Multi-aspect therapeutic effects of a polyphenolic herbal formulation Cirsium japonicum, Scutellaria baicalensis, Paeonia japonica, and Glycyrrhiza uralensis on ulcerative colitis: inflammation modulation, gut microbiota remodeling, and metabolite profiling
- Open Access
- 03.01.2026
- Research
Erschienen in:
Abstract
Introduction
Mounting evidence shows that multi-target therapeutic strategies, which simultaneously address inflammation, epithelial barrier dysfunction, and gut microbiota imbalance, are more effective in managing ulcerative colitis than traditional single-target approaches [1]. Studies have demonstrated that combining agents with complementary actions yields better outcomes than monotherapy [1‐3]. This growing recognition has led to increased research interest in interventions capable of modulating multiple disease axes.
Ulcerative colitis (UC) is a long-term inflammatory bowel disease (IBD) marked by inflammation and ulceration of the mucosal layer, predominantly impacting the colon and rectum. Despite significant improvements in treatment, managing UC continues to pose difficulties owing to the variable patient’s responses and the possible adverse effects linked to standard therapies, including 5-aminosalicylic acid (5ASA), corticosteroids, and immunosuppressive agents [2]. Biologics and small-molecule inhibitors, such as Janus kinase (JAK) inhibitors, have broadened the range of treatment alternatives; however, their long-term safety and availability remain a concern [3, 4].
Anzeige
Moreover, recent reports showed mesalamine (5ASA) rapidly depletes bacterial polyphosphate (polyP), a key stress-protection polymer, thereby sensitizing gut microbes to oxidative stress, and we know that oxidative stress is higher in inflammation gut situation [5]. This mechanism raises concerns for long-term use, since evidence indicates 5ASA can diminish microbial diversity even in healthy subjects. Especially while in inflammatory conditions, 5ASA’s cannot restore microbial balance fully, underscoring the inherent limitations of chronic mesalamine therapy [6‐9].
In contrast, one of the notable alternatives that has garnered interest is plant extracts rich in polyphenols, which are recognized for their anti-inflammatory, antioxidant, and gut microbiota-modulating effects. These polyphenols, prevalent in numerous plant species, have shown therapeutic promise in preclinical studies for UC. For instance, extracts derived from cherries have been noted for their ability to reduce symptoms of UC by inhibiting the Wnt/β-Catenin signaling pathway [10], while Green pea hull polyphenol extracts also demonstrated protective properties through the modulation of gut microbiota and the activation of the Nrf2 signaling pathway [11]. Honey polyphenols have also been demonstrated to alleviate DSS-induced colitis in rats by fostering the growth of beneficial gut microbiota, thereby reinforcing the theory that dietary polyphenols can play a significant role in maintaining gut homeostasis [12].
Our herbal CSPG extract contains Cirsium japonicum, Scutellaria baicalensis, Paeonia japonica, and Glycyrrhizae radix. These four notable medicinal plants are recognized for their anti-inflammatory effects and protective qualities for the gastrointestinal tract. These herbs are rich in bioactive polyphenols, which have demonstrated significant therapeutic benefits in inflammatory conditions. For example, baicalin derived from Scutellaria baicalensis and glycyrrhizin extracted from Glycyrrhizae radix have been found to influence inflammatory signaling pathways and help maintain the integrity of the gut barrier [13, 14]. Cirsium japonicum has also been investigated for its potential to modulate NF-κB signaling, which is a crucial inflammatory pathway involved in UC [15].
Combining these herbs into a polyphenol-rich extract (CSPG) also aims to produce a therapeutic agent that offers broad-spectrum anti-inflammatory, microbiota-modulating, and mucosal-protective effects. In more detail, various phenolic compounds found in our CSPG extract have shown significant effectiveness in treating UC. Chlorogenic acid, sourced from Cirsium japonicum, mitigates inflammation and enhances gut barrier function by influencing gut microbiota composition [16]. Biochanin A, extracted from Scutellaria baicalensis, reduces UC symptoms by inhibiting MAPK/NF-κB signaling pathways [17]. Gallic acid, derived from Paeonia japonica, boosts gut immunity through the induction of ILC3 and the modulation of bile acid metabolism [18]. Ferulic acid, prevalent in Glycyrrhizae radix, obstructs NF-κB and iNOS-NO pathways, thereby diminishing oxidative and apoptotic stress associated with colitis [19]. Ferulic acid improved Ruminococcus alongside other beneficial taxa abundance [20].
Anzeige
This study explores the prophylactic effects of CSPG extract in a DSS-induced murine model of UC, and focusing on its ability to reduce inflammation by suppressing TNF-α and IL-6, enhance gut barrier integrity by up-regulating ZO-1 and occludin, and modulating gut microbiota by promoting beneficial bacteria while reducing harmful strains. It also examines CSPG’s influence on metabolic pathways, particularly short-chain fatty acids (SCFAs), which support gut health and immune regulation. These findings aim to highlight CSPG’s potential as a multi-targeted therapy for UC. Interestingly, before DSS stimulation, we also compared the baseline effects of 5ASA and CSPG on the gut microbiome and metabolites in healthy mice, which showed many different results between the two treatments.
By integrating multi-omics approaches, microbiota sequencing, and metabolomic profiling, this study provides a comprehensive evaluation of CSPG’s therapeutic potential. Given the multifactorial nature of UC pathogenesis, the findings could pave the way for new phototherapeutic strategies that complement or enhance existing UC treatments.
Materials and experimental methods
CSPG extraction method and mixture preparation method
In this study, the four constituent herbs were Cirsium japonicum, Scutellaria baicalensis, Paeonia japonica, and Glycyrrhizae radix. The authenticated crude drugs (2 kg per type) were purchased and re-verified by Prof. Chang-Su Na (Dongshin University) from a certified supplier, K-herbal (Seoul, Korea). The herbs were added to a cleared pilot-scale stainless steel extractor (Dongshin University) along with 25 L of distilled water (Milli-Q® IQ 7000, 18 MΩ·cm).
Extraction was performed for 4.0 h at 120 ± 5 °C and 0.17 ± 0.02 MPa with subsequent cooling for 3.0 h. Supernatants with an °Brix of 3.9 (Cirsium japonicum), 4.9 (Scutellaria baicalensis), 3.3 (Paeonia japonica), and 6.1 (Glycyrrhizae radix) were obtained from the aqueous extracts after cooling.
Centrifugal extracts were performed 2 h after cooling and were subjected to 3000 ×g at 4 °C for 20 min (1248R, LaboGene). The Extractions (Scutellaria and Glycyrrhizae, Paeonia) underwent 3–5 h of second concentration (at 102 ± 2 °C) to 150 mL for Cirsium or 200–250 mL for the others.
Each individual thick extract was then prepared for vacuum drying, wherein 200 g aliquoted portions were processed for drying and subsequently cooled for 2 h after each cycle.
Vacuum drying (OV-11, Jeio-Tech) comprised two stages: (i) to a non-flowing syrup at 75 ± 10 °C, − 0.05 to − 0.08 MPa (gauge) for 20–28 h and (ii) at 60–70 °C, − 0.08 ± 0.01 MPa (gauge) for 24 ± 4 h to constant weight. The dried masses were pulverized. The yields (% of starting herb, w/w) were 38.75% (Cirsium), 36.5% (Scutellaria), 17.5% (Paeonia), and 12.25% (Glycyrrhizae).
The resultant four powders were then combined in proportions of 4:2:2:1 to make CSPG (Dae-gye Hwang-geum Jakyak-tang). The Processing parameters (durations/volumes) assigned to each herb were based on their viscosity, °Brix, thermal properties, and bioactivity for consistency across batches (Fig. 1).
Anzeige
The four powdered extracts were then mixed together in a 4:2:2:1 ratio to obtain the CSPG extract, based on both traditional and mechanistic considerations (provided in Explanation for the Choice of CSPG ratio Supplementary material).
Fig. 1
Extraction and Preparation Method of CSPG Mixture
Table 1
The composition of CSPG extract
Pharmacognostic | Ratio | Source |
|---|---|---|
Cirsium japonicum Fisch | 4 | China |
Scutellaria baicalensis | 2 | China |
Paeonia japonia | 2 | Korea |
Glycyrrhizae radix | 1 | China |
LC-MS/MS analysis
The active components of CSPG were identified through the analysis of polyphenols utilizing HPLC-MS/MS (4500 Q-Trap, AB SCIEX; LC-40 A System, Shimadzu). A 10 µL sample was injected into a Gemini C18 column (3 μm, 50 × 2.0 mm; Phenomenex, Cat# 00 F-4453-B0) and examined under the following parameters: column oven temperature set at 40 °C, autosampler temperature at 15 °C, and a flow rate maintained at 0.3 mL/min. The analysis was performed in both negative and positive ion modes employing Turbo Ion Spray.
Anzeige
The mobile phase comprised water with 0.1% formic acid (Solvent A; Sigma-Aldrich, Cat# 33015) and acetonitrile with 0.1% formic acid (Solvent B; Honeywell, Cat# 34851).
The gradient conditions for solvent B were established as follows:
0–0.1 min: B (5–5%)
0.1–2 min: B (5–40%)
Anzeige
2–min: B (40–80%)
3–min: B (80–80%)
5–min: B (80–5%)
1–8 min: B (5–5%)
The sample was analyzed according to these solvent conditions.
HT-29 cell culture
HT-29 cells (ATCC HTB-38, USA) were cultured in RPMI1640 (Welgene, Cat# LM011-01, Korea) with the addition of 10% FBS (Gibco BRL, Cat# 26140-079, New York, USA) and 1% penicillin/streptomycin (Welgene, Cat# LS202-02, Korea). The cells were incubated in a Heracell VIOS 160i CO₂ incubator (Thermo Fisher Scientific, USA) at 37 °C with 5% CO₂ and 95% humidity.
