Modulation of the gut-liver axis by oxymatrine alleviates metabolic dysfunction-associated steatotic liver disease

Abstract

BACKGROUND

Oxymatrine (OMT) has the potential to regulate intestinal microbiota and hepatic metabolites.

AIM

To explore the underlying mechanisms by which OMT exerts its effects on metabolic dysfunction-associated steatotic liver disease (MASLD).

METHODS

An MASLD rat model induced by a high-fat, high-sucrose diet was treated with OMT. Assessments included serum/Liver biochemical parameters, histopathology (hematoxylin eosin and oil red O staining), intestinal permeability, 16S rRNA gut microbiota sequencing, and hepatic metabolomics. Correlation analysis linked changes in microbiota to metabolic shifts. To test the hypothesis that gut microbiota mediates the efficacy of OMT, we conducted fecal microbiota transplantation experiments.

RESULTS

OMT effectively alleviated MASLD in rats by improving serum lipid profiles (P < 0.05), reducing hepatic steatosis (P < 0.05) and inflammatory cytokine levels (P < 0.05), and enhancing intestinal barrier function. It substantially restored gut microbiota diversity, increasing beneficial genera such as Lactobacillus, modulating hepatic metabolites such as luteolin (P < 0.001), and lowering adrenic acid (P < 0.001), which are linked to lipid and inflammatory pathways. Correlation analysis indicated a strong association between changes in specific microbiota and metabolic improvement. Fecal microbiota transplantation experiments indicated that transferring OMT-modulated microbiota recapitulated the therapeutic effects in recipients, suggesting that the gut microbiota contributed substantially to the efficacy of OMT.

CONCLUSION

This study indicated that OMT alleviated MASLD in rats by regulating intestinal microbiota and hepatic metabolites, highlighting its promise as a therapeutic agent.

Key Words: Metabolic dysfunction-associated steatotic liver disease; Alcoholic liver disease; Metabolic dysfunction-associated steatohepatitis; Insulin resistance; Oxidative stress

Core Tip: Oxymatrine alleviates metabolic dysfunction-associated steatotic liver disease, primarily through modulation of the gut microbiome (increasing Lactobacillus, reducing Firmicutes/Bacteroidetes ratio) and the hepatic metabolome (increasing luteolin, decreasing adrenic acid). Fecal microbiota transplantation from oxymatrine donors recapitulated this protection, establishing a causal gut-liver axis for this natural alkaloid and offering a microbiota-guided therapeutic perspective.

INTRODUCTION

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a chronic liver condition that affects approximately 25% of the global population, the prevalence of which is increasing[1]. MASLD involves excess fat buildup in the liver without heavy alcohol consumption[2], progressing from simple steatosis to steatohepatitis, cirrhosis, and hepatocellular carcinoma[3]. MASLD is often associated with obesity, diabetes, and cardiovascular diseases[4], and poses a substantial health threat[5]. Currently, no approved pharmacotherapy exists, underscoring the urgent need for an effective treatment[6]. Recent advances have been made, such as the thyroid hormone receptor-beta agonist, resmetirom, which achieved success in phase III clinical trials for metabolic dysfunction-associated steatohepatitis (MASH)[7]. However, the identification of effective and safe therapeutic agents, particularly natural products, remains an active area of research.

The pathogenesis of MASLD is complex and has not yet been fully elucidated. It is widely explained by the “multiple-hit” theory[8], which involves insulin resistance, gut microbiota dysbiosis, increased intestinal permeability, and inflammation[9]. Of all the potential factors, the intestinal microbiota is believed to be crucial for MASLD progression. Altered gut microbiota can promote the translocation of bacterial products into the liver, triggering chronic hepatic inflammation through metabolites and pro-inflammatory pathways, which are key drivers of MASLD progression[10,11]. Clinical studies have noted the enrichment of Proteobacteria in patients with MASLD, whereas metabolites such as indole may counteract inflammation[12,13]. Transplantation of gut microbiota from high-fat diet (HFD)-fed mice promoted steatosis progression and perturbed insulin synthesis[14]. Targeting intestinal microbiota may therefore represent a promising approach for the treatment of MASLD.

Hepatic metabolic reprogramming is central to MASLD[15], driving lipid dysregulation and accumulation that causes cellular dysfunction and inflammation[16,17], which is key to the progression to MASH and even cirrhosis. MASLD is invariably accompanied by changes in liver metabolite levels. Oleuropein improves HFD-induced obesity in mice by modulating liver metabolites[18]. Hedan has been reported to ameliorate MASH by regulating hepatic metabolites[19]. Thus, modulating hepatic metabolites may offer a promising approach to treating MASLD.

Oxymatrine (OMT), derived from Sophora flavescens[20], alleviates MASLD by activating hepatic pathways such as Sirt1/AMPK[21]. MASLD may also be ameliorated by rebalancing the intestinal microbiota, as has been reported in experimental autoimmune encephalomyelitis[22], experimental colitis[23], white matter injury[24], and the modulation of hepatic metabolites[24]. Nevertheless, the potential role of the intestinal microbiota and hepatic metabolites in mediating OMT’s anti-MASLD effects has not been established.

