al demonstrated that oral gavage of A. negoldii was able to effectively attenuate mouse colitis [38]. In our current study, we revealed that the abundance of Alistipes in both total (16S rDNA-based analysis) and active (16S rRNA-based analysis) gut microbial communities was largely reduced in the mouse model of DSS-induced colitis, whereas SH treatment showed little in uence on Alistipes level after assessed the composition of total gut bacteria (alive and dead bacteria) using the 16S rDNA-based metagenomics. However, SH treatment signi cantly increased the abundance of Alistipes after assessment of the active gut microbiota (alive bacteria only) using the 16S rRNA-based metatranscriptomics. Thus, such a contradictory effect of Alistipes on colitis development should be further evaluated and con rmed. Our current results revealed that oral administration of alive A. onderdonkii effectively prevented DSS-induced mouse colitis in vivo, suggesting the bene cial effect of Alistipes spp. on colitis management. Furthermore, we found that the effect of A. onderdonkii on colitis was strain-speci c because another A. onderdonkii strain was lead to more severe colitis in our mouse model of colitis.
In addition, indole compounds are broadly produced by both gram-positive and -negative bacteria [39]. Recent studies demonstrated that indole derivatives of different origins were potent to inhibit colitis, indicating of a tremendous potential in treatment of colitis [5,40]. Our current study revealed that SH treatment signi cantly induced 5HIAA production and 5HIAA treatment was able to signi cantly suppress DSS-induced colitis in mice. Furthermore, it was widely reported that indole compounds, as known ligands for AhR, could effectively suppress level of serum in ammatory cytokine/chemokines, like TNFα, IFN-γ, MCP-1, IL-17 in colitis [41], enhance expression of IL-10 and IL-22 [5,42], and strengthen the intestinal epithelial cell barrier [5]. In our current study, treatment with SH, alive A. onderdonkii, and 5HIAA, all signi cantly suppressed level of pro-in ammatory factors, including IL-1β, IL-6, IL-17, TNFα, and MCP-1, but increased level of anti-in ammatory factors (IL-10, IL-22) and expression of the tight junction genes, e.g., Occludin and Claudins. However, antibiotics treatment only eliminated the serum IL-22 level, which showed a closest correlation with the gut microbes. Although both SCFAs and indoles were reported to be able to enhance IL-10, only indoles can increase IL-22 mRNA [5,43]. Therefore, the increase in IL-22 expression induced by SH may mainly attribute to 5HIAA induction in the mouse model of colitis. Future investigation will assess the underlying molecular mechanism of SH-induced IL-22 expression; for example, whether this regulation is dependent on AhR activation.

Conclusion
Our results demonstrate that modulation of the gut microbiota, especially A. onderdonkii, is a novel mechanism underlying the anti-colitis effect of P gilvus extract. The gut microbiota-indole derivatives-AhR axis plays an important role in host immunity, while 5HIAA will be further evaluated as a promising agent in control of colitis.

SH preparation
The fruiting body of cultured P. gilvus was obtained from Sericultural Research Institute, Zhejiang Academy of Agricultural Science (Hangzhou, China) and naturally dried in the air and then extracted by boiling in water for 2 h. After that, the aqueous extract was concentrated and mixed with three volumes of ethanol. After centrifugation, the supernatant was collected and concentrated using rotary evaporation and lyophilized in a freeze vacuum and then collected, weighed, and stored at -20 o C until use. Our procedure was bale to make one kilogram of P. gilvus fruiting body to approximately 125.7 grams of SH, while the polyphenol content could yield 47.77% as determined by a colorimetric method 15 .