HT-29 cell viability measurements
For cell viability assay, HT-29 cells were seeded into 96-well clear plates (SPL Life Sciences, Cat# 30296, Korea) and allowed them to stabilize overnight. The cells were then treated with various concentrations of CSPG for 24 and 48 h. Following treatment, MTT solution (0.2 mg/mL; Sigma-Aldrich, Cat# M5655) was added to each well and incubated for 4 h. The supernatant was removed, and the resulting formazan crystals were dissolved in 200 µL of anhydrous DMSO (Sigma-Aldrich, Cat# D8418) per well. Finally, the absorbance at 540 nm was measured using a microplate reader (SpectraMax iD3, Molecular Devices, Sunnyvale, CA). Cell viability was compared to a control that did not receive any treatment.
$$\begin{aligned}&\text{[Survival rate}(\% \text{of control)}=\\&100\times\text{(absorbance of drug-treated sample)/(absorbance of control)}]\end{aligned}$$
Western blotting
HT-29 cells were plated in a 60 mm dish at a density of 20 × 10⁴ cells per dish and allowed to incubate for 24 h. The cells underwent pre-treatment with CSPG at varying concentrations for a duration of 2 h, after which they were exposed to 30 ng/mL TNF-α (recombinant, PeproTech, USA) and incubated further. Following an additional 48-hour incubation period, the cells were harvested using a scraper, centrifuged at 1,000 RPM for 3 min, washed once with DPBS (liquid, Welgene, Korea), and subjected to a second centrifugation at 13,000 RPM for 1 min to eliminate the supernatant.
Protein extraction was carried out utilizing M-PER™ Mammalian Protein Extraction Reagent (solution, Thermo Fisher Scientific, USA), supplemented with Halt™ Protease and Phosphatase Inhibitor Cocktail (solution, Thermo Fisher Scientific, USA), and the mixture was sonicated on ice for 3 s. The lysate was then centrifuged at 13,500 RPM at 4 °C for 10 min, and the supernatant was collected for further analysis. Protein concentration was determined using the Bradford Assay Dye Reagent (liquid, Bio-Rad, USA), and the proteins were denatured in 5× SDS-PAGE loading buffer at 95 °C for 5 min. The samples were separated via SDS-PAGE gel (Mini-PROTEAN TGX, Bio-Rad, USA) and transferred onto a PVDF membrane (0.45 μm, Amersham, UK) for 70 min.
The membrane was blocked with 5% skim milk in TBS-T buffer for 1 h, washed, and then incubated overnight at 4 °C with the following primary antibodies: β-Actin (4970, Cell Signaling Technology, USA), ZO-1 (33–9100, In vitrogen, USA), ZO-2 (61–7300, In vitrogen, USA), ZO-3 (ab150372, Abcam, UK), COX-2 (ab52237, Abcam, UK), IL-6 (ab9324, Abcam, UK), NF-κB p65 (8242, Cell Signaling Technology, USA), Lamin B1 (13435, Cell Signaling Technology, USA), Akt (9272, Cell Signaling Technology, USA), p-Akt (9271, Cell Signaling Technology, USA), Occludin (71–1500, In vitrogen, USA), Claudin-2 (32–5600, In vitrogen, USA), iNOS (ab15323, Abcam, UK), MUC1 (sc-7313, Santa Cruz Biotechnology, USA), MUC2 (sc-13312, Santa Cruz Biotechnology, USA), MUC4 (ab60720, Abcam, UK).
After washing, the membrane was treated with HRP-conjugated secondary antibodies for 2 h at room temperature:
HRP anti-rabbit IgG (112-035−045, Jackson ImmunoResearch, USA).
HRP anti-mouse IgG (115-035−146, Jackson ImmunoResearch, USA).
This was followed by additional washes. Protein expression was visualized using ECL substrate (WBKLS0500, Millipore, USA) and imaged using the iBright™ CL1500 Imaging System (Thermo Fisher Scientific, USA).
For nuclear protein NF-κB p65 and Lamin B1, cells after collecting were processed using the NE-PER™ Nuclear and Cytoplasmic Extraction Kit (78833, Thermo Fisher Scientific, USA) following the manufacturer’s instructions. In each replicate, the untreated control (UT) group was set as 1.0 for normalization. Therefore, when averaging, the UT group shows no deviation and appears as a fixed 1.0 value without error bar.
Animal grouping and in vivo experiment scheme
Fig. 2
Schematic diagram of the experimental design. The mice were divided into 5 groups: Normal, UT + DSS, 5ASA + DSS, CSPG50 + DSS groups, CSPG100 + DSS groups. The Normal group was given water for 24 days; the other 4 groups were given normal water at the first 14 days, and then given water containing 1.5% DSS for 10 days. The CSPG50 and CSPG100 groups were given CSPG(50 mg/kg, 100 mg/kg) supplemented with dietary during the whole experiment
Experimental animals purchased 6-week-old BALB/C male mice from Sam Taco Bio (Seongnam, Korea) and experimented with water and feed in a kennel that maintains constant temperature and humidity, with 1 mouse per cage. Prior to the experiment, the animals had an adaptation period of 1 week, and were divided into 5 groups: normal group, colitis-induced group (UT + DSS; 1.5% DSS after normal diet for 2 weeks), positive control group (5ASA-200 + DSS; 5ASA-200 diet for 2 weeks followed by 1.5% DSS), experimental group 1 (CSPG-50 + DSS; group receiving 1.5% DSS after eating 0.3 mg/g mixed feed for 2 weeks), and experimental group 2 (CSPG-100 + DSS; group receiving 1.5% DSS after eating 0.6 mg/g mixed feed with CSPG for 2 weeks) (Fig. 2). The doses were predetermined as mentioned in the Supplementary material explanation for the choice of CSPG ratio.
DSS (MW 36–50 kDa, MP Biomedicals, USA) was dissolved in 1.5% sterile distilled water in drinking water to induce acute colitis in mice and autonomously watered for 10 days from day 15. The normal group was provided with sterile distilled water.
This timeline outlines the experimental checkpoints during the CSPG intervention and DSS-induced colitis study. On day 14, fecal samples were collected to assess changes in the gut microbiome and metabolite profiles, mentioned in the following Figs. 6, 7 and 8. On day 24, mice were sacrificed for more complete analysis in Figs. 9, 13, 12 and 13.
In vivo drug administration
In previous studies using DSS-induced model, anti-colitis drugs mixed pellet is a widely utilized, intended to relieve the stress caused by repeated oral gavage in colitis models and increase adherence to the treatment voluntarily [21‐24]. It also aimed to circumvent possible physiological blunting like raised corticosterone owing to chronic stress—elements that would undermine animal welfare and experimental rigor. This approach to feeding—often called voluntary feeding—has been documented extensively in animal studies focusing on conditions involving the gastrointestinal tract, where continuous drug delivery is necessary with minimal behavioral and immunological interference [24].
For DSS, the dosage 1.5% DSS/drinking water administration and the dosage of 5ASA (200 mg/kg) as positive control were based on previous research, and our experience [25].
In this research study, CSPG and 5ASA treatments consisted of administering drug loaded pellets, each prepared for the specified group:
Type 1 for CSPG50 group: 2.5 mg CSPG per 1 g normal pellet
Type 2 for CSPG100 group: 5.0 mg CSPG per 1 g normal pellet
Type 3 for 5ASA group: 10 mg 5ASA per 1 g normal pellet
Control group: feeds identical to the experimental groups, but without any drug mixed in
All types of feed were produced under the same conditions to warrant comparability. The small ratio of drug to pellet mass, effectively masking any odor or taste. The selection of the dosage is mentioned in the Supplementary material explanation for the choice of CSPG ratio.
To enable precise dosage control, each mouse was placed in individual cages at all times. Every day, at 10 AM, we emptied the feeding rack of the cage with no feed left, then each cage is provided with a drug mixed pellet, and mice were permitted to be freely consumed. The calculation for drug pellet weight is explained below. In each cage, we active monitoring ensured that the drug mixed pellets were consumed in their entirety. As soon as this drug pellet in each cage was consumed with no leftovers and debris, unlimited standard (non-medicated) feed was provided in each cage.
Drug mixed pellet dosage was individualized based on each mouse’s body weight in each of the 3 drug administered groups, for example:
CSPG50 group (we used type 1 drug loaded pellet mentioned above): A 19 g mouse needed 0.38 g of this feed contained totally 0.95 mg CSPG. Therefore, achieve 50 mg CSPG/kg mouse weight per day.
CSPG100 group (we used type 2 drug loaded pellet mentioned above): A 19 g mouse needed 0.38 g of this feed contained 1.9 mg CSPG. Therefore, we achieve 100 mg CSPG/kg mouse weight per day.
5ASA group (we used type 3 drug loaded pellet mentioned above): A 19 g needed 0.38 g of this feed contained 3.8 mg 5ASA. Therefore, achieve 200 mg 5ASA/kg mouse weight per day.
Every 3 days, the body weight of each mouse was recorded for the recalculation of pellet mass to maintain dosing accuracy. We also noted that all mice fully consumed the drug mixed pellets within 1.5 to 8.5 h of instruction. This guaranteed complete adherence and exact drug administration during the duration of the study.
Sacrifice and mouse tissue processing
After fasting 12 h before the end of the experiment, mice under respiratory anesthesia were laid, cardiac blood was drawn, the large intestine was separated, the length was measured and fixed in 10% formalin. This animal experiment was conducted with the approval of the Animal Experiment Ethics Committee of Dongshin University (approval number: DSU 2024-03−02), and the handling and management of experimental animals followed the Animal Experiment Handling Regulations of the Laboratory Animal Ethics Committee.
To reduce variability in metabolic and inflammatory biomarkers, as well as decrease intestinal content for more precise colon length measurement and histological evaluation, mice were fasted for 12 h prior to sample collection. This fasting approach is consistent with prior works employing DSS-induced colitis models reported that a 12-hour fast enhances consistency in downstream analyses [26]. After fasting, mice were anesthetized with isoflurane, and cardiac blood was collected. Mice were then humanely euthanized via cervical dislocation, and colonic tissues were collected.
Mouse intestinal tissues were first subjected to washing with phosphate-buffered saline (PBS), followed by fixation in a 10% formalin solution and subsequent embedding in paraffin. Then, the paraffin blocks were then sectioned, deparaffinized, and stained using hematoxylin and eosin (H&E) (PetoBio, Korea). Representative images of the stained sections were obtained through a slide scanner. The histological assessment of colitis progression in the samples was categorized as follows: Normal colonic mucosa (0), cryptal defect less than one-third (1), cryptal defect between one-third and two-thirds (2), mild inflammatory infiltration in the proper lamina (3), and mucosal erosion or ulceration accompanied by significant inflammatory infiltration (4).
