This work acquired POG (B21157–20 mg) (purity>98%) in Shanghai Yuanye Bio-Technology Co., Ltd. (Shanghai, China). DMEM (SH30285. FS), 0.05% trypsin (SH30236.01), PBS solution (SH30256.01), fetal bovine serum (SV30087.03), penicillin, and streptomycin (SV30087.03) were purchased from HyClone Laboratories (UT, United States). We acquired TRIzol (93289), DMSO (D2650), LPS (L-2880), Evans Blue (E2129), and formamide (F7503) from Sigma-Aldrich (Saint Louis, MO, United States). Primary antibodies, including ERK1/2 (9102), p-ERK1/2 (9101), p-AKT (4060), AKT (4691), p38 (9212), p-P38 (9211), NF-κB p65 (4764), NF-κB p-p65 (3033), NF-κB IκB-α (L35A5), NF-κB p-IκB-α (ser32) JNK (9252), p-JNK (4668), β-actin (3700), COX-2 (ab62331), iNOS (ab178945), ZO-1 (D6L1E), Occludin (ab216327), and Claudin-3 (ab15102) were provided by Abcam (Cambridge, United Kingdom) or Cell Signaling Technology (MA, United States). In addition, the goat anti-mouse/rabbit secondary antibody was obtained from Boster Biological Technology (CA, United States). This work acquired ELISA kits for TNF-α (430907), IL-6 (431307), and IL-1β (432604) from Biolegend (San Diego, CA92121, United States). Reverse transcription kits (OligdT, RRI, dNTP Mix 10 mM, MLV, MLV Buffer) and 2 × SYBR Premix were provided by Takara Biomedical Technology Co., Ltd. (Kyoto, Japan).

In order to carry out the experiment, POG was dissolved in purified water and stored. The safe dosage was screened before the experiment, and it was found that 2.5, 5, and 10 mg/kg were safe for mice. Therefore, we chose 2.5/5/10 mg/kg of POG for animal experiments, and 12.5, 25, and 50 μmol/ml of POG were selected for the cell line studies.

Each animal experiment was carried out according to the legal regulations. The study protocols were approved by the Institutional Animal Management and Use Committee of Jilin University (Changchun, China) as per the protocol (Permit Number: SY202201008). This work obtained 6–8-week-old male C57BL/6 mice applied from Liaoning Changsheng Biotechnology Co., Ltd. In the experiment, 36 mice were randomized and grouped into six groups. Animals were divided into DSS, POG, control, DSS + POG (2.5, 5, and 10 mg/kg) groups, with six each. Three days before the experiment, all animals had free access to water and food. POG and DSS + POG groups (2.5, 5, and 10 mg/kg) were given corresponding doses of POG by intraperitoneal injection. After 3 days, all the groups were free to eat, the control group and POG group were free to drink, and the other groups were given 2% DSS (2 g added with 100 ml purified water). The POG and DSS + POG groups (2.5, 5, and 10 mg/kg) were given the corresponding dose of POG by intraperitoneal injection. Mice were euthanized on the 10th day after successful modeling for seven consecutive days. The specific experiments are as follows.

This work cultivated RAW264.7 cells within DMEM that contained 10% fetal bovine serum (FBS) under 37°C and 5% carbon dioxide (CO₂) conditions. The suitable concentration of the cell drug was selected by CCK8 assay. After counting, cells were inoculated into 96-well plates with 100 µL per well at the logarithmic growth stage. LPS stimulation was added for 2 h after pre-administration for 1 h, and drug treatment for 24 h, then the supernatant was discarded, and all 110 µL CCK8 dilution (Beyotime Inst Biotech, Beijing, China) wells were added to the wells. Following washing, OD _(450nm) values were detected after 3 h.

LPS and drugs were used to treat RAW264.7 cells to extract their mRNA, and cellular IL-1β, TNF-α, and IL-6 levels were analyzed by the qRT-PCR method according to the prior description (Hao et al., 2020). Table 1 displays all substrates utilized.

The body weight (BW), fecal occult blood and stool character, and DAI were determined using a scoring system (Table 2) according to the previous description (Tian et al., 2016).

The expression levels of inflammatory factors TNF-α, IL-1β, and IL-6 in the colon tissues of mice were measured as per the instructions given on the Elisa kit.

