The herbal materials for PWP, specifically Citrus reticulata Blanco, Atractylodes lancea, Magnolia officinalis, and Glycyrrhiza uralensis Fisch, were procured from the First Affiliated Hospital of Jinan University in Guangzhou, China. The specific product batch numbers are as follows: Glycyrrhiza (HX24C01), Magnolia (HX24AC01), Atractylodes (HX25MC01), and Citrus reticulata (HX24C01). Dextran sulfate sodium (DSS) salt was procured from MP Biomedicals Co., Ltd. in the United States. The fecal occult blood detection kit was procured from Yuanye Bio-Technology Co., Ltd. in Shanghai, China. The high-fat diet (60% fat content) was procured from Charles River Laboratories in Shunde, China. Sodium butyrate (S817488), neomycin (N799581), ampicillin (A800429), vancomycin (V871983) and metronidazole (M813526) were purchased from Macklin, China. Mouse Interleukin 1β(IL-1β) Quantification Kit was purchased from Nanjing BYabscience Biotechnology Co.

The four herbal materials—Atractylodes lancea (24 g), Magnolia officinalis (18 g), Citrus reticulata Blanco (18 g), and Glycyrrhiza uralensis Fisch (12 g)—were soaked in distilled water at a ratio of 10:1 (water to herb weight) for 30 minutes, followed by boiling for an additional 30 minutes. Following the initial extraction, a second extraction was conducted using an equal volume of distilled water. The extracts from both extractions were combined and concentrated under reduced pressure at 70°C to achieve a specific final volume.

Chromatographic analysis was conducted using an Agilent SB-C18 column (2.1 mm × 100 mm, 1.8 µm, at 40°C) with a 2 µL injection volume. Mobile phase A consisted of ultrapure water, while mobile phase B comprised acetonitrile (both with 0.1% formic acid added), and the flow rate was maintained at 0.35 mL/min. The gradient elution program was as follows: at 0.00 min, 5% B; at 9.00 min, B was linearly increased to 95% and maintained for 1 min; from 10.00 to 11.10 min, B was returned to 5% and held at that concentration until 14 min. The mass spectrometer was operated using an electrospray ionization (ESI) source at a temperature of 500°C, with the ion spray voltage (IS) set to 5500V (positive mode) and -4500 V (negative mode). The gas flow rates were maintained at 50 psi for Gas I, 60 psi for Gas II, and 25 psi for Curtain Gas. The collision-induced dissociation parameters were adjusted to a high setting. Compound identification was achieved by comparing the data against a self-constructed molecular weight database (MWDB, Metware Database).

A total of thirty-five 6-week-old C57BL/6J mice (SPF grade) were procured from Charles River Laboratories in Shunde, China. The mice were randomly divided into five experimental groups. All mice were housed in an environment with a temperature of 25°C and a humidity of 60%. All groups, except for the control group, were fed a high-fat diet consisting of 60% fat content, while the control group received a standard diet. Beginning on day 15, 3% DSS was administered via the drinking water for a duration of 7 days. Furthermore, the PWP high-dose group (PWP-H), low-dose group (PWP-L), and the 5-ASA group were treated with their respective concentrations of medication (10 mL/kg): herbal dosages of 18.72 g/kg (PWP-H) and 9.36 g/kg (PWP-L), as well as 5-ASA administered via gavage at 100 mg/kg. Both the control and model groups received saline via gavage. The Animal Experiment Protocol listed below has been reviewed and approved by Laboratory Animal Welfare and Ethics Committee of Jinan University under approved protocol number IACUC-20241115-08.

The construction of pseudo-germ-free mice was based on a previously established methodology (Li et al., 2021). Briefly, a cocktail of antibiotics was administered via drinking water for 14 consecutive days. Subsequently, 3% dextran sodium sulfate (DSS) was added to the drinking water of all mice to induce colonic inflammation. In addition, sodium butyrate was added to the drinking water of the ABX-PWP-BA group to achieve a final concentration of 200 mM. FMT interventions were performed as follows: the FMT-Mol group received fecal enemas from the model group, while the FMT-PWP group received fecal enemas from the PWP-H treatment group. Furthermore, the ABX-PWP and ABX-PWP-BA groups were subjected to intragastric administration of PWP herbal medicine at a dose of 18.72 g/kg once daily for 7 days.