This study investigated the therapeutic effects of OMT in rats with MASLD using 16S rRNA sequencing, metabolomics, and fecal microbiota transplantation (FMT) validation to provide a basis for its application.

MATERIALS AND METHODS

Ethical statement

All animal procedures were approved by the Animal Experimentation Ethics Committee of Zhejiang Eyong Pharmaceutical Research and Development Center (No. ZJEY-20230508-01). This study adhered to the guidelines for the Care and Use of Laboratory Animals.

Animal experiments

Male Sprague-Dawley rats (160-180 g) purchased from Shanghai SLAC Laboratory Animal Co., Ltd. [license SCXK(Hu)2022-0012]. Upon arrival, animals were maintained in a specific pathogen-free room (20-26 °C, 50%-60% humidity, 12-hour light/dark cycle).

Experiment 1: Following a 1-week acclimation, 57 rats were randomized to receive either a standard diet (control, n = 11) or a high-fat, high-sucrose diet (HFHSD) (n = 46). The HFHSD was prepared by supplementing standard rodent feed (74.25% of final weight) with lard (10%), sucrose (10%), egg yolk powder (5%), cholesterol (0.5%), and sodium taurocholate (0.25%)[25]. All dietary components were obtained from Beijing Boaigang Biological Technology Co., Ltd. Following an 8-week feeding period, rats were randomly chosen from each group (random number table method) and euthanized. Liver tissues were collected for histochemical analysis [hematoxylin eosin (HE) and oil red O (ORO) staining] to verify hepatic steatosis induction (Supplementary Figure 1)[25]. Rats were then randomized into three treatment groups (n = 15/group) for 27-week daily gavage: (1) Model (saline); (2) OMT (100 mg/kg/day; B21470, YuanYe)[25]; and (3) Metformin (MET) (200 mg/kg/day; B25331, YuanYe)[26]. Saline was administered to mice in the control group. Body weight was recorded weekly.

Experiment 2: Rats on HFHSD for 7-week were randomly assigned to three groups of six: (1) Antibiotics (Abx) + model; (2) Abx + FMT-sterile fecal filtrate (SFF); and (3) Abx + FMT. Following a protocol from Liu et al[27], animals received autoclaved water containing an antibiotic cocktail (1 g/L each of ampicillin, metronidazole, neomycin, and vancomycin) for 2-week to achieve gut microbiota depletion. To evaluate antibiotic clearance efficacy, fecal samples were collected from untreated rats, as well as from rats after antibiotic treatment but prior to FMT. DNA was extracted from these samples using centrifuge column-based fecal DNA extraction kit (CWY116S; CWBIO), and quantified via fluorescence to determine total bacterial DNA content[27]. Subsequently, the antibiotic water was replaced with autoclaved, deionized water. From week nine onward, rats were administered 1 mL daily by oral gavage according to group: (1) Abx + FMT-SFF (sterile fecal filtrate); (2) Abx + FMT (FMT preparation); and (3) Abx + model (sterile saline). This intervention was continued until week 27. Throughout the experimental period, all the rats were maintained on HFHSD. Both SFF and FMT samples were collected at the end of the final OMT administration in experiment 1, following established procedures[28-30]. Briefly, upon completion of the OMT regimen, fecal samples were collected on ice from rats housed in sterile cages without bedding to avoid urine contamination. All subsequent steps were conducted aseptically. Briefly, feces were homogenized in sterile saline (1:5, m/v) and centrifuged (2000 rpm, 5 minutes). The supernatant was filtered. FMT samples were prepared by mixing the filtrate with 10% glycerol followed by storage at -80 °C. SFF was generated by passing the supernatant through a 0.22 μm sterile filter.

Sample size and allocation

The initial group sizes were set to ensure the robustness of the primary physiological measurements. For subsequent endpoint analyses, biological samples were collected from these cohorts: Serum and liver tissues for core biochemistry and histopathology were analyzed from all surviving animals. From the control, model, and OMT groups, we randomly selected six rats each for 16S rRNA gene sequencing and untargeted metabolomics. For the independent FMT experiment (experiment 2), six rats per group were used based on the power considerations derived from the primary study. For more resource-intensive or confirmatory assays [e.g., western blotting and Evans Blue (EB) assay], three biological replicates per group were used, with samples randomly chosen, as the magnitude of change observed in pilot studies and similar experiments justified this sample size. The exact ‘n’ for each experiment is provided in the corresponding figure legend. All statistical tests were conducted based on the actual number of samples analyzed for each assay.

Biochemical analysis

After the final treatment, all rats were fasted for 12 hours and anesthetized using CO₂ inhalation. Blood was then drawn from the portal vein, and after centrifugation, serum levels of alanine aminotransferase (ALT), aspartate transaminase (AST), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) were quantified using a Hitachi 3110 automatic biochemical analyzer.

Enzyme-linked immunosorbent assay

The concentrations of tumor necrosis factor-α (TNF-α) (RX302058R), interleukin (IL)-1β (RX302869R), IL-6 (RX302856R) and IL-18 (RX302871R) in the serum of each rat were assessed using commercially-available enzyme-linked immunosorbent assay kits (Ruixin Biotech), following the instructions provided by the manufacturer.

