Myricetin and its derivative M10, myricetin-3-O-β-d-lactose sodium salt, modify the composition of gut microbiota in mice with chronic ulcerative colitis

Background Previous studies revealed that Myricetin and derivative M10, Myricetin-3-O-β-d-lactose sodium salt, prevented chronic ulcerative colitis (UC) in mice. We investigated whether the inhibitory effects of Myricetin and M10 on UC were associated to the modication of intestinal microbiota. Samples of intestinal microbiota were collected from the ileocecum of UC mice which demonstrated response to the treatment of Myricetin and M10. Gut microbiota was analyzed by 16S rDNA sequencing assay. Results UC model mice demonstrated the increases of Firmicutes and Actinobacteria as compared to healthy control mice. Oral M10 and Myricetin normalized the composition of Firmicutes and Actinobacteria. At genus level, the effect of M10 and Myricetin on ulcerative colitis was strongly associated to the increase of probiotics, such as Akkermansia, and the inhibition of pathogenic microorganisms, such as Ruminococcus and Parabacteroides. Myricetin’s derivative M10 signicantly increased both biosynthesis and degradation activities, resulting to strong improvements of the metabolism of sulfur, pyruvate, steroid biosynthesis and unsaturated fatty acid biosynthesis in gut microenvironment. Conclusions Natural product Myricetin and its derivative M10 could modify the modication of gut microbiota in UC mice. Combined with pharmacologic effects of Myricetin and M10 in these UC mice, we conclued that the effects of Myricetin and M10 on UC were associated to the modication of intestinal microbiota in the environment of chronic ulcerative colitis.


Background
In ammatory bowel diseases (IBD) describe the diseases of chronic in ammation and ulceration of mucosal and submucosal layers [1,2]. Chronic ulcerative colitis (UC), the most common forms of IBD, manifests clinically as contiguous, super cial mucosal in ammation in colon, originating at the rectum and moving toward the proximal colon [3]. Although the etiology of UC remains unknown, several factors including immune dysfunction [4,5], environmental exposures and more recently, compositional changes of gut microbiota (i.e. dysbiosis) [6][7][8][9] have been considered to associate with the development of disease. The dysbiosis of intestinal microbiota in UC patients was described in terms of richness and diversity [10]. Richness delineates the unique number of bacteria present in a microbial community and diversity extends this to account for their relative abundance. In UC patients, the richness and the diversity were reported the reduction as compared to healthy people. These changes of gut microbiota composition indicated the prognostic value, since the declines in both metrics were associated with an increased risk of disease relapse. Therefore, to modify gut microbiota in the patients with dysbiosis have been associated with therapeutic outcomes in the patients with UC and other IBD [11][12][13].
Myricetin, the extensively studied polyphenol, presents abundantly in various fruits and vegetables. Numerous studies showed that Myricetin exhibits diverse pharmacological properties including anti-chronic in ammation, anticarcinogenesis and angiogenesis, and even prevention of cancers [14][15][16]. Recently, Myricetin was found to have the biological activity of modulation gut microbiota in rats [17]. However, Myricetin does not meet the requirements as a therapeutic drug because of substantial limitations, such as complex isolation procedures, poor water solubility (< 100 ng/ml) and low bioavailability [18]. Previously, a series of novel compounds of Myricetin were designed in our group [19]. M10 is Myricetin-3-O-b-D-lactose sodium salt by adding a hydrophilic glycosylation group. M10 was found to have a higher e cacy of anti-UC and UC-induced colorectal tumors than Myricetin [20]. M10 inhibited chronic UC through inhibiting the IL-6 and NF-κB pathways in the mucosa of colorectum [21]. More evidences are needed to demonstrate the biological activity of Myricetin and its derivative M10 in the prevention of chronic UC.
We have noted that changes of gut microbiota play important roles in the development of UC. UC patients were found signi cant reduction of bacterial abundance and diversity in gut microbiota, which can be described as a general pattern for fewer Firmicutes, particularly Clostridial cluster XIVa and IV, and Bacteroidetes, while more Proteobacteria [22,23]. In DSS-induced ulcerative colitis mice, the abundance of bacteria from Firmicutes phylum was signi cantly decreased, while Bacteroides was consequently increased [24]. In our study, Myricetin and M10 by orally demonstrated strong acitvities of prevention UC in mice model [19]. We further analyzed the composition of gut microbiota in the ileocecum of these mice.
Both Myricetin and M10 reduced the proportions of Ruminococcue, Olsenella, Parvibacter. The composition of Akkermansia in the M10-treated mice were more abundant than control mice, and the proportions of Desulfovibrio, Parasutterella, Acetatifactor, Oscillibacter, Barnesiella, Parabacteroides, Flavonifractor, Alistipes, Lactobacillus, Prevotella, Odoribacter and Saccharibacteria_genera_incertae_sedis were prevented. Moreover, M10 increased both biosynthesis and degradation activities to improve gut microenvironment, including sulfur metabolism, pyruvate metabolism, steroid biosynthesis and unsaturated fatty acid biosynthesis. These results provided more evidences of Myricetin and M10 in the prevention of UC through modifying gut microbiota by increasing probiotics and inhibiting pathogens.