Body weight measurements were conducted on days 17, 19, and 21. Fecal consistency was assessed, and fecal samples were collected for fecal occult blood analysis. The Disease Activity Index (DAI) scores were calculated based on established methodologies as described in the literature (Li et al., 2022).

Colon tissue were fixed in 4% paraformaldehyde for 24 hours. Fixed tissues were dehydrated in graded ethanol (70%, 80%, 90%, 100%), followed by xylene for transparency. Tissues were embedded in paraffin wax and sectioned at 4µm thickness using a microtome. Sections were mounted on glass slides and dried. For H&E staining, sections were deparaffinized in xylene, rehydrated through graded ethanol (100%, 95%, 70%), and rinsed in distilled water. Nuclei were stained with hematoxylin for 3 minutes, followed by differentiation in citrate buffer (pH 4.8) for 10 seconds. Cytoplasm was counterstained with eosin for 1 minutes. Sections were dehydrated, cleared in xylene, and mounted with neutral mounting medium. Slides were examined under a light microscope for evaluation of tissue morphology.

After antigen retrieval of colon paraffin sections, the sections were placed in a mixture of primary antibodies ZO1 (GB111401, Servicebio, China) and Occludin (GB151981, Servicebio, China), and incubated at 4°C for 12 hours. Subsequently, secondary antibodies (GB23303, Servicebio, China) (GB21303, Servicebio, China) were incubated at room temperature in the dark for 1 hour, followed by DAPI staining of the cell nuclei before imaging.

Serum IL-1βconcentrations were measured using a commercial mouse IL-1βELISA kit (BYabscience, BY-EM220174) according to the manufacturer’s protocol. Briefly, standards and samples were added to antibody-coated 96-well plates and incubated with detection reagent for the recommended times, followed by wash steps to remove unbound material. After addition of substrate solution and termination of the enzymatic reaction, optical density was read at 450 nm using a microplate reader. Concentrations were calculated using a standard curve generated with SigmaPlot 15.0 software.

Colonic tissues were collected and lysed using high-strength RIPA lysis buffer (supplemented with protease and phosphatase inhibitors) to extract proteins. Following protein quantification, the samples were combined with the loading buffer. Proteins were separated using 10% SDS-PAGE and subsequently transferred to a PVDF membrane. The membrane was incubated with the primary antibody overnight, followed by incubation with the secondary antibody. Protein bands were visualized by chemiluminescence. Image J software was utilized to analyze the density of the protein bands.

Colonic tissues were collected, and RNA was extracted using the SteadyPure Universal RNA Extraction Kit (Accurate Biotechnology Co., Ltd; Hunan, China), followed by reverse transcription into cDNA using the Evo M-MLV RT Mix Kit (Accurate Biotechnology Co., Ltd; Hunan, China). Primers for the target genes were designed (see supplementary table), and qRT-PCR was performed. Relative quantification was determined using the 2^−ΔΔCT method with Gapdh as the internal reference gene.

Mice were euthanized, and cecal contents were promptly collected in 1.5 mL centrifuge tubes, which were subsequently stored in liquid nitrogen before being transferred to -80°C for long-term storage. Total DNA was extracted from fecal samples, and the concentration and purity of the DNA were assessed. The V3 and V4 regions of the 16S rDNA gene were amplified using specific primers. Following purification, quantification, and normalization of the amplification products, sequencing libraries were constructed and sequenced using the Illumina NovaSeq PE250 platform for 16S rDNA amplicon sequencing. Microbiota sequencing and data analysis were performed by APExBIO (Houston, USA).

