Mechanisms of Fuling Zexie decoction against metabolic dysfunction-associated steatotic liver disease: Integrative metabolomics, gut microbiota and serum pharmacochemistry

Abstract

BACKGROUND

Metabolic dysfunction-associated steatotic liver disease (MASLD) currently lacks specific therapeutic options. Fuling Zexie decoction (FZD) is a Chinese herbal formula that shows potential for treating MASLD, but its mechanism and active ingredients remain unclear.

AIM

To determine the mechanism and active ingredients of FZD against MASLD.

METHODS

Free fatty acid-stimulated steatotic HepG2 cells and high-fat diet-fed MASLD rats were used to assess the functionality of FZD. To investigate the mechanisms by which FZD affects endogenous metabolites and intestinal microbiota in rats, a combination of metabolomics and 16S rRNA sequencing was utilized. Serum pharmacochemistry-metabolomics integration identified FZD’s active ingredients and core targets, while molecular docking and cellular experimentation were utilized to validate the association of these ingredients and targets.

RESULTS

FZD ameliorated hepatic steatosis and hepatocellular injury in rats with MASLD, ameliorated serum and hepatic lipid levels, augmented antioxidant capacity, and mitigated inflammatory responses. Mechanistically, FZD restored 226 endogenous metabolites and regulated 17 metabolic pathways. This included reducing the Firmicutes/Bacteroidota ratio and diminishing the levels of potentially harmful microbes (Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter). A total of 39 FZD-originated constituents were detected in the serum. Among these, 24 were recognized as key pharmacodynamic constituents that modulated 17 targets, 10 metabolic pathways, and 18 endogenous metabolites. The key pharmacodynamic compound group downregulated the expression of NOS2, MAOA, and MIF, thereby alleviating hepatic steatosis.

CONCLUSION

These results demonstrate that FZD alleviates MASLD by regulating 10 metabolic pathways, decreasing the abundance of Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter, and normalizing 18 endogenous metabolite levels, which may be primarily attributed to the 24 FZD-derived active ingredients.

Key Words: Active components; Fuling Zexie decoction; Intestinal microbiota; Metabolic dysfunction-associated steatotic liver disease; Metabolites

Core Tip: This study employed an integrative strategy combining metabolomics, gut microbiome profiling, and serum pharmacochemical analysis to elucidate the therapeutic effects of Fuling Zexie decoction on metabolic dysfunction associated with steatotic liver disease. Fuling Zexie decoction normalized 226 metabolites and 17 pathways in rats, while also reducing the Firmicutes/Bacteroidota ratio and the abundance of three pathogenic bacteria. Among 39 serum-derived components, 24 key components regulated 17 targets, 10 pathways, and 18 metabolites, and downregulated nitric oxide synthase 2, monoamine oxidase A, and macrophage migration inhibitory factor to alleviate hepatic steatosis.

INTRODUCTION

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a clinicopathological syndrome characterized by multisystem metabolic disorders and hepatic steatosis. Currently, its worldwide prevalence is estimated at roughly 30%, making it the second leading liver disease globally, after viral hepatitis[1]. The progression of MASLD can lead to serious conditions such as hepatitis, liver fibrosis, cirrhosis, and even hepatocellular carcinoma. Moreover, MASLD often coexists with obesity and other metabolic disorders[2]. Despite the significant clinical and public health challenges posed by MASLD, there is no clear therapeutic strategy. Currently, management primarily focuses on dietary changes, physical activity, and lipid-lowering medications[3]. Consequently, there is a pressing need to develop novel therapeutic agents to enable effective management of MASLD.

Traditional Chinese medicine (TCM) exerts pleiotropic effects through its polypharmacological nature, acting on multiple signaling cascades and molecular targets, and has shown substantial therapeutic efficacy in MASLD[4]. From a TCM perspective, the core pathogenesis of MASLD lies in phlegm-dampness obstruction and stagnation of qi and blood[5]. Originating from the Synopsis of Prescriptions of the Golden Chamber, Fuling Zexie decoction (FZD) is formulated with Poria (PR), Alismatis Rhizoma (AR), Atractylodis Macrocephalae Rhizoma (AMR), Cinnamomi Ramulus (CR), Glycyrrhizae Radix et Rhizoma (GRR), and Zingiberis Rhizoma Recens (ZRR) in the proportion of 8:4:3:2:2:3. FZD exerts the effects of dissipating phlegm, eliminating blood stasis, and warming yang to resolve fluid retention, thereby enabling it to exert therapeutic effects on the MASLD symptoms[6]. Modern pharmacological studies have demonstrated that FZD inhibits the Janus kinase-signal transducer and activator of transcription 3 (STAT3) cascade, along with activation of the NOD-like receptor thermal protein domain associated protein 3 (NLRP3) inflammasome, thereby reducing renal inflammation and fibrosis in hyperuricemic mice[7]. In hyperlipidemic rats, FZD regulates myeloperoxidase activity and increases serum nitric oxide levels, contributing to blood lipid control[8]. Furthermore, clinical studies have revealed that FZD effectively reverses glucose and lipid metabolism disorders in type 2 diabetes mellitus patients[9]. Additionally, our previous research confirmed that PR (the principal herb of FZD) activates farnesoid X receptor (FXR) and peroxisome proliferator-activated receptor alpha (PPARα), downregulates CYP7A1 and SREBP-1c expression, improves mitochondrial function, and thus exerts a protective effect against MASLD[10,11]. Moreover, other component herbs of FZD, including AR, AMR, GRR, and ZRR, have been reported to alleviate symptoms associated with MASLD[12-15]. However, the mechanism and active ingredients of FZD in alleviating MASLD remain unclear.

Thus, this work explored the beneficial action of FZD in MASLD based on a cellular model of free fatty acid (FFA)-triggered lipid deposition in HepG2 cells and a rodent model of diet-induced hepatic steatosis. Additionally, a comprehensive analysis approach integrating metabolomics, gut microbiota, and serum pharmacochemistry was employed to explore the anti-MASLD mechanism of action and active ingredients of FZD. Furthermore, the anti-MASLD efficacy of the active principles was evaluated via molecular docking and in vitro assays. The present results are poised to serve as new evidence and offer novel perspectives on FZD in the management of MASLD.