Measurement of procalcitonin (PCT) in intestinal tissue
Procalcitonin in enteric tissue was measured using an ELISA kit (NBP2-81212, Novus Biologicals, USA). Intestinal tissue is cleaned of blood by PBS. Tissue weight: PBS (1:9) is placed in a tube for fracture, 5 m/s with a tissue shredder (Bead Ruptor 24, Omni International/PerkinElmer, USA), run for 10 s, and placed on ice for 5 min. Treat with an ultrasonic crusher (POWERSONIC 410, Hwashin Technology, Korea) for 3 s and place on ice for 5 min. 5000×g centrifugation at 4 °C for 5 min, then transfer the supernatant. Protein concentration was measured by the Bradford method (Protein Assay Dye Reagent Concentrate, 500-0006, Bio-Rad, USA). After the protein measurement, the same amount of protein is used. The ELISA analysis followed the manufacturer’s method.
Gut microbiota profiling using 16 S rRNA sequencing
Gut microbiota analysis was based on our previous published work with slight modifications [27]. Fecal samples were collected from mice administered CSPG for 2 weeks or from DSS-induced colitis models 5 days after DSS administration. DNA was extracted using the AccuFAST automated system (AccuFAST DNA Extraction System, AccuGene Inc., Incheon, Korea) according to the manufacturer’s instructions. For the DSS model, DNA purification followed a lithium chloride-based precipitation method: 0.1 volume of 8 M LiCl (L9650, Sigma-Aldrich, USA) was added to the DNA solution and incubated on ice for 30 min, followed by precipitation with 100% ethanol (459836, Sigma-Aldrich, USA) at − 20 °C for 2 h. The mixture was centrifuged at 13,000 rpm for 30 min at 4 °C, and the resulting pellet was washed twice with 70% ethanol (E7023, Sigma-Aldrich, USA), dried, and dissolved in TE buffer.
DNA quality was assessed prior to sequencing. For microbial community profiling, the V4 region of the 16 S rRNA gene was amplified using primers 515fb and 806rb with Nextera adapter sequences and the KAPA HiFi HotStart ReadyMix (KK2601, Roche Sequencing, Pleasanton, CA, USA), and amplicons were purified with Hi-AccuBeads (AccuGene Inc., Incheon, Korea). Sequencing was performed using the Illumina MiSeq platform (MiSeq Reagent Kit v2, 500 cycles, MS-102–2003, Illumina, USA).
Raw reads were denoised, filtered for chimeras, and trimmed for quality using DADA2 v1.16 to infer amplicon sequence variants (ASVs). Taxonomic classification of ASVs was performed using QIIME2 (version 2022.2) with a Naïve Bayes classifier trained on the SILVA 138 rRNA gene database at 99% sequence identity. Functional prediction of microbial communities was performed using PICRUSt2 based on ASV profiles. Microbial alpha diversity was assessed using Chao1, Shannon, Simpson, and Faith’s phylogenetic diversity (Faith_PD) indices, β-diversity was assessed using Jaccard and Bray–Curtis distances and visualized by PCoA.
Statistical analysis of microbial data
Alpha diversity was assessed quantitatively Chao1, Shannon, Simpson, and Faith’s PD indices. For Beta diversity, Jaccard and Bray–Curtis distance matrices were computed and visualized by PCoA. All statistical analyses were performed using GraphPad Prism v9.3.1. The Results are reported as mean ± SD. Significance was considered at p-value < 0.05. For multiple comparisons, the false discovery rate (FDR) procedure was used, and q-values were reported as appropriate.
Metabolite analysis and pathway profiling by GC-MS
Three fecal short-chain fatty acids (SCFAs)—acetic acid, propionic acid, and butyric acid—were quantified using gas chromatography–mass spectrometry (GC-MS) based on a modified method from previous reports. Approximately 50–100 mg of fecal sample was acidified, extracted with diethyl ether, and derivatized using propyl chloroformate to enhance volatility. The derivatized samples were injected into a GC-MS system operating in selected ion monitoring (SIM) mode for targeted detection. Internal standards were used for quantification [28]. This method provided high sensitivity, specificity, and reproducibility for SCFA profiling in fecal matrices, allowing accurate assessment of gut microbial fermentation activity in experimental groups.
In addition to SCFA detection, a broader panel of 29 metabolites was analyzed using a GC-MS–based method established in our previous study [29]. GC-MS data were processed using SIMCA-P version 16.0 (Umetrics, Umea, Sweden) to perform principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA), which enabled visualization of group-specific metabolic differences. For compound identification, the acquired mass spectral data were compared against the NIST 14.0 library, the Human Metabolome Database (HMDB, http://www.hmdb.ca), and interpreted with the aid of AIoutput software. Statistical significance was determined using Student’s t-test with false discovery rate (FDR) correction, applying a threshold of 5%. Metabolites with FDR-adjusted p-values below 0.05 were included in downstream pathway enrichment and topology analysis, which was carried out using MetaboAnalyst (https://www.metaboanalyst.ca) and referenced against the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Only pathways with both an adjusted p-value < 0.05 and impact value >0.1 were considered biologically meaningful.
Correlation analysis
The association between metabolites and microorganisms was analyzed using Pearson’s correlation analysis. The detailed method is the same as described in the previous study [30].
Pharmacology networking analysis
An investigation on the extraction of CSPG focused on identifying potential molecular targets using network pharmacology. The first step involved a comprehensive review of the literature and online databases for any listed bioactive CSPG components; with the goal to emphasis polyphenol and flavonoid compounds known for their anti-inflammatory and gut protective activities. Swiss Target Prediction was then utilized to predict potential protein targets based on bioactivity data that had been previously collated, and structure chemotype. To narrow down targets associated with UC, gene disease association data was collected from Gene cards database using the keyword “ulcerative colitis”. The predicted targets of CSPG were intersected with the UC-related genes to identify overlapping protein targets. These predicted targets were further analyzed using the KEGG (Kyoto Encyclopedia of Genes and Genomes) database to identify proteins related to inflammation pathways. The assembly of the protein-compound interaction network was done on Cytoscape 3.9.1. In this network, nodes represented the compounds and target proteins, while edges indicated their interactions. Key inflammatory targets were marked in order to emphasize their importance in UC pathophysiology. In addition, KEGG pathway enrichment analysis was performed to identify pathways associated with tight junction regulation. This method proved valuable for predicting CSPG’s molecular interactions with high confidence. Also, bioinformatics and experimental data were correlated, revealing the molecular mechanism underlying CSPG’s therapeutic effects in UC.
Analytics
Intergroup significance was assessed by ANOVA, and Newman-Keuls test was used to assess significance in two or more groups. Data were expressed as mean ± S.E. or mean ± S.D., and statistical significance was analyzed based on p < 0.05, p < 0.01, p < 0.001, or p < 0.0001.
Experimental results
Components of CSPG analyzed by LC-MS/MS
Fig. 3
LC-MS/MS Chromatographic Profiles of Identified Polyphenol Compounds in CSPG Extract: A. Representative chromatographic peaks of 32 compounds identified in the CSPG extract using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Each chromatographic curve represents the detection profile of an individual polyphenol standard matched with its corresponding peak in the CSPG extract. Peak alignment and ion fragmentation patterns were confirmed through LC-MS/MS analysis to verify compound identity. B. Quantitative Analysis of 32 Polyphenolic Compounds in CSPG Extract by LC-MS/MS (µg/g), the table lists the concentrations of 32 polyphenolic compounds detected in the CSPG extract using LC-MS/MS. Values are expressed in micrograms per gram (µg/g) of dry extract. Compound identification was confirmed by comparing retention times and mass spectra with authentic standards and database entries.
As a result of LC-MS/MS analysis of CSPG, a total of 32 components were identified (Fig. 3 A, B). These compounds—gallic acid, chlorogenic acid, naringenin, rutin, ferulic acid, riboflavin, formononetin, and protocatechuic acid—have all been reported to exert positive effects on gut health, which will be discovered in the following parts [31‐41].
CSPG extract regulates tight junction proteins and inflammatory markers in HT-29 cells
Fig. 4
Biological Network Analysis of CSPG’s Anti-inflammatory Activity: A. Biological networking explains the relationship between CSPG chemical compounds and anti-inflammatory activity: a high percentage of target interacted by 32 compounds from CSPG is related to Ulcerative colitis (200/312) within that, about one third (60/200) are related to inflammatory pathway, confirmed via KEGG pathway. B. Via KEGG enrichment, of the 200 targets mentioned, shows CSPG compounds can affect 4 pathways which are related to junction protein regulations
Anti-inflammatory interventions targeting tight junction (TJ) proteins recovery are crucial for managing UC [42‐47].
We conducted a network pharmacology study to explore the interactions between 32 bioactive compounds extracted from CSPG and 200 UC-related targets (Fig. 4A). These compounds include polyphenols and flavonoids such as quercetin, naringin, gallic acid, and resveratrol, which possess anti-inflammatory and gastroprotective activity. Target proteins are displayed in the right section with inflammation-related targets marked in red. Out of an initial pool of 312 candidate proteins, 200 relevant UC target proteins were identified as having been positively triaged with CSPG. Among those 200 genes, 60 are involved in inflammation-related pathways. This explains the efficacy of our drug samples to suppress inflammation. These proteins were filtered by cross-referencing the 3,738 genes from GeneCards to map out targets from data provided in SwissTargetPrediction. This indicates that CSPG produces its biological activity through several molecular mechanisms involving crucial inflammation target proteins.
The KEGG pathway enrichment analysis revealed the most relevant signaling pathways affected by CSPG compounds (Fig. 4B). The PI3K-Akt signaling pathway - a critical mechanism for cell death, inflammation and immune response modulation - appeared most significantly enriched, suggesting that CSPG compounds might find key focus in this pathway. Others include Focal adhesion, VEGF signaling, and Adherens junctions that are related to epithelial, angiogenic, and intercellular activities. The size of each dot indicates the magnitude of target genes associated with each pathway, with larger dots representing greater concern. This suggests that CSPG can be useful in treating UC because of its protective actions stemming from multi-pathway modulation through PI3K-Akt signaling, which is central to inflammation and epithelial restoration.
