The dry root of R. crenulata (Hook.f. et Thoms.) H. Ohba was purchased from Bozhou Materia Medica Market (Bozhou, China) and identified by Professor Zhenzhong Wang from Jiangsu Kanion Pharmaceutical Co., Ltd. (Nanjing, China). Monosaccharide references (mannose, ribose [Rib], rhamnose, glucose, glucuronic acid [GlcA], galactose, galacturonic acid, xylose [Xyl], arabinose and fucose [Fuc]) were purchased from Solarbio (Beijing, China). DSS (M.W 40,000 Da) was obtained from MACKLIN (Shanghai, China). Salicylazosulfapyridine (SASP) was sourced from Shanghai Fuda Pharmaceutical Co., Ltd. (Shanghai, China), while antibodies targeting ZO-1, occludin, and claudin1 were acquired from Abcam (Cambridge, UK). ELISA kits for interleukin-6 (IL-6) were procured from Thermo Fisher (MA, USA), and a myeloperoxidase (MPO) activity assay kit was provided by Wuhan Saipei Biotechnology Co., Ltd. (Wuhan, China).

C57BL/6 male mice, aged 8–10 weeks and weighing 22 ± 2 g, were obtained from the Experimental Animal Center of Hangzhou Medical College (Hangzhou, China), with the license No.: SCXK (Zhe) 2019–0002. All animals were housed in a controlled environment with an ambient temperature of 22–24 °C and relative humidity maintained between 50% and 60%, subjected to a 12-h light/dark cycle, with continuous access to food and water. All animal experimentation protocols received approval from the Institutional Animal Care and Use Committee (IACUC) of Jiangsu Kanion Pharmaceutical Co., Ltd. (Approval NO. 2023101712).

The dry rhizomes of 500 g R. crenulata were coarsely powdered and extracted twice by refluxing with distilled water (1:10, w/v) at 100 °C for 2 h each time. The combined extracts were filtered and centrifuged to remove precipitates. The supernatant was concentrated to an appropriate volume, and ethanol was added in a fourfold volume to precipitate the supernatant, which was then left overnight at 4 C. After centrifugation, the precipitate was redissolved in distilled water, and the solution was separated by the multifunctional membrane experimental equipment RNM-18G (Shandong Bona Biotechnology Group Co., Ltd., Jinan, China) with cut-off of 10 kDa. Subsequently, the solution was passed through D101 macroporous resin to remove pigments and proteins. Finally, the obtained fraction was collected and freeze-dried to yield a light brown crude polysaccharide from R. crenulata.

After a 7-day acclimation period, the mice were randomly allocated to five groups: normal control (NC), model control (MC), salicylazosulfapyridine (SASP), low-dose RCP (RCPL), and high-dose RCP (RCPH), with 10 animals per group. While the NC group received regular water, the other groups were given 2.5% DSS in their drinking water for 7 days to induce an UC model, with the DSS solution replaced on days first, third, and fifth. As shown in Figure 1, during the modeling process, medication was administered simultaneously. The RCPL and RCPH groups received RCP doses of 100 mg/kg and 200 mg/kg, respectively, while the SASP group was given 100 mg/kg. The NC and MC groups were given pure water. All treatments were administered via gastric lavage for 8 days.

During the experiment, the body weights of the mice were recorded daily, and their fecal morphology and presence of blood in feces were observed and scored according to specific criteria, with reference to the standards outlined by Xu et al. (2020).

After 8 days of administration, mice were fasted for 12 h and allowed to drink water freely. Perform orbital vein blood collection on mice at day ninth. Blood samples were collected and then centrifuged at 3000 rpm for 15 min to separate serum, which was subsequently stored at −80°C. Following blood collection, the mice were euthanized via cervical dislocation. Colon length from cecum to anus was measured and recorded. Colonic and cecal contents were harvested and frozen at −80°C for later analysis of microbiota and SCFAs. Colon tissues were rinsed in PBS and divided into three sections: Distal sections were fixed in 4% paraformaldehyde for histological examination with H&E staining, while the other two sections were kept at −80°C for further experiments.

IL-6 Cytokine level in blood serum was assayed using enzyme-linked immunosorbent assay (ELISA). Additionally, colon tissues were homogenized in a pre-cooled PBS solution. The level of MPO in the colon homogenate was measured in accordance with the manufacturer’s instructions.

The expression of TJ-associated proteins in colon was assessed using immunohistochemical analysis (Cui et al., 2021). Panoramic scanning images were captured through NanoZoomer-S210 slice scanner and read using NDP software. The results were quantified by ImageJ software.

Samples of mouse colonic contents were submitted to Sanshu Biotechnology Co., Ltd. (Shanghai, China) for 16S rRNA gene sequencing analysis. Quality control of the experimental results, Operational Taxonomic Units (OTUs) clustering, and annotation were performed.