MATERIALS AND METHODS

Extraction of FZD

The FZD extract was prepared by accurately weighing PR, AR, AMR, CR, GRR, and ZRR at a ratio of 8:4:3:2:2:3. The mixed herbs were decocted twice with 10-fold water (1 hour per decoction). The filtrates from both cycles were pooled, concentrated under reduced pressure at 65 °C using a rotary evaporator, and then freeze-dried, yielding 15% of the FZD aqueous extract as a lyophilized powder. The quality of the FZD extract was evaluated by measuring the contents of atractylenolide II (AT-II) and cinnamaldehyde, which were 8.55 mg/g and 0.37 mg/g, respectively (details are provided in the quality control of FZD extract section in the Supplementary material and Supplementary Figure 1).

Evaluation of the anti-MASLD ability of FZD in vitro

The HepG2 cells were cultivated in Dulbecco’s modified Eagle medium (100 mL/L fetal bovine serum, 10 mL/L antibiotics) at 37 °C with 50 mL/L carbon dioxide (CO₂). To determine the appropriate concentration, cells were plated into 96-well plates at a density of 3 × 10⁵ cells per well, allowed to adhere for 24 hours, and subsequently exposed to FZD at concentrations ranging from 50 μg/mL to 350 μg/mL for an additional 24 hours. Viability was evaluated by the cell counting kit 8 (CCK-8) to establish the optimal concentration. For the lipogenic model, cells in the exponential phase were plated into 6-well plates at the same density, cultured for 48 hours, and subsequently exposed to 1 mmol/L FFA (oleic acid: Palmitic acid, 2:1) over a 24-hour period. Subsequently, the cells were treated with FZD extract (50 μg/mL and 200 μg/mL) or fenofibrate (150 μM) for 24 hours, after which they were harvested for further analysis. Following Oil red O staining, cellular lipid deposition was examined under a light microscope. Commercial kits were used to evaluate biochemical parameters. HepG2, a human hepatocellular carcinoma cell line, was sourced from the Cell Resource Center of the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences.

Evaluation of the anti-MASLD ability of FZD in vivo

Animals and experimental design: Male specific pathogen-free Sprague-Dawley rats (8 weeks of age) were housed in the Laboratory Animal Center of Yunnan University of Chinese Medicine under a 12-hour light/dark cycle at 22 ± 2 °C and 45% ± 10% humidity. The animals were sourced from Beijing Speifu Biotechnology Co., Ltd. (Beijing, China; Certificate). They were randomly divided into five groups after one week of adaptive feeding (n = 6 per group): A control group (CON), a model group (MOD), a fenofibrate capsule group (FC) (21 mg/kg), a low-dose FZD group (LFZD) (7.25 g/kg), and a high-dose FZD group (HFZD) (29 g/kg). Of note, the LFZD dose for rats was derived from the human-equivalent dose of 69 g/person/day, as documented in the Synopsis of Prescriptions of the Golden Chamber. Rats in the CON and MOD groups received normal saline intragastrically, whereas the remaining three groups received their respective drug suspensions in an equal volume of saline. To induce MASLD, all groups except the CON group were fed a high-fat diet (HFD) for 12 weeks, during which all treatments were administered intragastrically once daily. Daily feed intake and weekly body weight were recorded throughout the study. All animal experiments were conducted in accordance with the ethical standards set by the Animal Ethics Committee of Yunnan University of Chinese Medicine (approval No. R-062022135).

Sample collection: At the 12-week stage of the study, rats were subjected to a 24-hour urine collection, which was then stored at a temperature of -80 °C. At the end of the experiment, the animals received their respective test samples, and 1 hour later, they were anesthetized with pentobarbital sodium (1.5 mL/kg). After blood withdrawal from the abdominal aorta and clotting at room temperature for 4 hours, the sample was centrifuged (1000 × g, 15 minutes, 4 °C) to isolate serum. The resulting serum was collected and frozen at -80 °C. In addition, liver tissues and colonic contents were harvested for subsequent analyses.

Oil Red O and hematoxylin-eosin histochemical staining: Liver tissues were processed for two histochemical assessments. For Oil Red O staining, samples were embedded in optimal cutting temperature compound, frozen at -20 °C for 30 minutes, and cut into 8-μm sections, which were then rinsed with 600 mL/L isopropanol and stained with Oil Red O. In parallel, other liver specimens were fixed in 40 g/L paraformaldehyde, after paraffin embedding, and 5-μm sections were obtained and stained with hematoxylin-eosin. Both types of stained specimens were visualized under a light microscope to evaluate lipid accumulation and cellular morphological changes, respectively.

Assessment of biochemical indices: The biochemical parameters were measured using commercial kits, and absorbance was recorded on a SpectraMax Plus 384 microplate reader.

Metabolomic analysis

Sample processing: Serum (200 μL) from each experimental group (CON, MOD, FC, HFZD) was deproteinized by adding 600 μL acetonitrile, held at 4 °C for 20 minutes, and centrifuged (12000 × g, 15 minutes, 4 °C). The clear supernatants were evaporated under nitrogen and then re-dissolved in 100 μL of acetonitrile, and centrifuged again (12000 × g, 10 minutes, 4 °C). Urine samples were treated in the same way.

Liver samples (0.2 g) from each experimental group (CON, MOD, FC, HFZD) were homogenized in 1 mL saline (6 minutes, 50 Hz, -10 °C), mixed with 3 mL acetonitrile for deproteinization, kept at 4 °C for 20 minutes, and centrifuged (12000 × g, 15 minutes, 4 °C). After concentration under nitrogen, the residues were reconstituted in 100 μL acetonitrile, and centrifuged again (12000 × g, 10 minutes, 4 °C) for the final analysis.