Key controllers of tight junction proteins recuperation, vital for the preservation of the intestinal barrier, include: The PI3K-Akt signaling pathway, Focal adhesion, the VEGF signaling pathway, and Adherens junction. PI3K-Akt signaling reduces inflammation and increases the expression of ZO-1 and occluding proteins which enhance the cellular tight junctions. Focal adhesion strengthens the stability of junctions by enhancing cell-matrix adhesion. While Adherens junctions improve epithelial intercellular connections, the VEGF signaling contributes to tissue maintenance and repair at the microvascular level. Modulating these pathways can lessen the inflammatory injury to the intestine barrier, thus serving as a potential treatment option for IBD [48‐51].
Fig. 5
Impacts of CSPG Extract on HT-29 cell protection, the status of the tight junction, and response to inflammation in TNF-α stimulated HT-29 cells. A. The cytotoxicity of CSPG extract was determined by performing cell viability assays in which HT-29 cells were treated with different doses of CSPG ranging from 0 to 600 µg/mL for 24 h and 48 h post treatment. It was noted that CSPG concentrations 25 and 50 µg/mL yielded insignificant cytotoxicity, whereas 100 µg/mL showed toxicity only when treated over 48 h. The right panel illustrates the effects on cell viability after 24 h and 48 h of TNF-α (30 ng/mL, similar with previous research) treatment, with or without CSPG (25, 50, 100 µg/mL) co-treatment. The results show that TNF-α showed more significant toxicity at 48 h, while 25, 50 µg/mL showed protection effect, 100 µg/mL showed toxicity, likely due to high dosage. B. Western blot analysis in HT-29 cells stimulated with TNF-α (30 ng/mL) and treated with CSPG (0, 25, 50, 100 µg/mL) for 24 h. Blots for COX-2, IL-6 (cell lysate), nuclear NF-κB p65, p-Akt, total Akt, ZO-2, and MUC2 are shown. β-actin was used as the loading control for COX-2, IL-6, ZO-2, and MUC2; Lamin B1 was the loading control for nuclear NF-κB p65. p-Akt was normalized to total Akt (p-Akt/Akt ratio). Bar graphs (right) show fold-change relative to untreated control (UT). TNF-α increased IL-6, nuclear NF-κB p65, and COX-2 and decreased ZO-2 and MUC2; CSPG attenuated inflammatory markers and restored ZO-2/MUC2, with significant effects at 50–100 µg/mL. CSPG also increased the p-Akt/Akt ratio. # p < 0.05 vs. UT; * p < 0.05, ** p < 0.01 vs. TNF-α-treated group
CSPG extract exhibited no significant cytotoxicity at concentrations below 100 µg/mL in 24 h; however, at 100 µg/mL cell viability decreased to around 85% only after 48 h. Higher concentrations (≥ 200 µg/mL) showed a dose-dependent decrease in viability. For CSPG 100 µg/mL that showed mixed result, significant lower cell viability only at 48 h; however, we still selected this concentration in later steps to see if this concentration can still show any protective effect, or it is really a toxicity over-dosage (Fig. 5A). Therefore, 25, 50, and 100 µg/mL were selected for further experiments.
In the next step, HT-29 cells viability were suppressed with TNF-α (30 ng/mL), consistent with previous publication, to mainly induce an inflammatory response, and resulted in only a slight decrease of viability in both 24 and 48 h: Only around 10% reduction in cell viability after 24 h, and around 20% reduction only after 48 h [52]. We observed cellular protection of CSPG: at 48 h after treatment, CSPG showed dosage dependent protection effect with 25 and 50 µg/mL, whereas 100 µg/mL showed toxicity, likely due to high dosage mentioned above.
The expression levels of key proteins related to tight junctions (ZO-2) and inflammation (COX-2, IL-6, NF-κB p65), and status of the PI3K pathway (observed via Akt activation) were assessed by Western blotting (Fig. 5B):
MUC2, a critical mucus-forming glycoprotein for epithelial defense, was upregulated after TNF-α stimulation in HT-29 cells, consistent with previous findings showing MUC2 mRNA induction via the MAPK pathway in these cells [53]. In our study, CSPG treatment at 25–50 µg/mL effectively suppressed MUC2 levels after being elevated by TNF-α, suggesting mucosal protection alongside barrier reinforcement.
In terms of inflammatory markers, TNF-α stimulation confirmed inflammatory up-regulation with increased COX-2, IL-6, and nuclear NF-κB p65. CSPG, especially at 50 µg/mL, reduced these markers in a concentration-dependent manner, demonstrating its anti-inflammatory potential.
Besides, TNF-α diminished the p-Akt/Akt ratio and CSPG was able to reverse this diminution, indicating activation of protective PI3K/Akt signaling, an important pathway that upregulates the tight junction barrier recovery [54]. We selected p-Akt as the checkpoint because it is a direct downstream target of PI3K and acts as a central regulator in the pathway [55]. Monitoring p-Akt levels gives a reliable indication of whether the PI3K-Akt pathway is functionally active (Fig. 5B). This also matches with the conducted pharmaco-informatic predictions above, that natural products in CSPG can interact with this pathway.
Next. We saw the decline in ZO-2 indicates insufficient barrier strength facing inflammation stress, which also align with previous studies [56‐58]. CSPG treatment restored ZO-2 expression in a dose-dependent manner, with greatest recovery at 50 µg/mL. We also tested ZO-1, but did not show a considerable increase, which continues to support the notion that ZO-2 may be a more precise indicator of epithelial stress in vitro, especially HT-29 cell when stimulated with TNF- α. ZO-1 only showed changes in in vivo model, in a later section of this data. We also added the Supplementary material for ZO1 in vitro experiment (Figure S1 – In vitro supplement), that showed treatment with TNF-α alone sharply diminished ZO-2 expression, a novel finding. Meanwhile, ZO-1 remained largely unchanged pattern consistent with previous findings in HT-29 cells [58].
These results suggest that CSPG extract can reduce inflammation and can likely restore epithelial barrier function primarily through ZO-2 modulation and Akt pathway activation, with an optimal dose of 50 µg/mL for efficacy with minimal cytotoxicity.
Effects of CSPG on intestinal microbiota before DSS exposure
Fig. 6
CSPG extract modulating the composition of Microbiota, before DSS exposure: A. Structural comparison of intestinal microbiota between Normal, 5ASA, CSPG50, CSPG100 groups at genus levels. B. Alpha diversity (pre-DSS, Days 1–14). Chao1 (richness) and Shannon (diversity) from fecal 16 S after 14 days of drug-only intake. C. Beta diversity analysis of the Normal(●), 5ASA(●), CSPG50(●) and CSPG100(●) groups. D. Pre-DSS (days 1–14) targeted taxa. Relative abundance of selected SCFA-linked or dysbiosis-associated taxa across groups (Normal, 5ASA, CSPG50, CSPG100). CSPG50 tended to preserve Ruminococcaceae_uncultured and Ruminococcus and reduce Gemella, compared with 5ASA. E. This table summarizes the most notable changes in gut bacteria before DSS treatment, when only pre-treat with 5ASA or CSPG. CSPG50 increased Ruminococcus and reduced harmful genera like Gemella, while 5ASA did not show meaningful changes. Ruminococcaceae_uncultured, a key butyrate-producing genus important for colon health, was maintained with CSPG50 but reduced by 5ASA. These changes suggest CSPG50 better supports a healthier microbial profile. All data are expressed as the mean ± SD (n = 6). *p < 0.05, **p < 0.01 as compared to normal. ↑, increase; ↓, decrease vs. Normal
This experiment was conducted after the first 2 weeks of drugs administration and prior to DSS exposure. This is to assess pre-treatment effects on systemic and mucosal immune readiness (Figs. 2, 6, 7 and 8). During this drug-only phase before DSS, we first looked at the overall population of gut bacteria across groups (Fig. 6A–D). The stacked bar plots showed that the composition of fecal bacteria was shifted by both CSPG and 5ASA, although the patterns were different. The 5ASA group appeared more dominated by a few taxa, while CSPG50 and CSPG100 showed broader changes, including an increase in some beneficial groups and reduction of others that are often linked with dysbiosis (Fig. 6A, D).
When alpha diversity was analyzed, a clear difference emerged. The Chao1 and Shannon indices were both significantly lower in the 5ASA group compared with Normal, showing that 5ASA reduced richness and evenness of the microbiota (Fig. 6B). CSPG50 maintained diversity at a level close to Normal, while CSPG100 showed lower richness (Chao1) and a modest reduction in Shannon diversity. This means that CSPG, especially at the lower dose, did not cause the same loss of baseline microbial diversity seen with 5ASA. This aligns with reported cases of long-term 5ASA intake can reduce gut microbial diversity and alter bacterial composition, even outside acute inflammation [7].
Beta diversity analysis using Jaccard and Bray–Curtis also supported these observations. The PCoA plot showed that the CSPG groups, both 50 and 100 mg/kg, separated along PC1 from Normal and 5ASA, which overlapped more closely (Fig. 6C). This indicates that CSPG changed the overall phylogenetic structure of the community at baseline, suggesting a reorganization of microbial abundance rather than just a loss of taxa.
CSPG50 showed a favorable impact on gut bacteria when compared to 5ASA. It preserved the butyrate-linked taxa Ruminococcaceae, Ruminococcus, and Lachnospiraceae_NK4A136 at levels close to normal, whereas 5ASA reduced these groups. CSPG50 decreased Gemella, while 5ASA produced no significant change in either. Taken together, these patterns indicate that CSPG50 supports butyrate-associated families and lowers dysbiosis-associated genera. Overall, CSPG50 seem to reshape the microbiota toward a steadier baseline with less diversity loss than 5ASA, this can provide a more stable community before DSS colitis challenge (Fig. 6D–E).
Effects of CSPG on fecal metabolites before DSS exposure
This experiment section, days 1 to 14, is the pre-DSS exposure duration. We assessed the drug effects on healthy individuals (Fig. 2A). A total of 32 metabolites were analyzed, including three SCFAs (Fig. 7B) and 29 other common metabolites. Among those 29 common metabolites, at this point, most significant changes were metabolites shown in the figure below (Fig. 7C). The rest are mentioned in Supplementary material for all metabolites profiling of fecal samples. Partial least squares discriminant analysis (PLS-DA) of GC-MS data was performed to study various metabolite profiles in feces. The PLS-DA model of stool samples showed a clear separation between the normal diet group and the CSPG intake group (Fig. 7A), which indicates that the metabolic profile of the CSPG intake group is different from that of the normal group.