Standard solutions of acetic acid, propionic acid, butyric acid, isobutyric acid, valeric acid, isovaleric acid, and caproic acid were prepared at a concentration of 100 mg/mL. The internal standard, 4-methylpentanoic acid, was prepared in ether at a concentration of 75 µg/mL. Working standard solutions of varying concentrations (0.02 µg/mL to 100 µg/mL) were prepared for gas chromatography-mass spectrometry (GC/MS) calibration. Fecal samples, measuring 200 µL, were mixed with 50 µL of 15% phosphoric acid, 100 µL of the internal standard solution, and 400 µL of ether. The mixture was homogenized for 1 minute, subsequently centrifuged at 12,000 rpm for 10 minutes at 4°C, and the supernatant was analyzed.

GC/MS analysis was conducted using a Thermo Trace 1300 system (Thermo Fisher Scientific, USA) equipped with an Agilent HP-INNOWAX capillary column (30 m × 0.25 mm ID × 0.25 µm). A 1 µL sample was injected at a 10:1 split ratio. The initial temperature was set at 90°C and ramped to 120°C at 10°C/min, then to 150°C at 5°C/min, and finally to 250°C at 25°C/min, where it was held for 2 minutes. The carrier gas used was helium, with a flow rate of 1 mL/min. Mass spectrometry was conducted utilizing a Thermo ISO 7000 mass spectrometer (Thermo Fisher Scientific, USA) operating in electron impact (EI) mode with an electron energy of 70 eV and in selected ion monitoring (SIM) scanning mode.

Colonic tissues were promptly fixed using an electron microscopy fixative at 4°C for 24 hours in the dark. The tissues were subsequently exposed to 1% osmium tetroxide for 2 hours, followed by a gradient dehydration process in ethanol at room temperature. The samples were embedded in 812 resin and polymerized at 60°C for a duration of 48 hours. Ultra-thin sections, measuring 60–80 nm, were stained with 2% uranyl acetate in alcohol for 8 minutes, followed by three washes with 70% ethanol and three washes with ultra-pure water. Finally, the sections were stained with 2.6% lead citrate for 8 minutes in a CO₂-free environment and subsequently washed again with ultra-pure water. The sections were dried overnight at room temperature and subsequently observed using a Transmission Electron Microscope (TEM).

Target Identification of gut microbiota metabolites: metabolites and their corresponding murine gut targets were retrieved from the gutMGene database (http://bio-annotation.cn/gutmgene/). The SCFAs detected in intestinal samples via GC/MS (acetic acid, propionic acid, isobutyric acid, butyrate, isovaleric acid, valeric acid, and caproic acid) were matched to their corresponding murine metabolite entries in the gutMGene database to identify associated molecular targets.

Intersection with key signaling pathways: The identified targets of the seven SCFAs were compared with genes annotated in the UC-associated PI3K/Akt/mTOR and autophagy signaling pathway and the autophagy pathway (KEGG pathway annotations).

Network visualization: The overlapping targets, their associated SCFAs, and the corresponding pathways were imported into Cytoscape version 3.9.1 to construct a metabolite-target-pathway interaction network.

Statistical analyses were conducted using SPSS version 25.0, and the results were expressed as the mean ± standard deviation (SD). Differences among groups were assessed using a one-way analysis of variance (ANOVA). Post-hoc multiple comparisons were performed using the LSD method for homoscedastic data or the Tamhane’s T2 method for heteroscedastic data. Statistical significance was defined as follows: Control group versus Model group: #: P<0.05, ##: P<0.01; Model group versus PWP group: *: P< 0.05, **: P<0.01.

Correlation analysis of gut microbiota abundance with SCFAs metabolites, colon tissue cytokines expression and autophagosome number was carried out using the spearman correlation calculation method. This process was performed using the Metware Cloud, a free online platform for data analysis (https://cloud.metware.cn).

The PWP was analyzed using UPLC/MS/MS, resulting in the detection of 3,940 compounds ( Supplementary Table 1 ). Based on secondary spectral information matched with the MWDB, we identified and characterized 15 quality control or key constituents derived from the four herbal components of PWP (Atractylodis Rhizoma, Magnoliae Officinalis Cortex, Citri Reticulatae Pericarpium, and Glycyrrhizae Radix) ( Table 1 , Figure 1 ).