Quality control samples, prepared from equal-volume mixtures of all samples, were interspersed at a frequency of one per five test samples during analysis to verify the repeatability of the assay. The liquid chromatography-mass spectrometry (LC-MS) analysis conditions are described in the Supplementary material.

Data processing: Raw LC-MS data were preprocessed using Progenesis QI software. Metabolite annotation was conducted by searching the HMDB, Metlin, and Majorbio databases. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were then applied to visualize variations in metabolic profiles. Differential metabolites were identified using a threshold of variable importance in projection (VIP) > 1 and a P value < 0.05. Finally, pathway enrichment analysis of the identified differential metabolites was performed using MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.kegg.jp/kegg/pathway.html).

16S rRNA gene sequencing analysis

DNA was extracted from the colonic contents of rats in the CON, MOD, and HFZD groups using the PF Mag-Bind Stool DNA Kit, strictly following the provided protocol. A 2 μL sample of each DNA extract was electrophoresed on 1% agarose gels (120 V, 20 minutes) to verify its quality. The V3-V4 region of the 16S rRNA gene was amplified using primers 338F/806R, and the resulting polymerase chain reaction products were then used for library construction with the NEXTFLEX Rapid DNA-sequencing Kit, followed by pairedend sequencing on an Illumina PE300/PE250 system. Quality control of raw sequences was conducted with fastp v0.19.6, and operational taxonomic units were defined at 97% sequence identity using UPARSE v7.1. Taxonomic classification was performed with the RDP classifier (v2.11) at a 70% confidence cutoff. All subsequent bioinformatic analyses α-diversity (Mothur), inter-group difference testing (Wilcoxon rank-sum test), and β-diversity [PCA and principal co-ordinates analysis (PCoA)] were executed on the Majorbio Cloud Platform.

Association analysis of gut flora with pharmacodynamic markers and differential metabolites

Correlation analysis between the gut microbial community, disease markers, and differentially abundant metabolites was performed using Spearman’s rank correlation, with a threshold of |r| > 0.6 and P < 0.05.

Identification of FZD-derived ingredients in rat serum by ultra performance liquid chromatography-tandem mass spectrometry

Sample preparation: For each group (MOD and HFZD), A 1.2 mL aliquot of serum was combined with acetonitrile at a ratio of 1:3 (v/v), vortexed for 30 seconds, and maintained at 4 °C for 20 minutes. The mixture was then centrifuged (12000 × g, 15 minutes, 4 °C), and the supernatant was dried under a stream of nitrogen. The resultant extract was reconstituted in 200 μL of acetonitrile-water (1:1), filtered (0.22 μm), and further analyzed by ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS).

UPLC-MS analysis: An Agilent 1290 Infinity II system coupled with a Waters Acquity HSS T3 column (30 °C, 2.1 mm × 100 mm, 1.8 μm) was used for UPLC using 0.1% formic acid in water (A) and acetonitrile (B) at 0.2 mL/minute, 5 μL injection. The gradient used was as follows: 5% B at 0-5 minutes, 20% at 5-10 minutes, 40% at 10-15 minutes, 50% at 15-20 minutes, 60% at 20-25 minutes, 70% at 25-30 minutes, 80% at 30-35 minutes, 90% at 35-40 minutes, 100% at 40-41 minutes, re-equilibrate to 5% at 41 minutes. MS detection used an Agilent G6545 quadrupole time-of-flight with electrospray ionization, scanning 50-1700 m/z in positive/negative modes (spray: ± 3.5/2.8 kV). Sheath gas was 350 °C, 11 L/minute, drying gas was 320 °C, 8 L/minute and collision energy was 35, 40 eV.

Data analysis: Qualitative analysis software (version B.06.00) was applied for processing the UPLC/MS spectral data. Initially, the chromatograms of the FZD and MOD groups were superimposed and compared to eliminate interferences originating from endogenous serum components. The FZD-derived compounds in rat serum samples were subsequently identified based on their molecular weights, MS/MS data, and relative retention times in PR, AR, AMR, CR, GRR, and ZRR as reported in the literature.

Comprehensive analysis of drug-derived compounds and metabolomics

Potential MASLD targets were retrieved from the OMIM Gene Map and GeneCards databases, whereas targets of FZD-derived compounds in serum were predicted using SwissTargetPrediction and PubChem. Additionally, the related targets in the 17 metabolic pathways regulated by FZD were obtained from KEGG. Overlapping targets among MASLD, FZD-derived compounds, and metabolic pathways were determined using Venny 2.1.0 software. Additionally, the “active compound-target-metabolic pathway-metabolite” network was visualized in Cytoscape to identify the key active ingredients, targets, pathways, and metabolites involved in FZD-mediated amelioration of MASLD.

Molecular docking analysis of key active compounds derived from FZD with their core targets

Information on core targets was obtained from the Uniprot database (https://www.uniprot.org/). Crystal structures of the target proteins were retrieved from the Protein Data Bank (http://www.pdb.org/), stored in PDB format, and subjected to preprocessing with PyMOL to remove water molecules and residual ligands. Subsequently, hydrogenation was performed using AutoDock Tools, and the target proteins were saved in pdbqt format. The structures of the key drug derived compounds were acquired from the PubChem database (http://pubchem.ncbi.nlm.nih.gov/). After structure optimization using Chem 3D software, they were saved in mol2 format, converted to PDB format via PyMOL, and their node information was calculated using AutoDock Tools (version 1.5.7) before being saved in pdbqt format. Finally, target proteins and compounds were imported into AutoDock Vina for docking, with binding energies calculated. Binding energies < -5.0 kcal/mol indicated stable binding between drug-derived compounds and targets.