An exact amount of colon content was meticulously weighed and subsequently treated with the addition of 50 μL of a 30% phosphoric acid solution, immediately followed by the introduction of 300 μL of acetone solution. This admixture was thoroughly homogenized for a duration of 3 min, thereafter subjected to centrifugation at 12,000 rpm for 10 min,and the supernatant was collected. Depending on the actual situation, the supernatant was diluted several times before the content of SCFAs was determined using gas chromatography (Agilent 7820, USA) coupled with a quadrupole mass spectrometry detection system (Agilent 5977, USA).

All statistical evaluations were conducted using GraphPad Prism 9. Analysis of variance (ANOVA) was applied, with the data expressed as mean ± standard deviation (SD). Statistical significance was set at p < 0.05, while p < 0.01 indicated highly significant differences.

The proportion of polysaccharides in RCP purified from the water extract of R. crenulata was 58.15% ± 0.016%, the standard curve was y = 9.3195x + 0.0281, R ² = 0.9915. There is a peak detected by MALLS-RI, indicating that the molecular weight of RCP is relatively concentrated in aqueous solution. The calculated number average molecular weight (Mn) of RCP is 20.834 kDa, weight average molecular weight (Mw) is 67.848 kDa, peak molecular weight (Mp) is 27.942 kDa, and polydispersity index (Mw/Mn) is 3.257. The monosaccharide composition was assessed using pre-column PMP derivatization followed by HPLC (Figure 2B), revealing that RCP comprised Man, Rha, GlcA, GalA, Glc, Gal, and Ara, with a relative molar ratio of 0.22: 1: 0.07: 7.03: 2.88: 0.64: 4.12, indicating that RCP was a type of heteropolysaccharide. Among RCP, the content of GalA was the highest, followed by Ara.

The body weight of mice experienced a notable decrease after DSS induction. However, the trend of weight loss was significantly reversed through RCP treatment (Figure 3A). DSS could also lead to a substantial increase in DAI, whereas, RCPL and RCPH treatments effectively mitigated the elevation of DAI, and this improvement was positively correlated with the administered dose (Figure 3B). To more accurately assess the severity of colitis and its impact on the mice, we specifically measured the colon length of mice in each group (Figure 3C). Experimental results revealed that in the MC group, the colon length of mice was significantly shorter than that of the NC group, visually reflecting the colonic tissue damage caused by colitis. Nevertheless, in the groups treated with RCPL and RCPH, the shortening of colon length was alleviated to varying degrees, demonstrating that RCP treatment significantly improves the colon shortening induced by colitis (Figure 3D).

HE staining results of the colon are shown in Figure 4. In the NC group, the mice had an intact intestinal mucosal layer, abundant intestinal glands with compact arrangement, normal morphology of epithelial cells, abundant goblet cells, normal morphology of the muscle layer, and no obvious tissue abnormalities. In the MC group, the mice showed colonic mucosal epithelial necrosis (blue arrows), crypts forming ulcers (yellow arrows), goblet cell reduction and replacement with proliferation of connective tissue (red arrows), more inflammatory cell infiltration (black arrows). In the SASP mice, small-scale crypt loss (yellow arrows), intestinal dilation, and glandular epithelium thinning were observed (green arrow). In the RCPL group, the lesions were significantly reduced compared with the MC group. The colonic mucosal epithelial integrity of the mice was relatively intact (blue arrows), inflammation was reduced (black arrow), and a small number of intestinal gland cavities were slightly dilated (green arrow). In the RCPH group, the mice exhibited intact mucosa (blue arrows), mild inflammation (black arrow), and restored crypt structure (yellow arrows), the lesions were significantly reduced compared with the MC group.

The results indicate that the DSS-induced mouse model of UC is successful and can be used to study the interventional effect of RCP on UC. RCP significantly protects the integrity of colonic mucosa and intestinal glandular cells, prevents abnormal connective tissue proliferation, and reduces inflammation. The intensity of its effects is positively correlated with the oral dose.

As shown in Figure 5A, a significant increase in the protein expression of the IL-6 pro-inflammatory cytokine was observed in the MC group. However, RCP significantly inhibited this elevation in expression. We determined the expressions of MPO in the colon homogenate of each group using ELISA. In Figure 5B, it was observed that the MC group had higher concentrations of MPO, while the RCP treatment groups could decrease the activity of MPO. This indicates that RCP could alleviate the inflammatory response in UC, enabling it to return to a normal state gradually.

Immunohistochemical analysis revealed a reduction in the expression of TJ proteins such as occludin, claudin 1, and ZO-1 in the MC group, indicating impaired intestinal barrier function. In contrast, RCP treatment notably enhanced the expression of TJ proteins compared to DSS-induced colitis (Figure 6). This result is consistent with the HE staining result. Therefore, RCP could exert a therapeutic effect on UC by repairing the intestinal barrier through increased expression of TJ proteins.