Quantification of fecal butyric acid, propionic acid, and acetic acid levels, revealing not much significant differences among groups. CSPG also maintained SCFA levels closer to normal.
Proline and glycerol were higher in the CSPG group. On the other hand, alanine, 3-hydroxybutyric acid, phosphate, uracil, glyceric acid, succinic acid, and maltose were lower in the CSPG intake group (Fig. 7C). CSPG50 treatment led to significant metabolomic shifts compared to 5ASA group. Specifically, CSPG50 increased proline and glycerol—metabolites associated with mucosal repair and barrier protection—while reducing uracil, glyceric acid, and succinic acid, which are linked to oxidative stress and sustained inflammation. These changes were not observed with 5ASA, highlighting CSPG50’s better ability in promoting positive shifts in gut chemical balance (Fig. 7C).
A summarized table is also prepared (Fig. 7D). This is consistent with earlier reports suggesting that herbal medicines may promote a more favorable gut microbiome and metabolites when compared to 5ASA treatment, supporting their potential advantage in supporting gut microbial function; and more importantly less adverse effect if the drug is needed to be administered in long term, which is not rare in UC treatment [9, 63].
Fig. 7
Effect of CSPG on metabolites production in the feces, before DSS exposure: A. Supervised PLS-DA (partial least squares discriminant analysis) score plots derived from GC–MS fecal metabolite data, illustrating group separation among Normal(●), CSPG50(●) and 5ASA(●) groups. B. Concentrations of three measure main short-chain fatty acids (butyric acid, propionic acid, and acetic acid) in fecal samples measured by GC–MS. C. We also measured other common metabolites in fecal samples of the normal and intervention groups. D. This table summarizes the most notable changes observed in metabolites, before DSS exposure and treatment with CSPG50 or 5ASA. CSPG50 increased proline and glycerol, which are linked to mucosal repair. It also reduced uracil and glyceric acid, which are related to inflammation and oxidative stress. These benefits are not seen in 5ASA treatment. *p < 0.05, **p < 0.01, ***p < 0.001 as compared to normal. A false discovery rate of 5% was applied to all tests to correct for multiple testing
Correlation between gut microbiota and fecal metabolites before DSS exposure and related pathway analysis
To understand the relationship between fecal microbiota, a correlation matrix (|r| >0.67) was generated using Pearson correlations, from 37 gut bacteria and 32 metabolites (3 main SCFAs and 29 other metabolites) changes above. The correlation heatmap reveals that CSPG both shapes gut microbe–metabolite networks and maintains stability without UC induced by DSS. The taxon Ruminococcaceae_uncultured displayed strong positive associations with SCFA-related metabolites, including butyric acid and 3-hydroxybutyric acid (r ~ 0.70), and was also linked to phosphate and glyceric acid (r ~ 0.60) as well as Anaerotruncus (r ~ 0.80). A weaker positive correlation was observed with urea (r ~ 0.50). By contrast, Ruminococcaceae_uncultured exhibited pronounced antagonism with members of the pro-inflammatory repertoire, notably Gastranaerophilales and Lachnospiraceae_NK4A136_group (r ~ − 0.80). This pattern highlights the dual positioning of Ruminococcaceae_uncultured as a core SCFA-aligned taxon opposed to inflammatory microbial clusters (Fig. 8A).
Fig. 8
Correlation between microbiota and metabolites in feces of mice fed without or with CSPG (Normal and CSPG50 group): A. The color was according to the Pearson correlation coefficient distribution. Red represented a significant negative correlation; Blue represented significantly positive correlation. A false discovery rate of 5% was applied to all tests to correct for multiple testing. B. Bubble plot visualizing pathway topology analysis. Each circle represents a metabolic pathway. The x-axis indicates the pathway impact based on topology analysis, and the y-axis shows the significance level. Larger and darker circles indicate higher impact and significance. C. Table summarizing the top six enriched metabolic pathways affected by CSPG treatment. Match status, p-values, and pathway impact scores are listed. Only pathways with FDR-adjusted p < 0.05 and pathway impact > 0.1 were included
A comprehensive analysis of metabolic pathways was performed to ascertain the pathways affected by the extract of CSPG. This investigation employed both enrichment and topological evaluations to pinpoint the pertinent metabolic pathways. Significant pathways were identified based on their influence and the − log (p) values. A schematic representation illustrating the metabolic pathways impacted by the ingestion of CSPG was presented (Fig. 8B). The principal metabolic pathways that were affected encompassed the biosynthesis of valine, leucine, and isoleucine (match status 1/8; p < 0.020), butanoate metabolism (match status 1/15; p < 0.037), glycerolipid metabolism (match status 1/16; p = 0.040), pantothenate and CoA biosynthesis (match status 1/20; p = 0.049), the pentose phosphate pathway (match status 1/23; p = 0.019), and galactose metabolism (match status 1/27; p = 0.066) (Fig. 8C). Prior to DSS exposure, CSPG preserved butanoate metabolism and branched-chain amino acid biosynthesis, supporting SCFA production and mucosal energy balance. Additional enrichment of glycerolipid and pantothenate/CoA biosynthesis indicated metabolic conditioning toward barrier stability. In contrast, 5ASA showed weaker impact on these pathways and modest reductions in microbial diversity.
Colitis inhibitory effect of CSPG extract on in vivo model
Fig. 9
Protective effect of CSPG against DSS-induced colitis in mice: A. Colon length was measured after DSS treatment. DSS-induced colitis significantly reduced colon length compared to the normal group, while treatment with 5ASA and CSPG (50 or 100 mg/kg/day) effectively attenuated this shortening, indicating anti-inflammatory effects. B. Procalcitonin (PCT) levels were measured using ELISA. PCT levels were markedly elevated in the DSS group, reflecting systemic inflammation. Both 5ASA and CSPG treatments significantly reduced PCT levels, with CSPG showing a dose-dependent effect. C. Representative Hematoxylin and Eosin (H&E)–stained colon tissue sections collected after DSS administration. Histological images show the degree of mucosal damage and recovery across experimental groups. Compared to the normal group, DSS-induced mice exhibited severe epithelial disruption, mucosal layer thinning, and crypt structure loss. CSPG50 and 5ASA treatments groups showed more normal monology. D. Western blot analysis showing the expression levels of iNOS and COX-2 (inflammatory markers), ZO-1, ZO-2, ZO-3, Occludin, Claudin-2 (tight junction proteins), and MUC1, MUC2, MUC4 (mucins) in colon tissues from Normal, UT + DSS, CSPG50 + DSS, CSPG100 + DSS, and 5ASA + DSS groups. Bar graphs represent the relative protein expression normalized to β-actin and expressed as fold change compared to the Normal group. DSS treatment increased iNOS and COX-2 levels while reducing tight junction proteins and mucins. CSPG treatment, particularly at 50 mg/kg, significantly suppressed inflammatory markers, restored tight junction proteins, and enhanced mucin expression, indicating protective effects on intestinal barrier integrity and inflammation. Statistical significances were determined using two-way ANOVA. # p < 0.05 as compared to normal, *p < 0.05, **p < 0.01 as compared to UT + DSS. Data were expressed as mean ± SEM (n = 6)
To investigate CSPG’s anti-inflammatory action in a living animal, we measured colon length, procalcitonin (PCT), and tissue slides after DSS-induced colitis. DSS cut colon length compared with controls (p < 0.01), showing classic swelling and mucosal shortening. CSPG50 and CSPG100 groups partially restored length, matching the clinically used drug 5ASA (Fig. 9A).
Acute proteins such as procalcitonin (PCT) and C-reactive protein (CRP) are the most widely used inflammatory biomarkers in clinical practice [70‐75]. During infection, PCT is widely released from non-endocrine tissues. PCT levels have been reported to have superior diagnostic accuracy in assessing disease activity in severe ulcerative colitis compared to CRP levels, ESR, and WBC [71]. Moreover, while PCT and CRP are both acute phase response proteins, PCT has been reported to be a more marker of bacterial infection [76‐78]. Severe UC is characterized by bacterial invasion due to a defensive deficiency of the mucous membrane. Therefore, PCT levels of intestinal tissue were analyzed by ELISA to assess the activity of mucosal damage and ulcerative colitis (Fig. 9B). The PCT level of normal tissue was 0.021 ± 0.002 ng/mL and the colitis-induced DSS group was significantly increased to 0.795 ± 0.311 ng/mL, which is judged to be severe ulcerative colitis (UC). On the other hand, CSPG50 + DSS: 0.281 ± 0.265 ng/mL, CSPG100 + DSS: 0.068 ± 0.035 ng/mL were reduced in the experimental group that took CSPG extracts. In the CSPG50 + DSS group, PCT levels were reduced to 0.281 ± 0.265 ng/mL, while the positive control 5ASA + DSS group showed a significant decrease to 0.212 ± 0.014 ng/mL, confirming the efficacy of CSPG extract in protecting the intestinal mucosa and inhibiting colitis.
Ten days after administering DSS (day 24), histological sections of colon tissue were prepared and stained with H&E to evaluate mucosal integrity and the extent of inflammation (Fig. 9C). In the DSS-only cohort, damage was severe: epithelium was eroded, crypts were destroyed, goblet cells were depleted, and a dense infiltrate of immune cells appeared, all classic indicators of acute colitis. By contrast, sections from the CSPG-treated animals showed marked improvement, especially at the 50 mg/kg/day dose, where crypt architecture and epithelial continuity were notably preserved. These findings mirror earlier reports using the same DSS model, which documented comparable mucosal injury and its moderation by other therapies [79, 80]. Collectively, the data support the conclusion that CSPG protects the colon by stabilizing epithelial barriers and curtailing inflammation-driven tissue damage.
The intracellular scaffolding ZO proteins, ZO-1, ZO-2, and ZO-3 are integral to linking actin filaments of the cytoskeleton to transmembrane tight junctions, thus controlling paracellular permeability. Reduced expression of Zonula Occludens proteins has been observed in ulcerative colitis, associated with dysfunctional barrier [56, 81, 82]. In our model, DSS treatment significantly suppressed all three ZO proteins. Treatment with CSPG restored ZO-1, ZO-2, and ZO-3 expression in a dose-dependent manner peaking at 50 mg/kg.