Determination of hepatic TG and free fatty acid content

Upon euthanasia, rat livers were excised. A portion of the liver sample (0.1 g) was homogenized and ground thoroughly into a 10% homogenate. For hepatic TG (BC0625, Solarbio) and free fatty acid (FFA) (BC0595, Solarbio) quantification, the supernatant obtained by centrifugation of the homogenate was analyzed using the respective kits.

Histopathological examinations

Liver and ileal samples underwent fixation with paraformaldehyde, followed by alcohol dehydration, paraffin embedding, and sectioning into slices for subsequent routine HE staining. Histopathological alterations in each sample were examined under a light microscope. To assess the lipid accumulation in the liver, fresh liver samples were fixed and rapidly frozen in liquid nitrogen. Subsequently, the liver tissues were cut into 10-μm sections and stained using ORO. ImagePro Plus software was used to quantify the percentage of the ORO-positive staining area.

Determination of plasma endotoxin levels

Plasma endotoxin levels were measured using limulus amaebocyte lysate kits (Solarbio, T7573), using a dynamic turbidimetric method. To ensure sterility and avoid heat exposure, the plasma samples were collected and aliquoted under aseptic and heat-free conditions. Following euthanasia, venous blood was collected from the rats and platelet-rich plasma was obtained using centrifugation. Endotoxin contents in the plasma from each group were assessed according to the manufacturer’s protocol.

Evaluation of bacterial translocation

After euthanasia, the abdominal cavities of the rats were opened under aseptic conditions. Mesenteric lymph nodes (from the ileocecal junction to the mesenteric root), along with representative sections of liver and spleen, were excised and collected. Following a rinse with sterile saline, tissues were transferred to a sterile homogenizer. Sterile saline was then added for mechanical grinding. The obtained homogenates of these tissues were then inoculated onto blood plate mediums and incubated overnight. Following incubation, the bacterial content was calculated and displayed as log10 colony forming units per gram of tissue.

EB assay

Before euthanasia, rats were anesthetized with isoflurane and injected with 2% EB dye (4 mL/kg, G1910, Solarbio) via the tail vein. After allowing the dye to circulate for 2 hours, the rats were perfused with saline and euthanized using CO₂. The liver tissues of the rats were harvested, and 100 mg of each liver sample was accurately weighed and homogenized in formamide solution. Following centrifugation, EB concentration in the supernatant was quantified at 620 nm with a microplate reader, according to a standard curve. The following formula was used: EB in liver (ng/mg) = EB concentration (ng/mL) × volume of formamide (mL)/weight of the wet liver (mg).

Measurement of ileal permeability

The ileal permeability was assessed using Ussing chambers (Beijing Donggong Technology Co., Ltd., UC-2M). Briefly, after euthanasia, fresh segments of the ileum were excised and opened along the mesenteric edge. The ileal tissues were then washed with Krebs solution and mounted as intact sheets in dedicated Ussing chambers. Each mucosal and serosal chamber was filled with 5 mL of Kreb’s solution. The buffer was continuously oxygenated with a mixture of 5% CO₂ and 95% O₂ and maintained at 37 °C throughout the experiment. Once the preparations reached equilibrium for 20 minutes, and the baseline potential difference and resistance were established, 0.1 mg/mL horseradish peroxidase (HRP) was introduced into the mucosal chamber. To measure the permeability to HRP, 0.22 mL of sample was taken (and replenished with an equal volume of fresh medium) from the serosal chamber every 30 minutes for 180 minutes to measure the HRP content.

Western blotting

Following euthanasia, the ileal tissues of the rats were harvested, and proteins from the ileal tissues were isolated using a radioimmunoprecipitation assay buffer. The protein concentration in each sample was determined. Proteins were separated by SDS-PAGE and subsequently electroblotted onto polyvinylidene fluoride membranes. Membranes were blocked with non-fat milk and subsequently incubated at 4 °C with primary antibodies: Occludin (1:1000, DF7504, Affinity), ZO-1 (1:1000, AF5145, Affinity), and β-actin (1:10000, 81115-1-RR, Proteintech). Membranes were washed and then incubated for 2 hours with secondary antibodies (1:6000, 7074, CST). After subsequent rinsing, bands were visualized via enhanced chemiluminescence and analyzed with ImageJ.

The 16S rRNA analysis

Post-euthanasia, cecal contents were collected from six rats per group (control, model, OMT). PCR amplification of the 16S rRNA gene V3-V4 region from sample genomic DNA was performed for library construction. The library was sequenced using an Illumina Novaseq-PE250 platform. Using the DADA2 algorithm coupled with Vsearch for quality control, we clustered amplicon sequences into amplicon sequence variants/operational taxonomic units at 97% similarity to construct the feature abundance table.