To investigate the potential regulatory role of RCPH on intestinal microbiota, we conducted 16S rRNA sequencing on colonic content samples derived from mice. In our microbiome analysis, we obtained 1,326,461 high-quality 16S rRNA reads, with an average of 73,692 reads per sample (ranging from 34,996 to 151,050). To assess variations in microbial diversity between groups, we aligned sequences to evaluate alpha and beta diversity. Alpha diversity indices, encompassing species richness and evenness, were analyzed. Ace index means the richness of samples, and shannon index means the evenness of samples. DSS significantly reduced the Ace (Figure 7A) and Shannon (Figure 7B) indices, indicating decreased bacterial richness and diversity. The RCPH group exhibited significantly elevated colonic microbiota richness compared to the DSS group. Beta diversity, which measures differences in microbial community composition, was analyzed using principal coordinate analysis (PCoA) and non-metric multidimensional scaling (NMDS), offering insights into intergroup differences. Weighted UniFrac distances were employed for PCoA to assess beta diversity, revealing distinct microbial structures between MC and NC mice (Figure 7C). RCPH treatment in colitis mice significantly altered microbial composition, resembling NC mice more closely. NMDS analysis further corroborated these differences in intestinal flora (Figure 7D), highlighting the efficacy of RCPH in modulating microbial profiles.

We analyzed microbial composition at various taxonomic levels (phylum, family, genus) to elucidate intestinal microbiota specifics. Key findings at the phylum level (Figure 7E) reveal ten major phyla, with Bacteroidetes, Firmicutes, and Proteobacteria dominating (>90% share), and inversely correlating Bacteroidetes abundance with GI inflammation. The model group exhibited a marked increase in Firmicutes and Proteobacteria, while Bacteroidetes levels were notably decreased. RCPH treatment restored the composition of Bacteroidetes phylum and decreased the Firmicutes in MC mice, which was similar to the NC mice (Figures 8A–C). The relative abundance of several families, including Bacteroidaceae, Lachnospiraceae, Erysipelotrichaceae, Tannerellaceae, Ruminococcaceae and Christensenellaceae were significant increased by DSS but reduced in the RCPH treatment group (Figures 8D–I). Another obvious change was observed in Muribaculaceae, MC mice showed a reduction in Muribaculaceae compared to NC mice, although levels of this bacterium were significantly higher in the RCPH group (Figure 8J).

The shifts in gut microbiota at the genus level were presented in Figures 8K–O. The relative abundance of Bacteroides, Faecalibaculum, Lachnospiraceae_unclassified, Parabacteroides and Ruminiclostridium_9 were enriched by MC, but RCPH could inhibit these microorganisms. Using Linear Discriminant Analysis (LDA) Effect Size (LEfSe) analysis, we identified significant genus-level biomarkers that differentiate the various groups. Compared to the NC and MC groups, the RCPH group exhibited significant differences in Oscillibacter, Ruminococcaceae_UCG_014, Ruminiclostridium, and Bifidobacterium (Figure 7H). The results indicate that DSS caused disruption in the intestinal microbial community, and RCP intervention led to changes in the intestinal microbial community that made it more similar to the NC group.

The contents of SCFAs in cecal contents from each group were determined by GC-MS. As illustrated in Figure 9, DSS treatment notably altered the levels of SCFAs, significantly decreasing acetic, isovaleric, and valeric acids, while no significant differences were observed in propionic acid, isobutyric acid, and butyric acid levels between the groups. RCPH therapy reversed this, restoring acetic, propionic, isovaleric, and valeric acids to NC group levels, with butyric acid levels showing a recovery trend. RCPH can restore the stability of the intestinal microecology by modulating the levels of SCFAs.

To explore the relationships among key intestinal bacteria, biochemical markers, and SCFAs, Spearman’s correlation analysis was conducted. The findings revealed that the abundance of Muribaculaceae and Bacteroidetes showed a positive correlation with colon length, structural proteins, and various types of SCFAs, while it was negatively correlated with the DAI score and inflammation (Figure 10). Conversely, the abundances of Firmicutes, Lachnospiraceae, Erysipelotrichaceae, Tannerellaceae, Ruminococcaceae, Faecalibaculum, Lachnospiraceae_unclassified, Parabacteroides, Ruminiclostridium_9, and Ruminiclostridium exhibited a positive correlation with both the DAI score and the inflammation index. Furthermore, the abundances of Firmicutes, Ruminococcaceae, Faecalibaculum, Lachnospiraceae_unclassified, Ruminiclostridium_9 exhibited a negative correlation with colon length and SCFAs. The correlations between Oscillibacter, Ruminococcaceae_UCG_014, Ruminiclostridium, and Bifidobacterium and the tested indicators were not significant, which may be related to polysaccharide degradation or other metabolic functions. Acetic acid and valeric acid exhibited negative correlations with DAI, IL-6, and MPO, while demonstrating positive correlations with the structural proteins occludin, claudin1, and ZO-1. These findings suggested that acetic acid and valeric acid played significant roles in the treatment of UC using RCP.