Chemicals
Myricetin (purity > 99%) was purchased from Sigma Chemical Co. M10 (purity ≥ 99%) was provided by Dr. Wang in Marine Biomedical Research Institute of Qingdao. Chemical structures of Myricetin and M10 were issued in the references [19,20].

Mice model of UC and drug treatment
Male C57BL/6 mice of 6-8 weeks were purchased from Charles River Laboratories (Beijing, China). Mice were caged under controlled room temperature, humidity and light (12/12h light/dark cycle) and allowed unrestricted access to standard mouse chow and tap water [25].
A total of 70 C57BL/6 mice were adapted to facility for one week, and then were randomly divided into following groups (n = 10 for each group): Group 1 (healthy control group, NEG group): normal control mice (distilled water); Group 2 (UC model group, M group) : DSS-induced colitis model mice (distilled water); Group 3 (Mesalazine-treated group, POS group) : mice exposed to DSS treated with 100 mg/kg of Myricetin; Group 4-6: mice exposed to DSS treated with 25, 50 (M10-treated group, MID group) and 100 mg/kg of M10, respectively; Group 7 (Myricetin-treated group, DMY group) : mice exposed to DSS treated with control drug Mesalazine (100 mg/kg). Mice were received water containing 1.0% DSS (36-50 kD, MP Biomedicals, LLC, Santa Ana, CA) for 7 days, and then given regular water for 14 days, followed by three additional DSS cycles [20]. Mice received either vehicle (CMC-Na) or drugs by gavage three days after DSS. Administrations were performed six times per week for 12 consecutive weeks. 16S rDNA sequencing assay Ampli cation of variable region V3-V4 from 16S rDNA gene was performed using speci c primer with barcode (Forward primer (5'-3'): CCTACGGGRSGCAGCAG (341F); Reverse prime (5'-3'): GGACTACVVGGGTATCTAATC (806R)). All PCR reactions were performed with 2X KAPA HiFi™ HotStart ReadyMix [26]. After 2% agarose gel electrophoresis, the ampli ed products were recycled with Axy Prep DNA gel extraction kit. Library was quali ed by Thermo NanoDrop 2000 Spectrophotometer UV trace and 2% agarose gel electrophoresis and quanti ed by Qubit® dsDNA HS assay kit. According to the amount of sequencing required for each sample, mixing ratio corresponding to predicted sequence libraries. The enriched library was sequenced on an Illumina NovaSeq PE250 platform according to manufacturer's speci cations to