Animals And Experiments
Male C57BL/6 mice of 6-weeks of age were purchased from Shanghai Experimental Animal Center (Shanghai, China) and acclimatized under standard animal care conditions for a week under controlled temperature and humidity and alternating 12-hour light and dark cycles. The mice will receive normal mouse chow and be allowed to drink sterile water ad libitum. To assess the anti-colitis effect of SH in the mouse model of colitis, we randomly divided 80 mice into two main groups, i.e., Group 1 (with no antibiotics supplementation) and Group 2 (with antibiotics supplementation, i.e., ampicillin + nor oxacin, 300 mg/kg/day, + Abs). These two groups of mice were then further divided into four subgroups (n = 10), namely Normal (or Abs), dextran sodium sulfate (DSS), DSS + a low dose of SH (SHL) and DSS + a high dose of SH (SHH). The normal or Abs group was administered an equal volume of the drinking water, the DSS group was established chronic colitis by supplementation of DSS (molecular weight 36-50 kDa) from MP Biologicals (Santa Ana, California, USA) in the drinking water (1.0% w/v) on day 1-5, 8-12, 15-19, 22-26, 29-33, and 36-42. The DSS + SHL and DSS + SHH groups were established colitis and treated with a low or high dose of SH (250 and 400 mg/kg/day, respectively). During the experiments, mouse body weight was measured every 3 days. The disease activity index (DAI) was evaluated on Day 42 (the end of experiments). In the end of the experiment, mice were fasted overnight and blood samples were collected for assessment of different parameters, while the colon tissues were resected and xed in 4% paraformaldehyde for histological analysis or frozen at -80 °C for biochemical measurements.
Furthermore, we also assessed the anti-colitis effect of 5-hydroxyindoleacetic acid (5HIAA) and indoleacetic acid (IAA) in the mouse model of colitis by divided 40 male C57BL/6 mice of 6-weeks of age randomly into four groups (n = 10), i.e., Negative control (NC), DSS-only, DSS + 5HIAA, and DSS + IAA groups. The NC was given an equal volume of the distilled water, while the other three groups of mice were rst established chronic colitis with DSS supplementation in the drinking water (1.0% w/v) (see details in the previous paragraph) and the DSS + 5HIAA and DSS + IAA groups were added treatment with 5HIAA (5 mg/kg) and IAA (5 mg/kg) orally, respectively. During the 42-day experimental period, mouse body weight was measured every 3 days. The disease activity index (DAI) was evaluated on Day 42. In the end of the experiment, mice were fasted overnight and blood samples were collected for analysis and the colon tissues were resected and xed in 4% paraformaldehyde for histological analysis or frozen at -80 °C for biochemical measurements.
In addition, we assessed the anti-colitis effect of Alistipes onderdonkii in mice using three human fecesderived Alistipes onderdonkii strains with different genome structures, namely A. onderdonkii-1, A. onderdonkii-2 and A. onderdonkii-3. However, due to the limited amount of alive A. onderdonkii microbes we could have collected, we were only able to evaluate their anti-colitis effect in the mouse model of acute colitis. In particular, we divided 40 male C57BL/6 mice of 6-weeks of age randomly into four groups (n = 10), i.e., NC, A. onderdonkii-1, A. onderdonkii-2, and A. onderdonkii-3. The NC was given an equal volume of the YCFA bacterial culture medium, while the other three groups were orally gavaged with respective A. onderdonkii strain (10 9 CFU/animal per day). From the fourth day after treatment, all animals were simultaneously given DSS in the drinking water (3.0% w/v) for additional seven days. Mouse body weight and the disease activity index (DAI) were assessed every day and in the end of the experiment, mice were fasted overnight and blood samples were collected for analysis and the colon tissues were resected and frozen at -80 °C for biochemical measurements.
Quantitative Reverse Transcriptase-polymerase Chain Reaction (qrt-pcr) Total cellular RNA was isolated from mouse colonic tissues using the Trizol reagent (Thermo-Fisher, Waltham, MA, USA) and reversely transcribed into cDNA using the Superscript Preampli cation System (Gibco BRL, Gaithersburg, MD, USA) according to the manufacturers' instructions. qPCR was then ampli ed using the SYBR Green PCR mix (Thermo Fisher) according to the manufacturer's protocol in a StepOnePlus real-time PCR system (Applied Biosystems, Foster city, CA, USA). Level of β-actin mRNA was used as the internal control and primer sequences used to amplify each gene are listed in Table S1.

Histology And Immuno uorescence
Length of the entire colons of each mouse was measured and recorded after tissue resections and a segment of the colon tissues from each mouse was taken and xed in 4% paraformaldehyde, embedded in para n, and sectioned into 5 µm-thick sections. Multiple sections from consecutive sections were stained with hematoxylin and eosin (H&E) or Alcian blue for histology assessment under a light microscopy (Nikon Co., Japan). Healthcare Life Sciences) and the relative intensity of protein bands was quanti ed using Image J software (National Institute of Heath, Bethesda, MD, USA).

Metabolomic Analysis
The fecal samples from each mouse were collected, snap-frozen in liquid nitrogen, and stored at -80 °C. Fecal DNA/mRNA extraction, PCR amplication and rDNA/rRNA pyrosequencing at the V4 -V5 regions were performed by Majorbio BioTech Co., Ltd. (Shanghai, China) as previously described [44].
Metabolomics analysis of fecal samples was performed according to a previous study [45]. Brie y, the stool samples were collected and underwent the untargeted metabolomics pro ling using the UHPLC system (Genedenovo, Guangzhou, China).

Statistical analysis
The data were expressed as the mean ± SEM and analyzed using the one-way analysis of variance (ANOVA) test and Dunnett's test (as a post hoc test) to compare differences among groups for pharmacological parameters. For metagenomic and metatranscriptomic analysis, the alpha diversity of the microbiome was calculated based on the operational taxonomic units (OTU) level using the Mothur software (version 1.30.1) [46]. The principal component analysis (PCA) and principal coordinate analysis (PCoA) were performed using the R and visualized by the R package (R Core Team) [47], whose signi cant differences were evaluated by the Adonis analysis in R. The P-values were adjusted for multiple comparisons using the Benjamini and Hochberg. False discovery rate and signi cance was set at q < 0.05. A p-value < 0.05 was considered to be statistically signi cant.

Consent for publication
Not applicable.

Competing interests
The authors declared that there is no con ict of interest in this work.                indicators. e, The correlation between 5-HIAA and genus abundance. The data are the mean ± sem (n=10). *P<0.05, **P<0.01, and ***P<0.001 using one-way ANOVA with Dunnett's test as a post hoc test. indicators. e, The correlation between 5-HIAA and genus abundance. The data are the mean ± sem (n=10). *P<0.05, **P<0.01, and ***P<0.001 using one-way ANOVA with Dunnett's test as a post hoc test. indicators. e, The correlation between 5-HIAA and genus abundance. The data are the mean ± sem (n=10). *P<0.05, **P<0.01, and ***P<0.001 using one-way ANOVA with Dunnett's test as a post hoc test.        SH and 5HIAA activation of the AhR signaling in mouse colon or sera. a, c, e, Effect of SH treatment on mRNA levels of AhR (a) and its downstream genes CYP1A2 (c) and CYP1B1 (e). b, d, f, The effect of 5-HIAA on mRNA levels of AhR (b) and its downstream genes CYP1A2 (d) and CYP1B1 (f). The data are the mean ± sem (n=10). *P<0.05 and ***P<0.001 using one-way ANOVA with Dunnett's test as a post hoc test.