We developed a mouse model using 3% DSS combined with a high-fat diet, and assessed the successful establishment of the model by utilizing 5-ASA as a positive control drug. Regarding body weight, following the introduction of 3% DSS in drinking water from day 15, overall body weights exhibited a decreasing trend. To assess body weight changes during the period of being given DSS free drinking (Day15-Day21), we compared the percentage of body weight change in each group of mice compared to that prior to the administration of DSS, and the results showed a significant reduction in the model group of mice compared to the control group (P<0.01). Following treatment with either 5-ASA or PWP, the degree of weight loss was less than that observed in the model group, particularly in the PWP-H group, which demonstrated a significant improvement compared to the model group (P<0.05) ( Figures 2A, B ). The DAI effectively reflected disease progression in the UC model, and treatment with 5-ASA significantly reduced the DAI on the fifth day of DSS administration (P<0.05), with a more pronounced reduction observed after treatment with PWP (P<0.01). On the seventh day, the DAI was significantly reduced (P<0.01) in all treatment groups compared to the model group ( Figures 2C–E ). Colon length serves as a reflective indicator of intestinal inflammation, with shorter lengths predicting more severe colonic inflammation. Compared to the control group, all groups subjected to UC modeling exhibited reduced colon lengths, with the model group showing the most significant reduction (P<0.01) and exhibiting bloody stool-like changes in intestinal contents. Conversely, both the 5-ASA and PWP-H treatment groups significantly increased colon length (P<0.01) ( Figures 2F, G ).

Disruption of the intestinal barrier is a characteristic pathological manifestation of UC. Histological examination of HE-stained colon sections revealed significant disruption of the villous structure of colonic epithelial cells in the model group, accompanied by extensive infiltration of neutrophils and macrophages compared to the control group. In contrast, the villous structure of the colonic epithelium remained intact, and inflammatory cell infiltration was attenuated in the groups treated with PWP or 5-ASA ( Figure 3A ). Immunofluorescence analysis demonstrated that the fluorescence intensity of ZO1 and Occludin, tight junction proteins found in colonic epithelial cells, was attenuated in the model group. Conversely, treatment with 5-ASA and PWP at both high and low doses significantly up-regulated the fluorescence intensity of these intestinal tight junction proteins ( Figure 3B ). Further examination of colonic tissues via Western blot analysis revealed that ZO1 and Occludin were significantly down-regulated in the colonic tissues of model mice, whereas their expression levels were restored following PWP treatment (P<0.05) ( Figures 3C–E ). In the model group, damage to the intestinal barrier was associated with a tissue inflammatory response, as evidenced by the up-regulation of pro-inflammatory factors IL-1β, IL-17, IL-6, and TNF-α (P<0.01). In contrast, treatment with PWP significantly down-regulated the expression of these pro-inflammatory factors in colonic tissues, particularly in the high-dose group (P<0.01) ( Figures 3F–I ).

Using the criteria of OB ≥ 30% and DL ≥ 0.18, a total of 104 active compounds of PWP were identified from the TCMSP database. Specifically, 43 targets were identified for Atractylodes lancea, 65 for Citrus reticulata Blanco, 25 for Magnolia officinalis, and 225 for Glycyrrhiza uralensis Fisch. After removing duplicates, a total of 235 unique pharmacological targets for PWP were obtained.

A comprehensive search for UC-related targets across the OMIM, DrugBank, TTD, and GeneCards databases was conducted, identifying 5,510 disease-associated targets. The intersection of PWP targets and UC disease targets yielded 177 potential therapeutic targets for PWP, which were visualized using a Venn diagram ( Figure 4A ). PPI networks for these 177 targets were constructed using the STRING platform ( Figure 4B ).

KEGG pathway enrichment analysis of the intersected targets revealed significant involvement in pathways including the IL-17 signaling pathway, p53 signaling pathway, Toll-like receptor signaling pathway, and PI3K/AKT signaling pathway ( Figure 4C ). Further network topology analysis using Cytoscape version 3.7.2 identified the core components and key targets involved in the therapeutic mechanisms of PWP. The constructed network consisted of 294 nodes and 1,990 edges, with dense connections highlighting the multi-component, multi-target, and multi-pathway therapeutic effects of PWP in the treatment of UC. This analysis demonstrated that the active components of PWP exert therapeutic effects on UC by interacting with distinct targets through various pathways ( Figure 4D ).