Occludin, a critical transmembrane protein responsible for preserving the integrity of the tight junction barrier, is essential for regulating intestinal epithelial barrier function. Its downregulation contributes to heightened permeability and increased risk of inflammation in colitis models [83]. In our analysis, Western blotting confirmed a marked drop in Occludin expression resulting from DSS administration, which suggests that there was impaired epithelial barrier function.
Claudin-2 is a tight junction protein which forms pores, increases cation permeability, and is often overexpressed in inflammatory bowel disease, exacerbating barrier dysfunction. Increased expression of Claudin-2 is viewed as an indicator of tight junctions’ pathology and mucosal inflammation [84]. Within the DSS group, Claudin-2 levels were considerably elevated, confirming barrier disruption. Both doses of CSPG treatment substantially reduced Claudin-2 expression toward normal levels, demonstrating functional recovery of selective permeability restoration after partially normalizing tight junction remodeling. This regulatory effect was comparable to that observed in the control group treated with 5ASA.
MUC1, MUC2, and MUC4 are major constituents of mucus secreted by goblet cells that create an intestinal mucus barrier; they offer initial defense against intraluminal organisms as well as physical abrasion. In colitis, both MUC2 and MUC4 are downregulated while low-expressing MUC1 is upregulated—resulting in loss-of-function changes in the protective mucosal barrier [85]. In this study, exposure to DSS led to dysregulation of mucin homeostasis by reducing MUC2 and MUC4 along with increased expression of MUC1. Restoration of these pathways was achieved through CSPG supplementation which significantly restored decreased values of both MUC2 and MUC4 while downregulating aberrant expressed MUC1 augmented expression demonstrating ability towards facilitation towards the healing response alongside immune system equilibrium disturbance rectification (Fig. 9D).
Effect of CSPG on intestinal microbiota in colitis after DSS exposure
In order to investigate the effect of CSPG on the intestinal microbiota in the process of DSS-induced colitis, feed mixed with extract was fed autonomously for 2 weeks, and 1.5% DSS was supplied as drinking water while the extract was continued to be fed and colitis was induced (Fig. 9). Bacterial 16 s rRNA gene sequencing of fecal samples was performed. The alpha diversity index Chao1, reflecting microbial richness, which looks at the differences between groups in how much fecal microbes each group had, had a significant difference from the normal group (Fig. 10 A). The alpha diversity Shannon index, reflecting microbial evenness, which measures differences in the uniformity of bacteria within a group, also showed a significant difference from the normal group after treatment with DSS, indicating that overall community evenness was reduced and not fully restored (Fig. 10B). In the evaluation of the beta diversity, which looks at differences between groups, analysis using the Jaccard distance PCoA technique showed a clear difference between the normal group and the DSS-fed group, and a clear difference between the DSS-only group and the CSPG-fed group with DSS (Fig. 10 C). Thus, although Shannon evenness remained suppressed (alpha diversity), the beta diversity analysis demonstrated that CSPG shifted the overall community composition away from DSS-only and toward the normal microbiota structure. These results show that CSPG affects the overall diversity and composition of the intestinal microbiota. Compared to the DSS alone, there was a genus-level difference in gut bacteria in the group fed CSPG with DSS (Fig. 10 A).
Fig. 10
CSPG extract modulating the composition of Microbiota after DSS exposure:A. Genus-level community composition (stacked bars) for Normal, UT + DSS, 5ASA + DSS, CSPG50 + DSS, and CSPG100 + DSS. B. α-diversity (Chao1 richness; Shannon diversity). DSS reduced diversity vs. Normal; CSPG and 5ASA partially restored it. C. Beta diversity analysis of the Normal(●), UT+DSS(●), 5ASA+DSS (●),CSPG50+DSS (●) and CSPG100+DSS (●) groups. Both CSPG doses shifted samples toward Normal; 5ASA showed a smaller shift. D. CSPG increased Prevotellaceae UCG-001 and Ruminococcus and decreased [Eubacterium] siraeum group and Erysipelotrichaceae. E. This table summarizes the most notable changes in gut bacteria. Generally, CSPG50 showed better improvement overall compared to 5ASA. All data are expressed as the mean ± SD (n=6), # p<0.05 as compared to Normal. *p< 0.05, **p< 0.01, ***p< 0.0001 as compared to UT+DSS.
We present microbial outcomes for the DSS-treated (therapeutic) phase. At the genus level, DSS alone (UT + DSS) shifted the community away from Normal, expanding opportunistic/inflammation-associated taxa and reducing health-linked commensals. CSPG shifted the community toward a healthier profile, whereas 5ASA showed a more modest effect. CSPG50 and CSPG100 increased the relative abundance of Prevotellaceae UCG-001 and Ruminococcus and reduced dysbiosis-associated taxa such as [Eubacterium] siraeum group and Erysipelotrichaceae (Fig. 10 A, D–E).
α-diversity dropped with DSS (lower Chao1 and Shannon). Post-intervention, CSPG and 5ASA partially restored α-diversity; differences between the two treatments were not significant, and values remained below Normal (Fig. 10 A, B). β-diversity showed clear separation between UT + DSS and Normal (PC1 15.0%, PC2 7.3%). Both CSPG doses shifted centroids toward Normal, whereas 5ASA clustered closer to UT + DSS, indicating a stronger global community correction by CSPG (Fig. 10 C).
Targeted genus/family comparisons were consistent with these patterns: Prevotellaceae UCG-001 and Ruminococcus increased with CSPG (more at 50–100 mg/kg than with 5ASA), while [Eubacterium] siraeum group and Erysipelotrichaceae decreased (Fig. 10D–E). Together, these data indicate that CSPG promotes a microbiota state closer to Normal and more compatible with SCFA-supported barrier function during DSS colitis. These findings align with reports that certain herbal therapies facilitate microbiota recovery more effectively than 5ASA [63].
Effects of CSPG on fecal metabolites in colitis recovery after DSS exposure
Targeted GC–MS profiling of fecal extracts quantified 32 metabolites (three SCFAs and 29 additional metabolites). A supervised PLS-DA model showed clear separation between UT + DSS and CSPG50 + DSS (PLS1 ≈ 71–72%; PLS2 ≈ 21–24%) and a smaller separation between UT + DSS and 5ASA + DSS (Fig. 11A), indicating a broader metabolic shift under CSPG treatment. Among those 29 common metabolites, at this point, most significant changes were metabolites shown in the figure below (Fig. 11C). The rest are mentioned in Supplementary material for all metabolites profiling of fecal samples.
Short-chain fatty acids (SCFAs). DSS reduced acetic, propionic, and butyric acids relative to Normal. Both CSPG doses increased SCFAs versus UT + DSS, with CSPG50 + DSS and CSPG100 + DSS frequently showing higher medians than 5ASA + DSS (Fig. 11B). These SCFAs are central to epithelial energy supply and barrier-supporting, anti-inflammatory signaling [89‐94].
Broad metabolite panel. Several metabolites linked to inflammation or epithelial stress were favorably shifted by CSPG (Fig. 11C). Succinic acid and uracil—often elevated in colitis—were reduced versus UT + DSS. In contrast, metabolites associated with mucosal repair and balanced host–microbe metabolism increased, including glycine, proline, valine, lactic acid, pyruvic acid, and myo-inositol. Alanine, which was high in UT + DSS, declined toward Normal with treatment. Across many readouts, CSPG50 + DSS displayed equal or greater recovery than 5ASA + DSS (see medians in Fig. 11B–C). A side-by-side summary of directional effects and literature-based roles is provided in Fig. 11D; the complete list of 32 metabolites is reported in the Supplementary material.
Overall, these metabolite shifts parallel the microbiota changes observed in Fig. 10, where Prevotellaceae UCG-001 and Ruminococcus (SCFA-associated taxa) increased with CSPG. Together, the data indicate that CSPG promotes a fecal metabolite profile closer to Normal during DSS colitis, with broader restoration of SCFAs and multiple non-SCFA metabolites than seen with 5ASA [63].
Fig. 11
Effect of CSPG on fecal metabolite profile, after DSS exposure: Supervised partial least squares discriminant analysis score plot derived from the GC-MS data of samples of UT+DSS(●), CSPG50+DSS(●) and 5ASA+DSS(●) groups. A. PLS-DA score plot of fecal metabolites for the UT+DSS and CSPG50+DSS (top) as well as the UT+DSS and 5ASA+DSS (bottom) groups showing PLS1 (71.9%/72.2%) and PLS2 (24.0%/21.7%). For further information about model validation, please consult Methods. B. Boxplots showing the amounts of short-chain fatty acids (acetic, propionic, and butyric) measured using GC-MS following alkyl-chloroformate derivatization. SCFAs were lower in DSS treatment as compared to the Normal group; both doses of CSPG were higher than SCFAs as compared to UT+DSS and in some instances, were higher than median 5ASA. C. Boxplots depicting additional fecal metabolites analyzed using GC- MS including but not limited to succinic acid, glycine, proline, valine, lactic acid, pyruvic acid, myo-inositol, alanine, and uracil. Increased levels of succinic acid and uracil were observed in DSS treatment as well as decreased levels of other metabolites. CSPG treatment shifted metabolite concentrations toward Normal as compared to UT+DSS. D. Summary table depicting the directional differences in the effect of CSPG50 compared to 5ASA along with some of their functional roles. A full list of 32 metabolites (comprising 3 SCFAs and 29 other metabolites) is included in the Supplementary material. A false discovery rate of 5% was applied to all tests to correct for multiple testing. #p <0.05 as compared to Normal. *p < 0.05, **p < 0.01 as compared to UT+DSS
Fig. 12
Correlation between microbiota and metabolites in fecal samples of Normal vs UT+DSS vs CSPG50+DSS groups: A. The color was according to the Pearson correlation coefficient distribution. Red represented a significant negative correlation; Blue represented significantly positive correlation. A false discovery rate of 5% was applied to all tests to correct for multiple testing. B&C. Metabolic pathways comparing Normal vs UT+DSS vs CSPG50+DSS
CSPG impact on correlation between intestinal microbiota and fecal metabolites after DSS exposure and pathway interaction analysis
To understand the relationships among fecal metabolites, a correlation matrix (|r| >0.65) was generated using Pearson correlations from 34 gut bacteria and 32 metabolites data collected in this stage (Fig. 13A). Succinic acid showed a strong inverse correlation with propionic acid (r = − 0.831) and positive correlations with thymine (r = 0.783) and glycerol (r = 0.749); it was also negatively correlated with linolenic acid (r = − 0.742). Pentitol was positively correlated with propionic acid (r = 0.767) but negatively correlated with succinic acid (r = − 0.733) and 3-hydroxybenzoic acid (r = − 0.721). Thymine displayed negative correlations with both propionic acid (r = − 0.730) and linolenic acid (r = − 0.727). Glycerol was negatively correlated with propionic acid (r = − 0.703) while remaining positively linked to succinic acid (r = 0.749).