Statistical analysis of abundance at the class level across the three groups was conducted using amplicon sequence variants/operational taxonomic units sequence abundance data. The microbial and evolution-based diversity in the cecal contents were evaluated using alpha diversity analysis (metrics including Chao1, observed species, and Faith’s phylogenetic diversity). Additionally, differences in microbial structure were estimated using beta diversity analysis according to weighted UniFrac such as principal coordinate analysis and nonmetric multidimensional scaling. To assess differences in bacterial communities, we utilized Bray-Curtis distances. Based on Euclidean distance and UPGMA clustering outcomes, a heatmap was plotted to show species composition at the genus level. Group-specific microbial taxa were identified by performing linear discriminant analysis effect size on the Huttenhower lab Galaxy server, using an LDA score cutoff of 2.0. Sequencing service was outsourced to Suzhou PANOMIX Biomedical Tech Co., Ltd.

Metabolome analyses

Liver tissues from six rats per group (control, model, OMT) were collected for metabolomics. After homogenization and centrifugation, the supernatant was concentrated, dried, redissolved in 50% acetonitrile (200 μL), filtered through a 0.22-μm membrane, and subjected to liquid chromatography-mass spectrometry detection. Metabolomic analysis was conducted by Suzhou PANOMIX Biomedical Tech Co., Ltd., using liquid chromatography-mass spectrometry technology. For analysis, 2 μL of supernatant was injected into a Vanquish UHPLC system coupled to an ESI-Orbitrap Exploris 120 mass spectrometer. Details of the liquid chromatography and mass spectrometry conditions are provided in Supplementary material.

The original data were transformed into mz XML format using MS Convert in ProteoWizard, followed by processing with R XCMS software for peak recognition, filtering, and alignment. Batch effects were eliminated by adjusting the data according to quality controls (QCs). Metabolites with relative standard deviations < 30% in the QCs were retained for further data processing. The tight clustering of QC samples in principal component analysis (PCA) (Supplementary Figure 2) demonstrated good instrument stability and reproducibility during the liquid chromatography-mass spectrometry runs. Following this, metabolite annotation was carried out by querying multiple databases, including Mass Bank, HMDB, Lipid Maps, Kyoto Encyclopedia of Genes and Genomes, mz Cloud, and a proprietary database from Suzhou PANOMIX Biomedical Tech Co., Ltd. Multivariate analysis included PCA and orthogonal partial least squares discriminant analysis; model quality was evaluated via permutation tests. Variable importance in the projection (VIP) scores from orthogonal partial least squares discriminant analysis identified important variables, and those with VIP > 1 underwent univariate t-testing (significance threshold: P < 0.05).

Statistical analysis

Values were mean ± SD. SPSS (version 20.0) was used for data analysis. Multiple groups were compared using one-way analysis of variance with Tukey’s test (for normally distributed data) or the Kruskal-Wallis H test followed by Dunnett’s T3 test (for non-normal distributions or heterogeneous variance); P < 0.05 defined significance. The Omic Studio Cloud platform was utilized to examine associations linking the intestinal microbiome to liver metabolic profiles. When conducting correlation analysis linking the gut microbiota to hepatic metabolic profiles using the platform, the option ‘Spearman’ was selected under ‘Correlation Calculation Method’, with all other parameters set to the platform's default settings. In animal experiments, researchers who collected samples and tested the indicators were blinded to the grouping situation.

RESULTS

OMT reduced serum and hepatic lipid contents in HFHSD-induced MASLD rats

A rat model of MASLD was established and investigated using OMT and MET interventions (Figure 1A). Model rats had higher body weights than control rats at multiple time points (P < 0.05; Figure 1B). OMT and MET treatments led to a slight but statistically insignificant weight reduction. The model rats had larger body sizes (Figure 1C) and livers with brighter colors and blunt edges, whereas the OMT and MET groups had smaller bodies and dark-red livers with sharp edges (Figure 1D). Serum contents of ALT, AST, TC, TG, and LDL-C were elevated and HDL-C content was reduced in the model group (Figure 1E, P < 0.01). Serum inflammatory factors (TNF-α, IL-1β, IL-6, IL-18) and hepatic lipid indicators (TG, FFA) were increased (P < 0.01; Figure 1F and G). However, after treatment with OMT or MET, most of these changes were effectively reversed (P < 0.01).

Figure 1 Oxymatrine improves liver function, lipid content, and inflammatory parameters in serum and liver in metabolic dysfunction-associated steatotic liver disease rats. A: Study schematic; B: Body weight trajectory post-oxymatrine, n = 6; C: Representative rat images (final day); D: Gross liver morphology; E: The contents of serum alanine aminotransferase, aspartate transaminase, total cholesterol, triglyceride, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol, n = 6; F: Levels of cytokines including tumor necrosis factor-α, interleukin (IL)-1β, IL-6, and IL-18, n = 6; G: Hepatic triglyceride and free fatty acid contents, n = 6. ^aP < 0.05, ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. ALT: Alanine aminotransferase; AST: Aspartate transaminase; FFA: Free fatty acid; HDL-C: High-density lipoprotein cholesterol; IL: Interleukin; LDL-C: Low-density lipoprotein cholesterol; MET: Metformin; OMT: Oxymatrine; TC: Total cholesterol; TG: Triglyceride; TNF-α: Tumor necrosis factor-α.