Evaluation of FZD-derived key compounds against MASLD in vitro

The steatosis model of HepG2 cells was established as described in the section concerning the in vitro anti-MASLD effect of FZD. The HepG2 cells were then incubated with 5 μg/mL and 20 μg/mL of the FZD-derived key compound group (FCG) [consisting of 16-hemagglutinin (HA), 11-deoxyalisol A (11-DA), betaine (Bet), alisol F-24-acetate (AF-24), 1-dehydro-[10]-gingerdione (D10G), trans-cinnamaldehyde (t-CA), liquiritigenin (Liq), 7-hydroxycoumarin (7-HC), 16,23-oxidoalisol B (16,23-OB), cinnamic acid (CA), dehydrotrametenolic acid (DLA), poricoic acid A (PA), AT-II, 8beta-methoxyatractylenolide I (8-MA), eburcoic acid (EA), glycyrrhetinic acid (GA), alisol F (AF), 6-shogaol (6-SG), and 4,4’-dihydroxy-2-methoxychalcone (D-2-MC) in the ratio 26.85:18.24:15.91:8.03:5.36:4.64:3.79:3.61:2.14:2.07:2.07:1.81:1.26: 1.23:0.98:0.95:0.43:0.39:0.22, which was derived from the serum chromatographic peak area ratios of the corresponding FZD constituents] and FC (150 μM) for 24 hours. After treatment, lipid deposition was determined by means of Oil Red O staining., and the biochemical parameters were measured with commercial kits following the provided protocols.

Determination of NOS2, MAOA, and MIF protein levels via western blotting

Radio immunoprecipitation assay lysis buffer containing protease inhibitors was used to lyse HepG2 cells. After centrifugation (12000 × g, 30 minutes, 4 °C), the supernatants underwent bicinchoninic acid quantification and were normalized to the lowestconcentration sample using lysis buffer. The normalization process was then followed by a combination of the proteins with the loading buffer, denatured (100 °C, 8 minutes), separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and electrotransferred to 0.45 μm polyvinylidene difluoride membranes. The membranes were then blocked (5% skimmed milk, 1 hour), incubated with primary antibodies against NOS2, MAOA, and MIF overnight at 4 °C, after which they were incubated with secondary antibody for 1 hour at room temperature. Images were acquired with a Tanon chemiluminescence imager and analyzed with ImageJ. The same procedures were used to detect these targets in rat liver tissues.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9.0 (GraphPad Software, CA, United States), and data were presented as mean ± SD. For multiple comparisons, one way analysis of variance followed by Dunnett’s post hoc test was used, and a P-value < 0.05 was considered statistically significant.

RESULTS

FZD ameliorates FFA-induced steatosis in HepG2 cells

Treatment with FZD at concentrations ≥ 300 μg/mL resulted in a significant decrease in cell viability, as determined by the CCK-8 assay, relative to the CON group (P < 0.05, P < 0.01) (Figure 1A). This indicates that FZD at 50-250 μg/mL has no effect on the viability of HepG2 cells. Oil red O staining demonstrated that FFA stimulation markedly elevated intracellular lipid content, whereas FZD treatment significantly attenuated lipid accumulation (P < 0.01, P < 0.001) (Figure 1B and C). As shown in Figure 1D-M, FFA stimulation significantly elevated the levels of total cholesterol (TC), triglyceride (TG), alanine aminotransferase (ALT), aspartate aminotransferase (AST), low-density lipoprotein cholesterol (LDL-C), tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6 (P < 0.05, P < 0.001) while reducing superoxide dismutase (SOD) and glutathione (GSH) levels (P < 0.05 and P < 0.001). In contrast, FZD treatment markedly decreased the TC, TG, ALT, AST, LDL-C, TNF-α, IL-1β, and IL-6 levels (P < 0.05, P < 0.01, P < 0.001) and concurrently increased SOD and GSH levels (P < 0.01 and P < 0.001). Collectively, these results indicate that FZD effectively counteracts FFA-induced steatosis in HepG2 cells.

Figure 1 Effect of Fuling Zexie decoction on free fatty acid-induced HepG2 cells. A: Viability of HepG2 cells at 50-350 μg/mL concentrations of Fuling Zexie decoction; B: Oil red O stained sections (scale bar = 50 μm); C: Analysis of relative lipid droplet content; D-M: Total cholesterol (D), triglyceride (E), alanine aminotransferase (F), aspartate aminotransferase (G), low-density lipoprotein cholesterol (H), superoxide dismutase (I), glutathione (J), tumor necrosis factor-α (K), interleukin (IL)-1β (L), IL-6 (M) levels in the free fatty acid-induced HepG2 cells. Data were obtained from 3 independent measurements. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. All P values vs the model group. CON: Control group; MOD: Model group; FC: Fenofibrate capsules group; FZD: Fuling Zexie decoction group; HFZD: High-dose Fuling Zexie decoction group; LFZD: Low-dose Fuling Zexie decoction group; TC: Total cholesterol; TG: Triglyceride; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; LDL-C: Low-density lipoprotein cholesterol; SOD: Superoxide dismutase; GSH: Glutathione; TNF: Tumor necrosis factor; IL: Interleukin.

FZD ameliorates the pathological status of HFD-induced MASLD rats

Following 12 weeks of treatment, model rats exhibited a significantly higher body weight than control rats (P < 0.001). In contrast, FZD administration significantly decreased this parameter compared with the MOD (P < 0.001) (Figure 2A and B). Furthermore, no significant differences in mean daily food intake were detected across all experimental groups (Figure 2C). Oil red O staining revealed marked lipid deposition in rat livers. Hematoxylin-eosin staining indicated both hepatic cellular disorganization and increased vacuolization, whereas FZD treatment effectively reduced lipid droplet accumulation and reversed the histopathological changes in MASLD rat livers (Figure 2D and E). As shown in Figure 2F-K, after 12 weeks of HFD feeding, the serum levels of TC, TG, ALT, AST, and LDL-C were significantly elevated (P < 0.01, P < 0.001), whereas high-density lipoprotein cholesterol (HDL-C) was markedly reduced (P < 0.001) in the model rats. By contrast, FZD treatment resulted in significant decreases in TC, TG, ALT, and LDL-C (P < 0.01, P < 0.001) along with a significant increase in HDL-C (P < 0.01, P < 0.001). Notably, serum AST exhibited a declining tendency following FZD administration. Collectively, these data indicate that FZD extract ameliorates dyslipidemia in MASLD rats.