Injury to the intestinal mucosa and the entry of antigenic substances from the intestinal lumen into the lamina propria can trigger the activation of the PI3K/AKT/mTOR pathway. The preceding network pharmacological analyses suggest that PWP may treat UC by modulating the PI3K/AKT-related pathway. The pattern of phosphorylation serves as an indicator of the activation state of these proteins. PI3K was significantly upregulated in the model group compared to normal controls (P<0.01); correspondingly, the pAKT/AKT and pmTOR/mTOR ratios were also higher in the model group (P<0.01). Following treatment with PWP, both the PI3K (P<0.01) and pAKT/AKT (P<0.05) ratios were significantly down-regulated. Meanwhile, a significant down-regulation of the pmTOR/mTOR ratio was observed in the PWP-H group (P<0.01) ( Figures 5A–D ). These findings suggest that the PI3K/AKT/mTOR pathway is significantly activated in the disease state, and that PWP inhibits the activation of proteins within this pathway, which may be a key mechanism for its efficacy.

mTOR is an inhibitory target for initiating the autophagy process. The up-regulated pmTOR/mTOR ratio in the colonic tissues of the model group is detrimental to the macro-autophagy process of these tissues, impeding the phagolysis and catabolism of damaged molecules. During the autophagy process, lysosomal-associated membrane protein 1 (LAMP1) serves as a structural component of lysosomal membranes, facilitating the fusion of autophagosomes with lysosomes during the autophagic process, and its expression level was significantly upregulated after treatment of PWP (P<0.01) ( Figure 6A ). P62 is consumed as a substrate, and the expression level of P62 in the colonic tissues of mice receiving high-dose PWP treatment was significantly down-regulated (P<0.05) compared to the model group ( Figure 6B ). The ratio of LC3B-II/LC3B-I reflects the level of autophagy, with higher levels of LC3B-II indicating increased autophagosome formation. Similarly, the ratio of LC3B-II/LC3B-I was down-regulated in the model group (P<0.05), whereas it was restored after treatment with PWP (P<0.05) ( Figures 6C, D ). Furthermore, TEM revealed that the control group exhibited normal morphological structures of the colon, including intact mitochondrial structures and clear endoplasmic reticulum structures without swelling or dilatation. In contrast, the model group displayed a significant reduction in the number of autophagosomes and autophagolysosomes (P<0.05), accompanied by mitochondrial swelling and endoplasmic reticulum dilatation. Compared to the model group, the PWP-treated group exhibited a significant increase in the number of autophagosomes and autophagolysosomes (P<0.01), with an increase in structurally intact mitochondria, reduced swelling, and decreased endoplasmic reticulum dilatation ( Figures 6E, F ).

The gut microbiota, comprising up to 100 trillion cells, plays a critical role in various diseases, including gastrointestinal, neurological, and respiratory disorders (Wang et al., 2023, 2024b; Wu et al., 2024). Oral administration is the predominant route for herbal medicines, and modulating gut microbiota along with their metabolites serves as a key mechanism underlying their therapeutic effects. Principal coordinate analysis (PCoA) revealed significant alterations in the gut microbiota composition of UC mice, which were notably influenced by PWP treatment ( Figure 7A ). At the genus level, 82 overlapping genera were identified across the three groups, with 17 unique genera present in the UC model group and 11 unique genera emerging following PWP treatment ( Figure 7B ). Alpha diversity indices, including the Shannon and Simpson indices, reflect microbial diversity, while the Chao indices indicate species richness. Both diversity and richness indices were significantly reduced in UC model mice, but they were partially restored following treatment with PWP. However, the alpha diversity indices did not achieve statistical significance (P > 0.05) ( Figures 7C–E ). Linear discriminant effect size (LEfSe) analysis revealed the enrichment of opportunistic pathogens including Staphylococcus, Ruminococcus torques group, Desulfovibrio, Clostridium sensu stricto 1, and Enterococcus in the model group. Conversely, beneficial bacteria, including Bifidobacterium and Bacteroides, were enriched in the PWP treatment group ( Figure 7F ).