Results
Myricetin and M10 reshaped gut microbiome taxonomic diversity in mice All pharmacological effects of Myricetin and M10 in mice were con rmed and issued in these UC mice [19]. To characterize the composition of gut microbiome in these UC mice, we randomly selected and analyzed eight stool samples of each group. Each stool sample was subjected to 16S rDNA Amplicon sequencing, which was followed by pro ling of microbial community taxonomic composition and potential function. 16S rDNA Amplicon sequencing retrieved an overall number of 1423136 clean reads, which were clustered in 10360 operational taxonomic units (OTUs) according to strict criteria. Goods coverage of all samples was greater than 99.9% (Table 1), indicating that this sequencing re ects the true nature of intestinal ora. Over 99% of sampled sequences belonged to 6 phyla: Firmicutes (38.80 ± 7.83%), Verrucomicrobia (31.20 ± 18.84%), Bacteroidetes (23.19 ± 15.25%), Proteobacteria (4.07 ± 3.71%), Actinobacteria (2.49 ± 0.83%), and Candidatus Saccharibacteria (0.21 ± 0.11%) (Fig. 1A). Bioinformatic analysis of 16S rRNA sequencing data Bioinformatic analysis of bacterial 16S rRNA amplicon data was conducted using a combination of QIIME software and the usearch platform. Paired-end raw sequences were quali ed to get clean reads [27]. After ltration and chimera checking, reads were clustered into operational taxonomic units (OTUs) at 97% identity [28]. Representative sequences for each OTU was screened for further annotation. Species-richness and alpha diversity were analysed by QIIME software [29].
Principal Coordinates Analysis (PCoA) ordination plot on weighted UniFrac distance matrix was used to describe Beta diversity difference. As for LEfSe analysis, linear discriminant analysis (LDA) scores were used to describe the difference among ve groups of mice [30].

Statistical analysis
Statistical analysis was performed using GraphPad Prism 8.0.2 (GraphPad Software, San Diego, CA, USA). All values are expressed as mean ± standard deviation (SD). After checking data for normal distribution and variance homogeneity, continuous data were compared using Mann-Whitney U test, multiple Student t tests or One-way ANOVA. All p values are two-tailed, and p values < 0.05 were considered signi cant (*P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001).
UC model mice are signi cantly increased as compared with healthy mice, whereas Bacteroidetes and Proteobacteria are signi cantly reduced (Fig. 1B). Although there were slight differences in intestinal ora content (  Data show as mean ± SD. N = 8.

M10 and Myricetin normalize alpha diversity of gut microbiota in mice
We further investigated the richness and evenness of gut microbiome among ve groups of mice. Alpha diversity (α diversity) revealed a signi cant change in drug-treatment groups as compared with healthy control mice. All parameters represented alpha diversity in drug-treatment groups were lower than healthy group, indicating that community richness was decreased in all experimental groups. Moreover, M10-treated mice and Myricetin-treated mice demonstrated signi cant difference as compared to UC model mice (Fig. 2). The detailed results are shown in Table 1.

M10 and Myricetin normalize Beta diversity in intestinal microbiota of UC mice
To examine the weighted UniFrac distance, all groups were separated on PCoA plot (P = 0.001), Adonis plot (P = 0.001) and NMDS plot (Fig. 3). It shows that the samples of Mesalazine-treated group have the largest intra-group difference, while other groups have small intra-group difference. At PCoA1 (57.24%; P < 0.001), healthy mice is signi cantly different with other four groups of mice. However, at PCoA2 (23.25%; P = 0.045), chronic UC model mice and Mesalazine-treated mice are different with healthy control mice, while M10-treated mice and Myricetin-treated mice are similar with healthy control mice (Fig. 3A). In Adonis analysis, we found that at PCoA1 (57.24%), healthy control group is signi cantly different with other four groups. However, at PCoA2 (23.25%), M10-treated mice and Myricetin-treated mice are very similar with healthy control mice (Fig. 3B). The results of NMDS analysis (Fig. 3C) are consistent with PCoA and Adonis analysis, indicating that M10 and Myricetin could normalize the gut microenvironment in mice with UC.
Identi cation of signature bacteria as mediators to accelerate the reshape of in ammed colon In the UC mice, oral M10 and Myricetin reshaped gut microbiota. To con rm both statistical and biological taxonomic differences between gut microbiota and de ne biomarker among ve groups mice, the LEfSe algorithm with a logarithmic LDA score cutoff ≥ 2.0 (LDA ≥ 2, P < 0.05) was performed, and a total of 110 biomarkers were found. In order to identify the signature bacteria as mediators which accelerate the repairing of in ammed colon, we adopted stricter screening criteria (LDA ≥ 4) than others. As expected in healthy control mice, species such as Bacteroidetes, Proteobacteria, Erysipelotrichaceae_incertae_sedis, Porphyromonadaceae, Enterobacteriaceae, Deltaproteobacteria, Escherichia_Shigella, Desulfovibrionales, Anaerotruncus, Betaproteobacteria, Sutterellaceae, Parasutterella, Acetatifactor, Streptococcus, Streptococcaceae and Barnesiella exhibited elevated proportions, whereas in M10-treated mice, the abundance of Akkermansia, Verrucomicrobiae, Gemmiger and Clostridium XlVa are signi cantly increased. Mesalazine-treated mice demonstrated increased abundance of Ruminococcaceae. However, UC mice are characterized by increased abundance of Enterobacter, Clostridium, Ruminococcus and Parvibacte (Fig. 5).