Figure 2 Effect of Fuling Zexie decoction on high-fat diet-induced metabolic dysfunction-associated steatotic liver disease in rats. A: Body weight; B: Body weight in last week; C: Average daily food intake; D: Oil red O stained sections (scale bar = 50 μm); E: Hematoxylin and eosin stained sections (scale bar = 50 μm); F-K: Total cholesterol (F), triglyceride (G), alanine aminotransferase (H), aspartate aminotransferase (I), low-density lipoprotein cholesterol (J), High-density lipoprotein cholesterol (K) levels in the serum; L-V: Total cholesterol (L), triglyceride (M), alanine aminotransferase (N), aspartate aminotransferase (O), low-density lipoprotein cholesterol (P), High-density lipoprotein cholesterol (Q), superoxide dismutase (R), glutathione (S), tumor necrosis factor-α (T), interleukin (IL)-1β (U), IL-6 (V) levels in the liver, n = 6 in each group, and the results are expressed as mean ± SD. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. All P values vs the model group. CON: Control group; MOD: Model group; FC: Fenofibrate capsules group; HFZD: High-dose Fuling Zexie decoction group; LFZD: Low-dose Fuling Zexie decoction group; TC: Total cholesterol; TG: Triglyceride; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol; SOD: Superoxide dismutase; GSH: Glutathione; TNF: Tumor necrosis factor; IL: Interleukin.

As depicted in Figure 2L-V, 12-week HFD feeding markedly elevated the hepatic levels of TC, TG, ALT, AST, LDL-C, TNF-α, IL-1β, and IL-6 (P < 0.05, P < 0.001), while significantly reducing HDL-C, SOD, and GSH (P < 0.001). Conversely, FZD treatment significantly lowered the elevated parameters (P < 0.05, P < 0.001) and restored the depleted antioxidants and HDL-C (P < 0.05, P < 0.001). These findings indicate that FZD extract attenuates hepatic lipid disturbances, oxidative stress, and inflammatory responses in MASLD rats.

FZD mitigates metabolic disorders in the serum, liver, and urine of MASLD rats

Changes in endogenous metabolites were detected by untargeted metabolomics. PCA results showed that samples in the same group were significantly clustered, while samples in different groups were significantly separated (Figure 3A). These findings indicate that the metabolic profiles of serum, liver, and urine differed among the CON, MOD, and FZD groups. Furthermore, OPLS-DA was employed to assess the metabolic profile differences among groups, and the analysis revealed clear separations between CON and MOD, as well as between FZD and MOD (Figure 3B-D). This confirms that there are distinct metabolic profiles among the three groups. Moreover, a permutation test (n = 200) was conducted on the OPLS-DA model, yielding R²Y values close to 1 and Q² values > 0.4, confirming the model’s robustness (Table 1).

Figure 3 Metabolome profiling of Fuling Zexie decoction in the serum, liver, and urine in rats with metabolic dysfunction-associated steatotic liver disease. A: Principal component analysis score plots; B-D: Orthogonal least partial squares discriminant analysis plot and permutation test plot of serum (B), liver (C), and urine (D); E: Volcano plot of differential metabolite expression; F: Bubble plot of metabolic pathway enrichment analysis. CON: Control group; MOD: Model group; FC: Fenofibrate capsules group; FZD: Fuling Zexie decoction group; QC: Quality control group; ESI+: Positive ion mode; ESI-: Negative ion mode; PC: Principal component; VIP: Variable importance in projection; FC: Fold change; KEGG: Kyoto Encyclopedia of Genes and Genomes; ESI: Electrospray ionization.

Table 1 Orthogonal least partial squares discriminant analysis model parameters.

Model	Group	R2Y value			Q2 value
		Serum	Liver	Urine	Serum	Liver	Urine
ESI+	CON vs MOD	0.998	0.998	0.998	0.777	0.866	0.968
	FZD vs MOD	0.997	0.994	0.997	0.799	0.412	0.962
ESI-	CON vs MOD	0.998	0.992	0.993	0.873	0.882	0.962
	FZD vs MOD	0.999	0.998	0.998	0.906	0.576	0.979

CON: Control group; MOD: Model group; FZD: Fuling Zexie decoction group; ESI: Electrospray ionization; ESI+: Positive ion model; ESI-: Negative ion model.

As shown in Figure 3E, differential metabolites were identified using VIP > 1.0 and P < 0.05 as selection criteria. Following FZD treatment, 226 differential metabolites were significantly modulated in MASLD rats, with 30 in serum (Supplementary Table 1), six in liver (Supplementary Table 2), and 192 in urine (Supplementary Table 3). Subsequently, KEGG pathway enrichment analysis was conducted to elucidate the metabolic pathways associated with FZD intervention. As depicted in Figure 3F, FZD modulated 17 metabolic pathways in MASLD rats, primarily encompassing amino acid, lipid, carbohydrate, cofactor, vitamin, terpenoid, polyketide metabolism, biosynthesis of other secondary metabolites, and xenobiotic biodegradation and metabolism.

FZD restores the dysbiosis of gut microbiota in MASLD rats

16S rRNA sequencing was employed to investigate the impact of FZD on the intestinal microbiota of MASLD rats. Alpha diversity analysis revealed no significant differences in the Ace, Chao, and Sobs indices between the FZD and MOD groups, suggesting that FZD did not markedly affect the diversity of the gut microbial community (Figure 4A-C). PCA and PCoA were applied to assess the dissimilarity in microbial community structure across samples. The analysis revealed a clear separation between the FZD and MOD groups, suggesting that FZD attenuated the gut microbiota dysbiosis in MASLD rats (Figure 4D and E).