Reduced liver injury and steatosis after OMT treatment in MASLD rats

Control rats exhibited intact liver lobules on HE staining, whereas model rats displayed features of steatosis and inflammation. However, rats supplemented with OMT or MET showed reduced hepatic injury (Figure 2A). The percentage of ORO-positive area was significantly higher in model rats than in controls (P < 0.01). Both OMT and MET treatments effectively reduced hepatic lipid deposition, as demonstrated by a lower ORO-positive area (P < 0.01; Figure 2B).

Figure 2 Oxymatrine alleviates hepatic pathological damage and lipid accumulation in metabolic dysfunction-associated steatotic liver disease rats. A: Liver histology (hematoxylin eosin staining, representative images), 100 × (200 μm); 400 × (50 μm); B: Liver histology (oil red O staining, representative images). n = 6, 100 × (200 μm); 400 × (50 μm). ^aP < 0.05, ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. HE: Hematoxylin eosin; MET: Metformin; OMT: Oxymatrine; ORO: Oil red O.

OMT improved intestinal permeability in HFHSD-induced MASLD rats

Model rats had elevated plasma endotoxin and bacterial content in tissues compared to the controls (P < 0.01), which was alleviated by OMT or MET (P < 0.01; Figure 3A and B). HE staining indicated that the ileal tissues of rats in the model group were severely damaged, as evidenced by extensive inflammatory cell infiltration and missing and broken chorionic villi. In contrast, ileal tissue damage was ameliorated in rats treated with OMT or MET (Figure 3C). EB levels increased, whereas HRP, occludin, and ZO-1 levels decreased (P < 0.01). Treatment with either OMT or MET substantially reversed these changes (P < 0.05; Figure 3D-F).

Figure 3 Oxymatrine improves intestinal permeability in metabolic dysfunction-associated steatotic liver disease rats. A: Plasma endotoxin levels of the rats from the portal vein, n = 6; B: Bacterial contents in mesenteric lymph nodes, liver tissues, and spleen tissues, n = 6; C: Ileal histology (hematoxylin eosin staining, representative images), 200 × (100 μm); 400 × (50 μm); D: The concentrations of Evans Blue in liver samples, n = 3; E: Permeability of horseradish peroxidase in segments of ileum incubated in Ussing chambers, n = 6; F: Protein levels of occludin and ZO-1 in ileal tissues, n = 3. ^aP < 0.05, ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. MET: Metformin; OMT: Oxymatrine.

OMT reversed intestinal microbiota changes for HFHSD-induced MASLD rats

Alpha diversity analysis showed that the model group had lower Chao1, observed species, and Faith’s phylogenetic diversity indices than the control group, although the differences were not statistical. OMT substantially increased these indices (P < 0.05; Figure 4A). Principal coordinate analysis and nonmetric multidimensional scaling analyses indicated that the microbial community of model rats was markedly different from controls, whereas that of OMT-treated rats resembled normal rats (Figure 4B and C). Hierarchical clustering confirmed a shorter distance between the control and OMT groups than between the control and model groups (Figure 4D). At the phylum level, HFHSD feeding decreased Bacteroidetes and increased Proteobacteria. At the class level, it increased Clostridia and Erysipelotrichi but decreased Bacilli. OMT treatment reversed these changes (Figure 4E and F). OMT treatment reversed these changes. The HFHSD resulted in an increase in the Firmicutes/Bacteroidetes (F/B) ratio (P < 0.01), whereas OMT treatment decreased the F/B ratio (Figure 4G). There was a significant increase in the relative abundances of Clostridialesand and Streptococcus (phylum Firmicutes), and a concurrent decrease in S24-7 (phylum Bacteroidetes) in model rats vs controls (P < 0.05). However, OMT treatment reduced the relative abundance of Clostridiales and Streptococcus but enhanced the relative abundance of S24-7 in model rats (Figure 4H). The heatmap displays the top 50 genera (Figure 4I).

Figure 4 Oxymatrine enhances intestinal microbiota diversity in metabolic dysfunction-associated steatotic liver disease rats. A: Chao1, observed species, and Faith’s phylogenetic diversity indices of the alpha diversity analysis, n = 6; B: Principal coordinate analysis results of the beta diversity analysis; C: Nonmetric multidimensional scaling results of the beta diversity analysis; D: Hierarchical clustering tree of the rats; E: Composition of the top 20 intestinal microbiota (phylum level); F: Composition of the top 20 intestinal microbiota (class level); G: Firmicutes/Bacteroidetes ratio at the phylum level, n = 6; H: The relative abundance of Clostridiales, Streptococcus and S24-7 in each group, n = 6; I: A heatmap of the top 50 different bacterial genera among the three groups (control, model, oxymatrine). ^aP < 0.05, ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. Group A: Control; Group B: Model; Group C: Oxymatrine; OMT: Oxymatrine; PD: Phylogenetic diversity.

The linear discriminant analysis effect size analysis identified 27 substantially different taxa across the groups (Figure 5). From the phylum to the genus level, 27 taxa were identified across all groups. Enriched taxa differed between groups: Bacilli (class), Lactobacillaceae, Pseudomonadaceae (families), Lactobacillales (order), Lactobacillus, Pseudomonas, and Streptococcus (genera) were predominant in controls, while Lachnospiraceae (family) and Blautia (genus) were predominant in the model group.