Additional noteworthy pairs included butane-2,3-diol, which was positively correlated with propionic acid (r = 0.683) and negatively correlated with glyceric acid (r = − 0.740) and glycerol (r = − 0.659). Valine correlated positively with ribose (r = 0.720), lactic acid with threonine (r = 0.681), glycolic acid with leucine (r = 0.679), and serine with pentitol (r = 0.680), whereas glycolic acid was inversely related to uracil (r = − 0.667). Together, these associations highlight coordinated shifts in energy, nucleotide, and amino-acid metabolism during the colitis process (Fig. 13A).
Fig. 13
Crosstalk between polyphenol-mediated microbiota restoration and colon healing pathways. A. Comparative summary showing comprehensive effects of CSPG50 (a polyphenol-rich extract) versus 5ASA in modulating microbiota, metabolites, and inflammation at two key time points in DSS-induced colitis. CSPG50 demonstrates superior effects in enriching SCFA-producing bacteria, suppressing inflammatory metabolites. B. Mechanistic diagram illustrating how CSPG50 disrupts the inflammatory loop in ulcerative colitis. CSPG50 suppresses inflammation, restores gut barrier proteins (ZO-1, occludin), promotes beneficial microbiota and facilitates recovery of SCFAs (butyrate, acetate, propionate), ultimately leading to colon healing and remission. C. Summary table highlighting key microbiota–metabolite–barrier cross-talks mediated by CSPG polyphenols. Each row describes a specific axis, such as how polyphenols promote beneficial microbes, SCFA recovery, barrier restoration, and inflammation resolution, supported by relevant citations
To identify the metabolic pathways affected by CSPG extract, metabolic pathway analysis was conducted using enrichment and topology-based approaches. Key pathways were selected based on pathway impact and statistical significance (− log p). The most significantly affected pathways included Galactose metabolism (match status 5/27, p < 0.00001), Valine, leucine, and isoleucine biosynthesis (3/8, p < 0.0001), Pyruvate metabolism (3/18, p < 0.001), Glyoxylate and dicarboxylate metabolism (3/32, p < 0.005), Glycine, serine, and threonine metabolism (3/33, p < 0.005), Butanoate metabolism (2/20, p < 0.02), and Glycerolipid metabolism (2/23, p < 0.03) (Fig. 13B&C). The pathway impact ranking emphasizes SCFA and amino acid metabolism as central hubs of CSPG’s effect. These convergent data indicate that CSPG’s polyphenol matrix enhances microbial functions that sustain epithelial energy use, antioxidant defense, and immune balance, indicating broader metabolic normalization than 5ASA in reconstituting metabolic homeostasis.
Discussion
This study provides compelling evidence that polyphenol-rich CSPG extract exerts multi-targeted effects in a murine model of UC, demonstrating its potential as a complementary therapeutic approach. The findings align with a growing body of research highlighting the anti-inflammatory, gut barrier-enhancing, microbiota-modulating, and metabolomic-regulating properties of polyphenol-based interventions. While conventional UC treatments, such as 5ASA and corticosteroids, primarily target inflammation, CSPG extract appears to work through broader mechanisms, offering additional benefits in restoring gut homeostasis and microbial balance. This highlights the potential advantages of multi-targeted natural therapeutics over single-pathway pharmaceutical interventions, particularly in chronic inflammatory conditions like UC, which involve complex interactions between immune dysregulation, epithelial dysfunction, and gut microbiota disturbances. A structured overview of CSPG’s advantages over 5ASA is summarized (Fig. 13A-C).
Chronic inflammation plays a central role in UC pathogenesis, given that polyphenols are known to modulate inflammation through numerous signaling pathways. This study shows that CSPG extract significantly decreased levels of key pro-inflammatory cytokines such as TNF-α, IL-6, and IL-1β, which corroborates previous studies, suggesting that polyphenol-rich extracts lower NF-κB activation as well as NLRP3 inflammasome. There is evidence supporting active suppression of inflammation by polyphenolic NLRP3 antagonists [108, 109]. Similar effects have been reported from extracts of mango polyphenols, which also suppress inflammation via the PI3K/AKT/mTOR signaling pathway; and from cherry polyphenol extract, which mitigates colitis through the Wnt/β-Catenin pathway [10, 109]. These mechanisms suggest that polyphenols can modulate diverse inflammatory pathways simultaneously, offering advantages over single-target pharmaceutical therapies. The observed reduction in pro-inflammatory cytokines suggests that CSPG extract may not only suppress acute inflammatory responses but also have long-term immune-modulating effects that could help maintain remission in UC patients.
Emerging evidence strengthens the claim that polyphenols exert targeted control over epithelial signaling by directly modulating the PI3K-Akt pathway, which is implicated in both barrier integrity and inflammation. Flavonoids and other phenolic compounds have demonstrated the ability to inhibit PI3K/Akt phosphorylation, thereby suppressing TNF-α–driven cytokine cascades and enhancing tight junction expression [110, 111]. In this context, compounds like baicalin and dihydroartemisinin have been reported to mitigate inflammation through dual inhibition of PI3K-Akt and NF-κB pathways, which are jointly responsible for epithelial disruption in colitis models [112]. These actions lead to an upregulation of ZO-1/ZO-2 and occludin, with normalization (reduction) of claudin-2—markers of tight-junction restoration - which aligns with the observed effects of CSPG in this study. Other studies also highlight how plant-derived polyphenols such as mango and pomegranate extracts attenuate epithelial leakage and reduce colonic damage via modulation of the PI3K-Akt/mTOR axis [45]. These findings reinforce CSPG’s mechanism of action by showing that multi-targeted suppression of inflammatory signaling can directly improve epithelial junction resilience.
Taken together, the CSPG-induced effects on tight junction recovery and inflammation may be attributed, at least in part, to the collective actions of its polyphenolic compounds, which act synergistically on signaling pathways like PI3K-Akt and focal adhesion—both central to colonic barrier maintenance.
CSPG extract not only augments the antioxidant defenses, which are essential in alleviating oxidative stress-related damage to the colon but also contributes to cytokine suppression. This aligns well with studies showing that polyphenols from pomegranate activate Nrf2 signaling, thereby enhancing the gut barrier [113]. CSPG extract exhibits antioxidant properties that mitigate epithelial damage and mucosal ulceration. which helps in mitigating the epithelial damage and mucosal ulceration caused by oxidative stress injury in UC. CSPG contains polyphenols such as gallic acid, baicalin, and naringenin that reduce the generation of reactive oxygen species (ROS) and improve the functioning of mitochondria in intestinal epithelial cells [114]. Such antioxidant activity reduces oxidative damage and epithelial permeability, reinforcing CSPG’s role in gut barrier protection.
A distinguishing feature of UC is a dysfunctional gut barrier which permits luminal toxins and bacteria to aggravate inflammation. The study discovered that the CSPG extract upregulates key epithelial integrity maintaining proteins such as ZO-1 and occludin that are vital for the preservation of epithelial integrity. In the same manner, polyphenol-rich interventions are known to increase tight junction expression and decrease intestinal permeability [115]. As an illustration, extracts from licorice that are abundant in flavonoids have shown protective effects through the stabilization of ZO-1 and occludin that precludes the damage of epithelial tissues in the colon [14].
Tight junction (TJ) proteins, including ZO-1 and occludin, are crucial for maintaining intestinal barrier integrity in inflammatory bowel disease (IBD). Inflammation-driven dysregulation of these proteins increases intestinal permeability, exacerbating disease progression [116, 117]. TNF-α reduces ZO-1 expression, while occludin downregulation further disrupts the barrier [118, 119]. COX-2 overexpression worsens TJ dysfunction, but its inhibition restores barrier integrity [82, 120]. Targeting these pathways through anti-inflammatory strategies may enhance intestinal protection and serve as a therapeutic avenue for IBD treatment [121, 122].
These findings reinforce the notion that CSPG extract may prevent UC progression by strengthening intestinal barrier function.
Dysbiosis, defined as a disproportionate ratio of good and bad bacteria in the gut, is important in the development of UC. This study shows that CSPG extract increases the beneficial microbial populations and lowers inflammatory taxa which was in accordance with earlier studies on polyphenol-microbiota interactions [11, 123]. Recognizing different types of quercetin monoglucosides has also increased microbiome diversity and facilitated microbial gut homeostasis [124]. It is important to note that the CSPG extract is capable of shifting the microbial composition and seems to restore functional activity, which is an extremely important aspect in UC remission and disease maintenance.
The restoration of Prevotellaceae UCG-001, and Ruminococcus populations observed in CSPG-fed groups suggests these genera play synergistic roles in gut healing. Lactobacillus, a well-established probiotic, exerts barrier-protective and anti-inflammatory actions through the release of immunomodulatory extracellular vesicles and stimulation of IL-22 production from innate lymphoid cells [125, 126].
Similarly, Prevotellaceae UCG-001, often enriched by fiber-rich diets, contributes to SCFA biosynthesis and may shift the local metabolic profile toward a more anti-inflammatory milieu [127, 128]. Its association with improved glucose and lipid metabolism also suggests a broader metabolic impact beyond the gut.
Ruminococcus, known for fermenting complex polysaccharides, produces acetic and formic acids—metabolites crucial to epithelial energy supply. Members of Ruminococcus ferment complex polysaccharides and contribute to SCFA pools [129, 130].
CSPG’s apparent ability to elevate these microbial community, hints at a microbiome reshaping strategy, unlike 5ASA or corticosteroids which often suppress microbial diversity [7]. This microbial enrichment may underline the anti-colitic effects observed, including reduced COX-2/iNOS and improved epithelial markers.