Figure 4 Effect of Fuling Zexie decoction on gut microbiota in rats with metabolic dysfunction-associated steatotic liver disease. A: Chao1 index; B: Observed species index; C: Abundance-based coverage estimator index; D: Principal component analysis; E: Principal coordinate analysis; F: Species composition at the phylum level of intestinal flora in rats; G: Species composition at the genus level of intestinal flora in rats; H: Firmicutes/Bacteroidota ratio; I: Relative abundance of Unclassified_f_Lachnospiraceae; J: Relative abundance of Candidatus Saccharimonas; K: Relative abundance of Turicibacter. ^aP < 0.05. All P values vs the model group. CON: Control group; MOD: Model group; FC: Fenofibrate capsules group; HFZD: High-dose Fuling Zexie decoction group; OTU: Operational taxonomic unit; PCA: Principal component analysis; PCoA: Principal coordinate analysis; PC: Principal component; F/B: Firmicutes/Bacteroidota.

Community bar charts were employed to further analyze the taxonomic composition of each group across different taxonomic ranks. At the phylum level, the five most abundant bacterial phyla in the FZD group were identified as Firmicutes, Actinobacteriota, Bacteroidota, Patescibacteria, and Desulfobacterota (Figure 4F). At the genus level, after FZD treatment, the relative abundances of Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter were significantly decreased (P < 0.05) compared with the MOD group, FZD treatment significantly reduced the Firmicutes/Bacteroidota ratio compared with the MOD group (P < 0.05) (Figure 4G-K). These observations imply that FZD reshaped the gut microbial community structure and restored its homeostasis in MASLD rats.

Association analysis of intestinal microbiota, pathological parameters, and differential metabolites in response to FZD treatment

Unclassified_f_Lachnospiraceae showed a significant positive correlation with serum TC, TG, ALT, and LDL-C levels in the Spearman correlation analysis between gut microbiota and MASLD pathological indicators (P < 0.05, P < 0.01, P < 0.001). Candidatus Saccharimonas was identified to be significantly negatively correlated with the serum level of HDL-C (P < 0.01). Turicibacter exhibited a significant positive correlation with serum TC, ALT, and LDL-C levels (P < 0.05, P < 0.01), whereas it showed a significant negative correlation with HDL-C (P < 0.001) (Figure 5A). In addition, as shown in Figure 5B, Unclassified_f_Lachnospiraceae exhibited significant positive correlations with hepatic AST and TNF-α (P < 0.01) and a significant negative correlation with hepatic SOD (P < 0.05). Candidatus Saccharimonas showed a significant positive association with liver IL-6 (P < 0.05). Turicibacter demonstrated significant positive correlations with hepatic TNF-α, IL-1β, and IL-6 (P < 0.01, P < 0.001) and significant negative correlations with GSH, HDL-C, and SOD (P < 0.05, P < 0.01). These findings demonstrate that Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter may act as the core gut microbial taxa through which FZD exerts its effects in restoring gut microbiota homeostasis and alleviating MASLD. To further examine the relationship between gut microbiota and endogenous metabolites, we performed Spearman correlation analysis using genus-level microbial taxa and FZD-regulated differential metabolites.

Figure 5 Spearman’s correlation analysis among gut microbiota, pathological indicators, and differential metabolites regulated by Fuling Zexie decoction. Spearman’s correlation heatmap between gut microbiota at the genus level and serum pathological indicators (A), liver pathological indicators (B), serum differential metabolites (C), liver differential metabolites (D), and urine differential metabolites (E). ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. TC: Total cholesterol; TG: Triglyceride; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol; SOD: Superoxide dismutase; GSH: Glutathione; TNF: Tumor necrosis factor; IL: Interleukin.

As shown in Figure 5C, Unclassified_f_Lachnospiraceae was identified to be significantly negatively correlated with the serum fenofibric acid and ferulic acid 4-O-sulfate (P < 0.05, P < 0.001). Candidatus Saccharimonas was observed to be significantly negatively correlated with the serum flavonol 3-O-D-glucoside, phosphatidylethanolamine (16:0/0:0), and mayzent (P < 0.05). Turicibacter was found to be significantly negatively correlated with the serum indolelactic acid, ornithine, ferulic acid 4-O-sulfate, and 3-phenyllactic acid (P < 0.05, P < 0.001), while significantly positively correlated with the serum cytidine diphosphate diacylglycerol [18:3 (9Z, 12Z, 15Z)/22:6 (4Z, 7Z, 11E, 13Z, 15E, 19Z)-2OH (10S, 17)] (P < 0.05).

As shown in Figure 5D, Unclassified_f_Lachnospiraceae was observed to be significantly positively correlated with the liver 7-imidazo[1,2-a]pyridin-6-ylisoquinolin-1-amine (P < 0.05). Turicibacter was identified to be significantly negatively correlated with the liver rubiadin, 4-guanidinobutanoic acid, and glutaminylvaline (P < 0.05).

As depicted in Figure 5E, Unclassified_f_Lachnospiraceae was observed to be significantly positively correlated with the urine methyl (Z)-5-(1-propynyl)-2-thienylacrylate, 2-methylhippuric acid, p-cresol glucuronide (P < 0.05), while showing a significant negative correlation with the urine 2-guanidinobutanoic acid, 4-methoxy-3-(sulfooxy)benzoic acid, americanin B, tavulin, dihydroferulic acid 4-O-sulfate, [(2S,3S,4S,5S,6R)-4,5-dihydroxy-2,6-dimethyloxan-3-yl] hydrogen sulfate, chrysin-7-O-glucuronide, salicyluric acid, p-aminobenzoic acid, daidzein 4’-O-glucuronide, homarine, pinocembrin, vinclozolin M2, L-1,2,3,4-tetrahydro-beta-carboline-3-carboxylic acid, and protocatechuic acid 3-O-sulfate (P < 0.05, P < 0.001). Turicibacter was discovered to be significantly positively correlated with the urine D-erythroascorbic acid 1’-a-D-xylopyranoside, solerol, stercobilin, methyl (Z)-5-(1-propynyl)-2-thienylacrylate, 2-methylhippuric acid, and p-cresol glucuronide (P < 0.05, P < 0.001), while significantly negatively correlated with the urine cefcapene, pyrocatechol sulfate, osmundalactone, protocatechuic acid, N-acetyl-L-glutamate 5-semialdehyde, ethinamate, 2-hydroxynevirapine, 5,6-dihydroxyprostaglandin F1a, 4-methoxy-3-(sulfooxy)benzoic acid, americanin B, tavulin, chrysin-7-O-glucuronide, salicyluric acid, p-aminobenzoic acid, daidzein 4’-O-glucuronide, homarine, pinocembrin, and L-1,2,3,4-tetrahydro-beta-carboline-3-carboxylic acid (P < 0.05, P < 0.001). Collectively, these findings indicate that FZD attenuated HFD-induced MASLD by modulating the abundance of key gut microbiome taxa and restoring corresponding metabolite levels.