Figure 5 Oxymatrine alters intestinal microbiota composition in metabolic dysfunction-associated steatotic liver disease rats. A: Linear discriminant analysis effect size analysis results; B: Cladogram of linear discriminant analysis effect size analysis; C: The relative abundance of Lactobacillus, Bacilli, Lactobacillales, Lactobacillaceae, Streptococcus, Lachnospiraceae, Blautia, Pseudomonadaceae, and Pseudomonas among groups (control, model, oxymatrine). Class A: Control; Class B: Model; Class C: Oxymatrine; LDA: Linear discriminant analysis.

OMT reversed hepatic metabolite changes for HFHSD-induced MASLD rats

Both PCA and OPLS-DA indicated clear group separations (Figure 6A), with the latter specifically highlighting significant metabolomic differences between control vs. model and model vs OMT groups (Figure 6B). Furthermore, permutation tests were applied, with R2X = 0.346 cum, R2Y = 0.974 cum, and Q2 = 0.846 cum for the positive ion mode and R2X = 0.370 cum, R2Y = 0.968 cum, and Q2 = 0.869 cum for the negative ion mode, suggesting that the OPLS-DA model used in this study was stable and reliable (Figure 6C). Figure 6D shows the detailed trends in the top 150 differential metabolites (based on the VIP values) among the three groups. Relative to the model group, luteolin, oleoylethanolamide, and pomiferin levels higher, whereas adrenic acid was lower in the OMT group (Figure 6E). Enrichment analysis using the Kyoto Encyclopedia of Genes and Genomes database showed primary involvement of these metabolites in the PPAR signaling pathway, along with glutathione and arachidonic acid metabolism (Figure 6F).

Figure 6 Oxymatrine reverses hepatic metabolite changes in metabolic dysfunction-associated steatotic liver disease rats. A: Principal component analysis scores (positive/negative modes); B: Orthogonal partial least squares discriminant analysis score plots (positive and negative ion modes); C: Orthogonal partial least squares discriminant analysis permutation tests: Both ion modes; D: Top 150 hepatic metabolites: Heatmap visualization; E: The levels of adrenic acid, luteolin, oleoylethanolamide and pomiferin in the hepatic metabolites among the three groups, n = 6; F: Enrichment analysis of differential hepatic metabolites in Kyoto Encyclopedia of Genes and Genomes pathways (control, model, oxymatrine). ^aP < 0.05, ^bP < 0.01 and ^cP < 0.001. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. Group A: Control; Group B: Model; Group C: Oxymatrine; NS: Not significant.

Correlation of intestinal microbiota and hepatic metabolites

The OmicStudio Cloud platform was used to investigate the association between the intestinal microbiota and hepatic metabolites (focusing on the 30 metabolites based on VIP values). Verrucomicrobia, Verrucomicrobiales, Akkermansia, Verrucomicrobiae, Verrucomicrobiaceae, Actinobacteria, Bifidobacteriales, Bifidobacteriaceae, and Bifidobacterium were strongly associated with most differential hepatic metabolites (Figure 7). In addition, Blautia, Streptococcaceae, and Streptococcus were positively correlated with the hepatic metabolites alpha-tocopherol, prostaglandin F3a, 3-dehydro-2-deoxyecdysone, and corticosterone.

Figure 7 Association between different intestinal microbiota and different liver metabolites. ^aP < 0.05, ^bP < 0.01 (color-coded: Red = positive, blue = negative). Data = mean ± SD. Groups were compared by one-way analysis of variance (Tukey’s test).

FMT from OMT-treated donors partially recapitulated the therapeutic phenotypes in recipient MASLD rats

To validate the causal role of gut microbiota, FMT experiments were conducted in antibiotic-pretreated rats (Figure 8). To verify the efficacy of the antibiotic treatment in depleting the host microbiota, we quantified the total bacterial DNA load in fecal samples collected immediately prior to FMT. As shown in Supplementary Figure 3, the antibiotic regimen led to a profound reduction (> 99%) in fecal DNA content compared to untreated controls, confirming successful microbiota ablation. The Abx + FMT group received microbiota from OMT-treated donors. As shown in Figure 8B, this group demonstrated markedly lower levels of serum ALT, AST, TG, TNF-α, IL-1β, IL-6, and IL-18 relative to the Abx + FMT-SFF group (P < 0.05).

Figure 8 Improvement of serum parameters and cytokines by fecal microbiota transplantation in metabolic dysfunction-associated steatotic liver disease rats. A: Study schematic; B: The contents of serum alanine aminotransferase, aspartate transaminase, total cholesterol, triglyceride, n = 6; C: The concentrations of tumor necrosis factor-α, interleukin (IL)-1β, IL-6, and IL-18, n = 6. ^aP < 0.05, ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. ALT: Alanine aminotransferase; AST: Aspartate transaminase; FMT: Fecal microbiota transplantation; IL: Interleukin; MET: Metformin; SFF: Sterile fecal filtrate; TC: Total cholesterol; TG: Triglyceride; TNF-α: Tumor necrosis factor-α.