Butyrate, one of the most potent SCFAs, is essential for maintaining gut barrier function. It enhances ZO-1 and claudin-1 expression and serves as an energy source for colonocytes, thereby facilitating wound healing in inflamed tissue [131, 132]. Importantly, butyrate inhibits HDACs, leading to chromatin remodeling that promotes epithelial survival and IL-10 receptor expression [133, 134]. Propionate also contributes to mucosal integrity and has been found to block IL-8 and TNF-α expression in cytokine-stressed environments, thereby lowering the inflammatory load [135, 136]. Beyond direct effects on epithelial cells, SCFAs play immunological roles by modulating inflammasome activity (e.g., NLRP3) and promoting the differentiation of Tregs. These Tregs enhance IL-10 production, reinforcing tolerance in mucosal tissues [110, 137].
Given that CSPG increased SCFA levels in fecal samples, these immunological and barrier-repairing benefits are likely integral to its therapeutic action. SCFA production thus bridges the observed microbial shifts with improved gut integrity.
In previous research, Ferulic acid treatment augmented Ruminococcus alongside other beneficial taxa which is consistent with our Ruminococcus rebound under CSPG [20]. Complementary to this, dietary tannic acid encouraged Ruminococcaceae in animal microbiota studies [138]. biochanin A [139], caffeic acid [36], Chlorogenic acid [140] enhance the genus of the family Prevotellaceae UCG-001 which is an important SCFA producer. At the family level, an LPS-inflammation model showed ferulic acid significantly promoting Prevotellaceae, which supports the recovery of the broader SCFA-guild [141]. Extracts rich in polyphenols (e.g. pomegranate peel) reduced Erysipelotrichaceae in vivo, which suggests the responsiveness and suppression of polyphenols as dysbiosis resolves [142]; this is consistent with our Erysipelotrichaceae decline during CSPG treatment.
Concerning the interaction of polyphenols and metabolites in UC, gallic acid treatment increased total SCFAs and acetic acid in feces [143]; chlorogenic acid could reversed dysbiosis and enriched SCFA-linked bacteria [144]; naringenin increased propionate during fermentation [145]; and rutin enriched barrier- and SCFA-producing taxa [38] and, with chlorogenic acid, increased glycine [146, 147] Gallic acid also reduced the inflammatory metabolite succinic acid in cases of colitis [148]. Altogether, these polyphenol–microbe–metabolite interactions reinforce the therapeutic promises of CSPG in UC.
In addition, the polyphenols such as those in Fu Brick tea are known to stimulate the formation of indole-3-acetic acid, a microbial metabolite known to boost immune response and alleviate colitis [149]. These findings validate the initial assumption that polyphenols have anti-inflammatory properties and modify the gut microbiota for better disease outcomes [149].
CSPG extract vs. conventional UC treatment – 5ASA has often altered the gut microbiota functionality, whereas corticosteroids are frequently associated with microbial dysbiosis [150, 151]. It is worth mentioning that a clinical trial that compared mesalazine (5ASA) with herbal polyphenol preparation showed that the formulation provided greater additional effects by stabilizing the microbiota, which may offer an advantage over other standard therapies [152]. Network-pharmacology predictions are consistent with the Western-blot pathways: PI3K-Akt and inflammation-related signaling. This connection serves to corroborate the Western blots since the predicted targets, particularly within the PI3K-Akt and inflammation related pathways, had their functional protein expression changes. These results justify the multifaceted effects of CSPG and provide mechanistic understanding of its modulation of inflammatory and epithelial pathways in UC.
Microbes have important metabolites like butyrate, acetate, and propionate that modulate inflammation, immunity, and epithelial functions. This work showed that CSPG extract improved the activity of SCFA-related metabolic pathways, indicating that it has a positive effect on microbial metabolic function restoration. Earlier studies indicate that SCFAs have the ability to activate GPR41/GPR43 receptors which in turn increase the production of anti-inflammatory cytokines [153]. In UC patients, lower SCFA levels showed an association with higher disease activity, strengthening the rationale of microbiota-based interventions [154].
Beyond SCFA production, CSPG-driven microbiota modulation may exert therapeutic effects through a range of microbial metabolites that influence immune cell differentiation, cytokine production, and epithelial homeostasis. These include tryptophan metabolites, vitamin A derivatives like retinoic acid, and SCFA-mediated signals. SCFAs such as butyrate and propionate act through G-protein coupled receptors (e.g., GPR43/FFAR2 and GPR41/FFAR3) expressed on immune cells including macrophages and T cells [155, 156]. Receptor activation shifts immune signaling toward anti-inflammatory profiles, promoting IL-10 production and dampening the expression of TNF-α and IL-6—two cytokines central to UC pathogenesis. Butyrate specifically promotes macrophage polarization from pro-inflammatory M1 to regulatory M2 phenotypes, aiding in mucosal healing and tissue repair [157]. Propionate, meanwhile, enhances the differentiation of regulatory T cells (Tregs), further supporting immune tolerance in the inflamed gut [158].
Another layer of immune modulation is provided by microbial tryptophan metabolism. Indole derivatives, especially indole-3-acetic acid (IAA), generated by gut bacteria, play a central role in Treg induction and IL-10 secretion [159]. These effects are particularly relevant in UC, where Treg deficiency and elevated pro-inflammatory cytokine profiles drive disease severity. Dysregulation of tryptophan metabolism is a known feature in UC patients and correlates with increased inflammation [160]. Retinoic acid, derived from vitamin A and metabolized by gut bacteria, represents yet another critical signal in immune regulation. It suppresses Th1/Th17 cell expansion while promoting Treg differentiation and IL-10 production—mechanisms vital to controlling chronic intestinal inflammation [161]. In inflammatory models, retinoic acid supplementation restores immune balance, underscoring the microbiome’s role in maintaining vitamin A metabolism.
Recent findings also highlight that microbial metabolites influence energy metabolism in immune cells, altering their functional fate. SCFAs can promote oxidative phosphorylation over glycolysis in T cells and macrophages, reinforcing a regulatory rather than inflammatory phenotype [162].
Altogether, these findings suggest that CSPG-induced microbial remodeling not only impacts the gut barrier and metabolome but also broader immunological reprogramming. This metabolite-driven crosstalk offers a compelling explanation for the systemic immunomodulatory effects observed in CSPG-treated colitis models.
Although expectations from these findings are positive, there are challenges that still exist concerning the application of polyphenols in clinical practice. One key issue is bioavailability, as a large number of polyphenols have low absorption and high metabolism rates [163]. Another challenge is this particular variation in the microbiota structure of individuals, which impacts the metabolism of polyphenols [10].
Using combination therapies could improve the efficacy of CSPG extract even further. The combination of polyphenols with probiotics or prebiotics may provide a joint benefit for microbial and gut barrier protection. In addition, further study of polyphenol-drug interactions is necessary since some components could affect the bioavailability or effect of the active UC drugs [164].
These findings support further exploration of polyphenol-based therapeutics as adjuncts or alternatives to conventional UC treatments. However, addressing challenges related to bioavailability, formulation optimization, and clinical validation will be essential for translating these findings into effective human therapies. Future research should focus on improving polyphenol delivery systems, exploring combination treatments, and conducting well-designed clinical trials to establish commercialization of polyherbal extracts such as CSPG’s potential.
This work is one of the few to directly compare a standardized herbal mixture with 5ASA, the current clinical standard, across both prevention and treatment models. That rare head-to-head comparison creates a reference baseline for herbal–drug evaluation, highlights specific molecular and microbial pathways for targeted follow-up, and provides a reproducible methods framework. These findings can guide and accelerate future causal, mechanistic, and translational studies. Although we evaluated CSPG in both prophylactic (pre-DSS) and therapeutic (post-DSS) settings, some limitations remain. Deeper work with pathway inhibition, knockdown models, confocal imaging is a good, recommended direction in the future. Our microbiome and metabolome data rely on 16 S sequencing and targeted GC-MS, which constrain resolution; future studies should apply shotgun metagenomics, meta-transcriptomics, and untargeted LC-MS/MS. In addition, we used only male BALB/c mice under a single diet and housing condition, and CSPG remains a multi-component botanical without fractionated active identification or pharmacokinetic data.
Furthermore, implementing an advanced drug delivery system can also help our formula. Recent developments in nanotechnology, particularly the utilization of phospholipid-based nanocarriers, have been investigated to enhance bioavailability and facilitate more effective delivery [165]. This aim of implementing an enhanced drug delivery system can also be the next step of our research to compare effects from different preparations forms of CSPG.
Conclusion
In conclusion, this study provides strong evidence that CSPG extract—a polyphenol-rich combination of four traditional herbs—has a multitarget protective effect against ulcerative colitis. The extract significantly attenuated the inflammatory response by downregulating IL-6, COX-2, and nuclear NF-κB p65 in HT-29 cells; in vivo we observed reductions in COX-2/iNOS and restoration of barrier proteins. It also restored intestinal barrier integrity by enhancing the expression of key tight junction proteins (ZO-1, ZO-2, Occludin) and mucins (MUC2, MUC4). CSPG activated the PI3K/Akt signaling pathway, a key regulator of epithelial healing.
Additionally, gut microbiota analysis showed that CSPG increased the abundance of SCFA-producing beneficial bacteria such as Prevotellaceae UCG-001 and Ruminococcus, while decreasing harmful species such as Erysipelotrichaceae, Gastranaerophilales, and the [Eubacterium] siraeum group. Corresponding changes in metabolite profiles, especially increases in butyric acid, proline, and glycerol, indicate improved intestinal metabolic balance.
Together, these findings highlight the therapeutic potential of polyphenol-containing herbal formulations in modulating inflammation, gut microbiota, and intestinal barrier function, opening promising avenues for future colitis management strategies.
Acknowledgements
The authors would like to express their sincere gratitude to the researcher Huy Hieu Phung, from the Biomedicine Discovery Institute and Department of Biochemistry and Molecular Biology, School of Biomedical Sciences, Monash University, Australia, for his invaluable assistance in drawing the pharmacology network for this study. His expertise and contributions significantly enhanced the analysis and visualization of the network pharmacology data, providing critical insights into the mo-lecular interactions of CSPG bioactive compounds.
Declarations
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