The mice were weighed and scored clinically according to the scoring criteria in Table 2. Finally, the colon and feces of the mice were collected for relevant experimental analysis.

After collecting mice colonic samples, they were fixed with 4% formaldehyde, paraffin-embedded tissue, followed by slicing into 5 µm sections and H&E staining. Histology was observed by an optical microscope after staining. Finally, histological scores were performed according to Table 3. The specific methods were referred to from the reported method (Zhang et al., 2021).

About 5 µm tissue sections were dewaxed and dehydrated with ethanol gradient, antigen retrieval, and rinsed by PBS for 5 min thrice. 5% donkey serum resulted in blocked sections for 30 min. The primary antibody [ZO-1 (1:100), Occludin (1:100), and Claudin-3 (1:100)] was supplemented dropwise and kept at 4°C in the refrigerator overnight. It was washed with PBS 3 times (5 min each time). After 1 h at room temperature, binding to goat anti-rabbit IgG antibody (1:2000, Santa Cruz), cells were washed thrice and stained with DAPI.

This work extracted total colonic tissue protein using RIPA lysate after grinding. The protein concentration of mice colon tissues was determined using a BCA reagent (Beibo Biological Technology, shanghai, China). The proteins (15 µg) were later separated by 12% SDS-PAGE, followed by transfer onto PVDF membranes (Millipore, Darmstadt, Germany). Thereafter, 5% defatted milk was utilized to block PVDF membranes at room temperature for a 2 h period. PVDF membrane and primary antibody ERK1/2 (1:1000), p-ERK1/2 (1:1000), p-AKT (1:1000), AKT (1:1000), p38 (1:1000), p-P38 (1:1000), NF-κB p65 (1:1000), NF-κB p-p65 (1:1000), NF-κB IκB-α (1:1000), NF-κB p-IκB-α (1:1000) JNK (1:1000), p-JNK (1:1000), β-actin (1:1000), COX-2 (1:1000), iNOS (1:1000), ZO-1 (1:1000), Occludin (1:1000), and Claudin-3 (1:1000) were subject to overnight preservation under 4°C and rinsing by tbst lotion (Guo et al., 2019). Then, the PVDF membrane was subject to 2 h incubation with goat anti-mouse/rabbit secondary antibody (1:5,000) and washed in the TBS-T washing solution. Then, enhanced chemiluminescence solution was used to detect the changing trend of protein bands.

The Illumina Novaseq sequencing platform is used to examine microbial diversity. A small fragment library was created using paired-end sequencing to carry out HTS. In order to compare sample species composition, abundance and annotation analyses of species were performed, as well as filtering, clustering, or denoising of reads splicing. α-Diversity, β-diversity, correlation analysis, and significant species difference analysis were used to examine inter-sample differences. Beijing Biomark Biology Co., Ltd. completed the bacterial population sequencing analysis.

The results were presented as mean ± standard deviation (SD) deviation. Unpaired Student’s t-tests were used in Prism 8.0 to compare two groups, while general linear model ANOVA was used in Prism 8.0.20 to compare multiple groups. Pathological and clinical scores were evaluated using non-parametric tests (Liu K. et al., 2020).

To begin testing the protective efficacy of POG treatment in DSS-induced mice model of UC (Figure 1A). The images of the collected samples from the colon showed that the colon of the DSS group mice had hematochezia, along with significantly shortened colonic length (Figure 1B). The weight of the mice was recorded daily, colon samples were collected, and the length was measured (Figure 1C). DSS mice had a shorter average colon length than other groups, and the DSS + POG group (10 mg/kg) significantly improved the damage of DSS to the colon length of mice, as seen in Figure 1D.

In order to explore the protective effects of POG on the UC mice model, mice colon tissue samples were stained with hematoxylin-eosin. As a result, DSS mice showed fewer colonic goblet cells, lost colonic crypts, significantly damaged intestinal mucosa, and exhibited typical inflammatory tissue edema infiltration relative to other groups (Figure 2A). Histological and clinical scores showed that the DSS + POG group (2.5, 5, and 10 mg/kg) significantly decreased the histological and clinical scores of DSS mice dose-dependently (Figures 2B,C). In addition, levels of inflammatory factors (TNF-ɑ, IL-6, and IL-1β) in each group of mice were inhibited by POG within DSS mice; this concluded effective control of the development of inflammation by POG (Figures 2D–F).