Histopathological assessment demonstrated that hepatic tissue injury was markedly alleviated in the Abx + FMT group. This was evidenced by improved hepatic cord architecture, reduced steatosis, diminished patch necrosis, reduced inflammatory cell infiltration, and decreased lipid droplet accumulation (as visualized by ORO and HE staining; Figure 9A and B). Quantification showed significantly lower lipid deposition in the Abx + FMT vs the Abx + FMT-SFF group (P < 0.01; Figure 9C).

Figure 9 Fecal microbiota transplantation improved histopathological liver damage in metabolic dysfunction-associated steatotic liver disease rats. A: Liver histology (hematoxylin eosin staining, representative images), 100 × (200 μm); 400 × (50 μm); B: Liver histology (oil red O staining, representative images). n = 6, 100 × (200 μm); 400 × (50 μm); C: Quantitative oil red O staining of rat liver tissue, n = 3. ^bP < 0.01. Statistical analysis was conducted using one-way analysis of variance with Tukey’s post hoc test, and results are presented as mean ± SD. Abx: Antibiotics; FMT: Fecal microbiota transplantation; HE: Hematoxylin eosin; ORO: Oil red O; SFF: Sterile fecal filtrate.

DISCUSSION

MASLD is a major global public health concern. Lifestyle changes aimed at weight loss are the only effective treatment, but are challenging to sustain, underscoring the urgent need for new therapies[31,32]. Natural medicines such as OMT, a quinazoline alkaloid derived from Sophora flavescens, offer advantages such as rich resources and minimal side effects[33,34]. In the current study, rats with HFHSD-induced MASLD were used to replicate human MASLD. MASLD rats exhibited characteristic increases in serum (ALT, AST, TC, TG, LDL-C, TNF-α, IL-1β, IL-6, IL-18) and hepatic (TG, FFA) lipid and inflammatory markers, coupled with a decrease in HDL-C. Consistent with established MASLD pathology[9,35]. OMT treatment effectively reversed these alterations, ameliorating both pathological injury and hepatic steatosis. OMT could therefore serve as an efficient treatment option for MASLD.

The intestinal microbiota is strongly linked to MASLD development. Targeting the intestinal microbiota is a promising therapeutic approach for MASLD[36,37]. The liver is more susceptible to intestinal microbiota imbalance because of its connection to the gut via the portal vein[38]. Under normal physiological conditions, the intestinal mucosa serves as a defensive barrier, shielding the body from exogenous hazardous substances, and efficiently preventing enteric pathogens from entering the liver[9,39]. Liver inflammation and damage can be caused or exacerbated by gut-derived toxins that, following intestinal injury and microbiota dysbiosis, enter the liver via the portal vein[32]. Liver-derived inflammation can further reduce intestinal mucosal barrier function by disrupting tight junctions[40]. he interactions between intestinal microbiota imbalance and liver inflammation are bidirectional and contribute to the development of MASLD. OMT alleviated intestinal barrier deterioration and systemic inflammation in mice with intracerebral hemorrhage[24]. OMT has been reported to improve intestinal barrier function in rats with cirrhosis[41]. Consistently, in the current study, exposure to a HFHSD led to a deterioration in intestinal barrier function, as evidenced by elevated plasma endotoxin levels, accelerated bacterial translocation, damaged ileal tissues, increased ileal permeability, and reduced Occludin and ZO-1 protein expression in ileal tissues. However, after OMT treatment, these conditions were reversed.

Due to the effects of OMT on intestinal microbiota rebalance, the effects of OMT on the intestinal microbiota were investigated in MASLD rats[22]. Alpha diversity analysis indicated that after treatment with OMT, Chao1, observed species, and Faith’s phylogenetic diversity indices of the model rats were elevated, suggesting that OMT enhanced microbial and evolution-based diversity in MASLD rats. Beta diversity analysis indicated that OMT effectively restored the microbial composition in MASLD rats. OMT administration restored the gut microbiota of MASLD rats toward normality, evidenced by the recovery of the F/B ratio and adjusted abundance of specific taxa, namely Clostridiales, Streptococcus, S24-7, Lactobacillaceae, Lachnospiraceae, Lactobacillus, and Blautia. An elevated F/B ratio leads to reduced short-chain fatty acid generation and increased energy obtained from the diet, which can promote the onset and progression of MASLD[42]. Scientists have indicated that Clostridiales is a biomarker for MASLD, and compared to healthy subjects, the proportion of Streptococcus is enhanced in patients with MASLD[43,44]. Moreover, S24-7 exhibits substantial negative correlations with TG, CHOL, LDL-C, IL-6, IL-1β, TNF-α, and body weight and positive correlations with HDL-C[45]. Shen et al[46] found that patients with obvious liver fibrosis or MASH exhibited higher abundance of Blautia and Lachnospiraceae, which are strongly linked to MASLD development. Lactobacillus plantarum ZJUIDS14 alleviates liver steatosis and HFD-induced injury caused by HFD[47]. Thus, the mechanism by which OMT improves MASLD may be associated with the restoration of the intestinal microbiota.