FZD-derived potential active compounds against MASLD

UPLC/MS was employed to profile the FZD-derived compounds in the serum of MASLD rats. As shown in Table 2, 39 FZD-derived compounds were identified in the serum. Among these compounds, there were 21 terpenoids, five aromatic compounds, five gingerols and derivatives, four flavonoids, one monoacylglycerol, one aliphatic dicarboxylic acid, one alkaloid, and one steroid.

Table 2 Fuling Zexie decoction-derived compounds in serum of rats with metabolic dysfunction-associated steatotic liver disease.

tR (minute)	[M+H]+ [M+Na]+ m/z	ESI-MSn (+) m/z	[M-H]- m/z	ESI-MSn (-) m/z	Predicted formula	Predicted (m/z)	Measured (m/z)	Diff (ppm)	Assigned identification	Source	Type
1.572	118.0864	118.0854			C5H11NO2	118.0863	118.0864	0.85	Betaine	AMR	Alkaloid
2.729	263.1673	263.878, 163.1489, 145.0991, 117.0701			C16H22O3	263.1649	263.1673	9.12	8beta-Methoxyatractylenolide I	AMR	Sesquiterpenoids
3.091	133.0648	105.0697, 115.0542			C9H8O	133.0647	133.0648	0.75	Trans-cinnamaldehyde	CR	Aromatic aldehydes
3.195	119.0350	119.0863			C4H6O4	119.0339	119.0350	9.24	Succinic acid	CR	Aliphatic dicarboxylic acids
3.352	149.0595	130.9506, 119.9283, 103.944, 102.9436			C9H8O2	149.0597	149.0595	-1.34	Cinnamic acid	CR	Aromatic carboxylic acids
5.184	217.1585	217.0204, 199.0299, 189.0469, 175.0313, 161.018, 119.0869, 107.0496			C15H20O	217.1587	217.1585	-0.92	Atractylone	AMR	Sesquiterpenoids
7.496	475.3741	457.3129			C30H50O4	475.3782	475.3741	-8.62	11-deoxyalisol A	AR	Triterpenoids
8.113	455.3501	295.1916			C30H46O3	455.352	455.3501	-4.17	Dehydrotrametenolic acid	PR	Triterpenoids
10.392	433.1126	416.8033, 398.7295, 367.146, 323.2313, 165.1012			C21H20O10	433.1129	433.1126	-0.69	Vitexin	GRR	Flavonol glycosides
10.401	433.1134	257.0902, 239.0705, 137.0232			C21H20O10	433.1129	433.1134	1.15	Liquiritigenin-4-O-glucuronide	GRR	Flavanone glycosides
11.530	271.0964	148.9916			C16H14O4	271.0965	271.0964	-0.37	4,4’-dihydroxy-2-methoxychalcone	GRR	Chalcones
11.834	277.1800	259.5875, 189.0878			C17H24O3	277.1798	277.1800	0.72	6-Shogaol	ZRR	Shogaols
12.867	163.0393	163.1483, 135.0831, 107.0865			C9H6O3	163.039	163.0393	1.84	7-Hydroxycoumarin	AMR	Coumarins
13.118	347.2215	177.1277, 137.0955			C21H30O4	347.2217	347.2215	-0.58	1-Dehydro-[10]-gingerdione	ZRR	Dehydrogingerdiones
13.3	429.1734	295.1494, 163.1091			C20H28O10	429.1755	429.1734	-4.89	Cinnamylalcohol-6’-o-α-furanara-binose-O-β-glucopyranoside	CR	Aromatic alcohol glycosides
14.149	233.1538	233.1524, 205.0327, 189.0586, 159.0813, 145.0636, 133.1017, 131.0857, 105.0699			C15H20O2	233.1536	233.1538	0.86	Atractylenolide II	AMR	Sesquiterpenoids
16.872	415.3211	157.0082, 143.0864, 129.0775, 105.0696			C27H42O3	415.3207	415.3211	0.96	Diosgenin	AMR	Steroids
18.027	231.1380	231.0799, 189.0884, 161.0607, 128.0625, 105.0311			C15H18O2	231.138	231.1380	0.00	Atractylenolide Ι	AMR	Sesquiterpenoids
18.314	469.3314	391.0493			C30H44O4	469.3312	469.3314	0.43	3-oxo-16α-hydroxylanosta-7, 9(11), 24-trien-21-oic acid	PR	Triterpenoids
18.314	469.3314	471.2854, 337.0586			C30H44O4	469.3312	469.3314	0.43	Alisol L	AR	Triterpenoids
27.023	471.3465	453.336, 435.3262, 425.3419			C30H46O4	471.3469	471.3465	-0.85	16α-hydroxydehydrotrametenolic acid	PR	Triterpenoids
27.084	453.3363	435.3329, 339.2752			C30H44O3	453.3363	453.3363	0.00	Dehydrotrametenonic acid	PR	Triterpenoids
27.627	471.3467	471.3462, 453.3377			C30H46O4	471.3469	471.3467	-0.42	16,23-oxidoalisol B	AR	Triterpenoids
28.744	471.3474	453.3377, 189.1636			C30H46O4	471.3469	471.3474	1.06	Glycyrrhetinic acid	GRR	Triterpenoids
28.849	203.1792	161.0607, 105.0697			C15H22	203.1794	203.1792	-0.98	Atractylenolide VI	AMR	Sesquiterpenoids
33.482	293.1774	155.0371, 112.9707			C17H24O4	293.1747	293.1774	9.21	[6]-gingerdione	ZRR	Gingerdiones
33.962	331.2841	331.2614			C19H38O4	331.2843	331.2841	-0.60	2-monopalmitin	AMR	Monoacylglycerols
34.949	319.1932	151.0802			C19H26O4	319.1904	319.1932	8.77	1-dehydro-[8]-gingerdione	ZRR	Gingerdiones
35.241	531.3684	167.1446			C32H50O6	531.3680	531.3684	0.75	3-O-acetyl-16α, 26-dihydroxytrametenolic acid	PR	Triterpenoids
4.398			167.0352	167.0228, 123.0091	C8H8O4	167.035	167.0352	1.20	Vanillic acid	CR	Phenolic acids
7.341			499.3416	437.1592, 421.3151	C31H48O5	499.3429	499.3416	-2.60	16α, 25-dihydroxyeburiconic acid	PR	Triterpenoids
8.385			255.0665	237.2247, 135.7461	C15H12O4	255.0663	255.0665	0.78	Liquiritigenin	GRR	Flavanones
13.728			281.1395	281.2485, 263.238, 245.2271	C15H22O5	281.1394	281.1395	0.36	4, 6-dihydroxy-3, 3a-dihydro atractylenolide III	AMR	Sesquiterpenoids
17.552			469.3658	339.2162	C31H50O3	469.3687	469.3658	-6.18	Eburcoic acid	PR	Triterpenoids
17.792			499.3066	367.1447	C30H44O6	499.3065	499.3066	0.20	3-oxo-16-oxo-11-anhydroalisol A	AR	Triterpenoids
18.757			487.3426	453.279, 435.0019, 381.019	C30H48O5	487.3429	487.3426	-0.62	Alisol F	AR	Triterpenoids
25.201			497.3272	423.8392, 379.2716	C31H46O5	497.3272	497.3272	0.00	Poricoic acid A	PR	Triterpenoids
34.320			347.2204	211.843	C21H32O4	347.2227	347.2204	-6.62	10-gingerdione	ZRR	Gingerdiones
35.423	531.3683	309.2785, 391.1912			C32H50O6	531.368	531.3683	0.56	Alisol F-24-acetate	AR	Triterpenoids