Based on our findings that PWP modulates known SCFAs-producing bacteria, we further quantified the levels of SCFAs in the intestinal contents. PWP treatment significantly upregulated the levels of butyric, isovaleric, and valeric acids in the gut contents compared to the model group ( Figures 8A–C ). GPR43, a recognition receptor for SCFAs, was significantly up-regulated in the PWP-treated group ( Figures 8D, E ). Further correlation analysis revealed that Desulfovibrio, which was enriched in the model group, was negatively correlated with fecal levels of propionic acid, isobutyric acid, and isovaleric acid, while Ruminococcus torques group was negatively correlated with butyric acid levels. Both were positively correlated with the levels of inflammatory factors in colonic tissue. In contrast, Alistipes and Parabacteroides, which were enriched in the feces of the PWP-treated group, were positively correlated with fecal levels of propionic acid, butyric acid, isobutyric acid, and isovaleric acid, while negatively correlated with the colonic tissue levels of IL-17, IL-6, TNF-α, and IL-1β ( Figure 8F ). A search of the gutMgene database revealed that the targets of action of SCFAs generated 10 crossover repeats in UC-associated PI3K/AKT/mTOR as well as autophagy pathway-associated targets, among which butyric acid was identified as a core target ( Figure 8G ). Correlation analysis also revealed that the abundance of Alistipes and Parabacteroides in feces was positively correlated with the number of autophagosomes formed in colonic epithelial cells, whereas Desulfovibrio exhibited a negative correlation. The results suggest that the upregulation of colonic autophagy levels by PWP, through the inhibition of the PI3K/AKT/mTOR pathway, may be mediated by its regulatory effects on intestinal SCFAs-producing bacteria and their associated SCFAs.

To further confirm that the gut microbiota serves as a central mediator of PWP’s therapeutic effects, we performed a FMT experiment. In a DSS-induced colitis model, mice receiving fecal gavage from the PWP-H treatment group—rather than those mice treated with PWP after gut microbiota clearance with antibiotics—exhibited markedly attenuated intestinal inflammatory injury ( Figure 9A ) and significantly higher expression of intestinal tight junction proteins (P < 0.01) ( Figures 9B–D ) compared with mice that received feces from the model group. Concurrently, activation of PI3K/AKT/mTOR pathway related proteins was significantly inhibited (P < 0.01) ( Figures 9E–G ), while LAMP1 and LC3BII/LC3BI levels were elevated, and the autophagy substrate P62 was more extensively consumed (P < 0.01) ( Figures 9H–J ).

Butyrate, a key gut microbiota derived metabolite, may represent the principal effector molecule in this process. In a butyrate supplementation experiment, mice in the ABX-PWP-BA group displayed reduced colonic inflammatory injury compared with the ABX-PWP group. This improvement was accompanied by upregulation of the SCFAs receptor GPR43 (P < 0.05) ( Figure 9K ) and increased expression of tight junction proteins ZO-1 and Occludin (P < 0.01) ( Figures 9B–D ). Additionally, butyrate supplementation suppressed phosphorylation of PI3K/AKT/mTOR pathway–related proteins (P < 0.01) ( Figures 9E–G ), markedly increased the expression of autophagy-related proteins LAMP1 and LC3BII/LC3BI, and further enhanced consumption of P62 (P < 0.05) ( Figures 9H–J ). ELISA analysis revealed that serum IL-1β levels were significantly reduced in both the ABX-PWP-BA and FMT-PWP groups (P < 0.05) ( Figure 9L ). Collectively, these findings indicate that during PWP treatment of HFD-induced UC, the gut microbiota and its metabolite butyrate are key intermediaries mediating its therapeutic efficacy. This mechanism contributes to suppression of PI3K/AKT/mTOR pathway activation, thereby enhancing autophagy in colonic tissue, promoting repair of the intestinal barrier, and mitigating inflammatory injury.