M10 and Myricetin changed the metabolic function of gut microbiota
The PICRUSt algorithm (phylogenetic investigation of communities by reconstruction of unobserved states) was used to evaluate functional differences of bacteria in different groups. The differential pathways analysis was performed against KEGG, and analyzed and plotted (Fig. 6)

Discussion
In current study, we aim to determine if the effects of Myricetin and its derivative M10 on UC are associated to their activity of modi cation the richness and diversity of gut microbiota. Previously, M10 demonstrated higher activities of antiulcerative colitis than Myricetin. M10 prevented DSS-induced UC by inhibiting colonic mucosal cell necrosis and remodeling intestinal barrier [21]. M10 possessed higher activity than Myricetin in prevention of UC. Both M10 and Myricetin did not absorb into systemic circuit but played local activities in intestinal tract [20]. We thus speculate that the effects of M10 on UC are associated to its biological role in the modi cation of gut microbiota. Our result demonstrated that M10 and Myricetin affect the gut microenvironment (α-or β-diversity) as compared with UC model mice ( Fig. 2 and   Fig. 3). We have noticed that although the abundance of gut microbiota is lower in M10-treated mice and Myricetin-treated mice as compared with UC mice, the diversity of these mice are more similar with healthy control mice, which means M10 and Myricetin could reshape gut microenvironment in UC mice model ( Supplementary Fig. 1). Current results are consistent with a previous study that chronic UC gut microbiomes exhibit general decreases in taxonomic diversity relative to healthy gut microbiomes [30]. The differentially affected bacteria generally fell into two major categories, Firmicutes and Actinobacteria. The proportions of Firmicutes and Actinobacteria in UC mice are signi cantly increased as compared to healthy control group, while M10 and Myricetin treatment resulted in normalizing the proportion of Firmicutes and Actinobacteria in gut microbiota.
At genuss level, we noted that the protective effect of M10 and Myricetin on DSS-induced ulcerative colitis was associated to the increase of probiotics, such as Akkermansia, and the inhibition of pathogens, such as Ruminococcus and Parabacteroides. Ruminococcus contribute to in ammation through production of in ammatory polysaccharide and are associated with thinning of intestinal mucosa and impairment of intestinal barrier function [31][32][33]. Paraprevotella is a pathogenic bacterium. Indeed, according to previous studies, its abundance is correlated with the incidence of UC and proin ammatory cytokine secretion by colonic epithelial cells [34]. A previous study showed that stachyose could change gut microbiota composition with a higher level of Akkermansia [35]. Akkermansia can protect from liver injury, reinforce gut barrier function, and reduce in ammation [36]. Correlation coe cient analysis demonstrated that Akkermansia abundance was negative correlated with Paraprevotella ( Supplementary Fig. 2). Hence, M10 could deplete the proportion of Ruminococcus and Parabacteroides, resulted to the proportion of Akkermansia, thus leading to alleviating chronic UC. Furthermore, M10 reshaped intestinal barrier, which prevented intestinal ora invasion into blood in UC mice [21].
Overall, UC mice demonstrated the increases of Firmicutes and Actinobacteria as compared to healthy control mice. Oral