MASLD also results in changes in the liver metabolite profiles; ameliorating liver metabolic abnormalities can serve as a promising strategy for the treatment of MASLD[48]. By using non-targeted metabolomics, a few potential liver markers were identified from the differentially expressed metabolites. Adrenic acid was decreased in the OMT group. Adrenic acid exacerbates inflammation in MASLD, and its accumulation is linked to the progression of MASLD[49]. Luteolin was increased in the OMT group. Luteolin is a promising hepatoprotective agent that can improve liver lesions through multiple mechanisms, such as inhibition of inflammation, reduction of oxidative stress, and regulation of lipid homeostasis[50]. An elevation in oleoylethanolamide was observed following OMT treatment. Evidence indicates that oleoylethanolamide confers benefits in MASLD by regulating the expression of genes involved in lipid metabolism and improving serum NRG4 levels[51]. Pomiferin levels increased after OMT treatment; pomiferin has been reported to mitigate the harmful influence of NiCl2 on liver tissues via various cell and signaling mechanisms[52].

Correlation analysis indicated a strong relationship between intestinal microbiota and hepatic metabolites. The hepatoprotection conferred by OMT likely involves modulation of the gut-liver axis (intestinal microbiota and hepatic metabolites). As this study aimed to elucidate OMT’s specific mechanisms rather than compare it with MET, 16S rRNA sequencing and metabolomics were performed exclusively in the control, model, and OMT groups.

That OMT reshapes the gut microbiota and hepatic metabolome adds a novel dimension to previously-reported mechanisms in MASLD. OMT exerts hepatoprotective effects, partly by activating the hepatic Sirt1/AMPK signaling pathway, which promotes fatty acid oxidation and inhibits lipogenesis[21]. The present study suggests that this “microbiota-metabolite” axis may not operate in isolation but could interact with these intracellular pathways. OMT treatment increased the abundance of Lactobacillus, a genus known for producing short-chain fatty acids (SCFAs) such as butyrate[53]. These SCFAs, notably, activate AMPK in various tissues, including the liver[54]. Thus, it is plausible that OMT-induced enrichment of Lactobacillus contributes to AMPK activation via SCFAs, thereby reinforcing the direct pharmacological action of OMT on the Sirt1/AMPK axis. Furthermore, the hepatic metabolite, luteolin, which was elevated upon OMT treatment, is not merely associated with the PPAR signaling pathway in a broad sense but has been identified as a modulator of PPARγ activity, potentially acting as an agonist or expression regulator[55]. PPARγ activation can improve insulin sensitivity and regulate lipid metabolism, whereas its cross-talk with the AMPK pathway is also documented[56]. Therefore, OMT may orchestrate a multi-target therapeutic network: It directly activates Sirt1/AMPK, concurrently fosters a Lactobacillus-enriched gut environment that may produce AMPK-activating SCFAs, and upregulates hepatoprotective metabolites such as luteolin that fine-tune complementary pathways such as PPARγ. This integrative perspective bridges the gut-liver axis with fundamental metabolic regulators, providing a more holistic understanding of the anti-MASLD action of OMT.

Subsequent FMT experiments yielded compelling functional evidence that solidified the causal role of gut microbiota in mediating the therapeutic effects of OMT. The transfer of OMT-modulated microbiota alone was sufficient to recapitulate key therapeutic outcomes in recipient rats, including the amelioration of hepatic steatosis, tissue injury, and systemic inflammation. This result strongly indicates that the restructured microbial community acts as a functional effector responsible for the hepatoprotective phenotype. These findings not only reinforce the causal linkage within the gut-liver axis but also highlight that the therapeutic efficacy of OMT is intrinsically tied to its capacity to induce a host-protective gut microbiome.

Limitations

The evaluation of OMT at a fixed dose of 100 mg/kg hindered assessment of its dose-response relationship. It remains unclear whether higher doses would enhance efficacy or induce toxicity or if the observed microbial and metabolic alterations represent the optimal effect. Our study design focused on the control, model, and OMT groups for 16S rRNA and metabolomic analyses to directly investigate OMT’s mechanism of action. Including a positive control (MET) group in future -omics analyses would improve the specificity of the effects of OMT. In the FMT experiment, although a therapeutic phenotype was successfully transferred, the gut microbiota in the recipient rats was not directly profiled post-transplantation. The efficacy of antibiotic pretreatment for microbiota depletion, although based on an established protocol, was not independently quantified. Future studies should include longitudinal tracking of the recipient microbiota and verification of depletion efficiency (e.g., via fecal bacterial DNA load) to directly confirm microbial engraftment and strengthen the model.

CONCLUSION

OMT administration (100 mg/kg) attenuated MASLD, primarily through modulating gut microbiota balance and hepatic metabolite profiles. Through the FMT experiment, we demonstrated that the gut microbiota underlies the mechanism of action for OMT’s beneficial effects. By demonstrating that OMT acts via the gut-liver axis, our work offers key mechanistic insights supporting its promise as a MASLD treatment, an essential step prior to future translational dose-response studies.