In order to explore the protective mechanism of POG in UC mice, the mapped protein bands of MAPK, NF-κB, and AKT pathways were analyzed (Figure 3A). POG effectively hindered the phosphorylation activation of AKT, p65, and IκB-α proteins in NF-κB and AKT pathways (Figures 3B–D). Moreover, P38, JNK1/2, and ERK1/2 had significantly increased phosphorylation within the MAPK pathway in the DSS group, while POG administration reversed the above effect (Figures 3E–G). Therefore, POG suppressed MAPK, NF-κB, and AKT pathway activation, while alleviating inflammatory injury of mice colon.

The tight connectivity of intestinal tight junction (TJ) proteins is an important factor in forming the intestinal immune barrier. Immunofluorescence staining of the three tight junction proteins, namely, Occludin, Claudin-3, and ZO-1, and Western Blot experiments showed reduced expression levels in each group of colon samples (Figures 4A–G). This fully proved that DSS mice had the lowest TJ protein levels compared to other groups. It demonstrated that POG effectively alleviated the decrease of Occludin, Claudin-3, and ZO-1 protein levels in the DSS-induced mice ulcerative colitis model.

In this 16S rDNA, HTS was carried out to evaluate differences in gut microbiota. The differences in the Venn diagram and auto-number diagram were not significant among the four groups (Figures 5A,B). There were adequate sequencing data for reflecting species diversity in the sample (Figure 5C). Relative to DSS mice, mice in the rest three groups show more species of bacteria in the flora, and this richer species indicates that the test samples cover most of the microbial species information and sufficient sample size (Figures 5D–F). PCoA analysis and NMDS analysis show that the closer the distance on the coordinate map, the higher the similarity. Some studies believe that the effect of NMDS is better than PCA and PCoA. A stress value < 0.2 reflected the partial reliability of NMDS analysis (Figures 5G,H). Alpha (α)-diversity analysis revealed higher flora species diversity in three groups except for the DSS group (Figure 5I). Based on the analysis of similarities, that is, ANOSIM analysis, beta diversity did not present obvious differences among the four groups (Figure 5J).

Histogram of LDA value distribution showed that intestinal microbial species were significantly different between the four groups (Figure 6A). By analyzing the relative microbial abundance of phyla and genus level, the DSS group had decreased species richness of Lactobacillus, Firmicute, and Bacteroidetes compared with the POG group, whereas the level of Proteobacteria markedly decreased in DSS (Figures 6B,C). The cladogram was plotted from linear discriminant analysis effect size (LEfSe) analysis indicating a significant difference in all four groups. Diverse colors in cladogram represent diverse groups, whereas distinct node colors reflect the degree of microbiota in the grouping. Therefore, microbial levels were significantly different across the four groups (Figure 7A). DSS group had a markedly increased abundance of Enterobacteriales, Gammaproteobacteria, and Helicobacteraceae compared with the remaining three groups, which indicated that DSS-mediated UC model mice had disrupted gut microbial population structure (Figures 7B–D). POG positively regulates the colonization of intestinal flora, protects beneficial flora, and maintains the homeostasis of the intestinal environment.

The effect of POG on affected LPS-mediated RAW264.7 cells was examined in vitro. Firstly, the CCK8 method tests whether POG between 12.5 and 200 μmol/mL has toxic side effects on RAW264.7 cells (Figure 8A). RAW264.7 cells were subject to 1 h POG pretreatment, followed by 12 and 24 h LPS stimulation, respectively. Further mRNA and protein expression analyses were performed. qRT-PCR analysis indicated POG suppresses IL-1β, TNF-α, and IL-6 levels dose-dependently (Figures 8B–D). Meanwhile, protein bands showed that POG effectively inhibited iNOS and COX-2 upregulation within LPS-mediated RAW264.7 cells (Figures 8E–G).

To further investigate the effect of POG treatment on inflammation-related pathways, we extracted the proteins within LPS-mediated RAW264.7 cells for analyzing protein phosphorylation levels in MAPK, NF-κB, and AKT pathways (Figure 9A). According to the results, POG treatment significantly inhibited phosphorylation of AKT, IκB-α, and p65 proteins within RAW264.7 cells (Figures 9B–D). Meanwhile, POG treatment also decreased P38, JNK1/2, and ERK1/2 phosphorylation within the MAPK pathway (Figures 9E–G).