ESI: Electrospray ionization; MS: Mass spectrometry; PR: Poria; AR: Alismatis Rhizoma; AMR: Atractylodis Macrocephalae Rhizoma; CR: Cinnamomi Ramulus; GRR: Glycyrrhizae Radix et Rhizoma; ZRR: Zingiberis Rhizoma Recens.

Table 3 Docking information between Fuling Zexie decoction-derived key compounds and core targets.

Component	Target	Binding-energy (kcal/mol)	Amino acid residues
8beta-Methoxyatractylenolide I	PLA2G2A	-5.2	Gly-32
Liquiritigenin	PLA2G2A	-6.1	Lys-62
Atractylenolide II	MAOA	-9	Ile-325, Ile-335, Ile-180, Phe-208, Phe-352, Leu-337, and Glu-216
7-Hydroxycoumarin	MAOA	-7.8	Val-244, Leu-277, and Ala-44
[6]-gingerdione	MAOA	-7.9	Glu-216, Tyr-407
4,4’-Dihydroxy-2-methoxychalcone	MAOA	-9.3	Glu-43, Ala-44, and Arg-51
1-Dehydro-[8]-gingerdione	MAOA	-8.9	Val-244, Glu-43, Arg-45, Gly-21, and Gly-22
1-Dehydro-[10]-gingerdione	MAOA	-8.1	Tyr-444
Atractylenolide II	MPO	-5.1	Glu-181
7-Hydroxycoumarin	XDH	-5.8	Arg-1053, Ser-1185, and Gln-562
7-Hydroxycoumarin	MIF	-8	Asn-97, Ile-64, and Lys-32
[6]-gingerdione	MIF	-7.1	Tyr-36, Tyr-95, Lys-32, and Pro-1
6-Shogaol	MIF	-6.8	Lys-32, Asn-97
Dehydrotrametenonic acid	MIF	-4.7	Pro-15, Glu-21
1-dehydro-[8]-gingerdione	MIF	-6.3	His-40, His-62, Asn-97, and Tyr-99
1-dehydro-[10]-gingerdione	MIF	-6.1	Tyr-99, His-62, Ser-60, and Asn-6
10-gingerdione	MIF	-6.5	Tyr-99, His-62
7-Hydroxycoumarin	CYP1A2	-8.8	Gly-316, Asp-313
[6]-gingerdione	CYP1A2	-7.8	Thr-124
7-Hydroxycoumarin	ALDH2	-5.9	Arg-329
6-Shogaol	ALDH2	-7.3	Asp-457, Gln-289
10-gingerdione	ALDH2	-8.8	Leu-269, Trp-168
7-Hydroxycoumarin	ACHE	-7.6	His-447
[6]-gingerdione	ACHE	-8.1	Ser-125, Tyr-124, Gly-121, and Ser-203
4,4’-dihydroxy-2-methoxychalcone	ACHE	-10.0	His-447, Tyr-341
Liquiritigenin	ACHE	-9.6	Trp-86, Gly-121, Gly-122, and Ser-203
Eburcoic acid	ACHE	-9.2	Phe-295
16,23-oxidoalisol B	NOS2	-11.2	Arg-381
Alisol F-24-acetate	NOS2	-9.3	Trp-463, Tyr-491, Phe-369, Trp-194, and Ile-201
Trans-cinnamaldehyde	NOS2	-7.5	Phe-369, Leu-209, and Trp-194
Glycyrrhetinic acid	NOS2	-8.9	Lys-497
4,4’-dihydroxy-2-methoxychalcone	NOS2	-9.5	Trp-372, Tyr-489
Dehydrotrametenonic acid	NOS2	-10	Trp-372
Eburcoic acid	NOS2	-10	Trp-372
16α, 25-dihydroxyeburiconic acid	NOS2	-10	Arg-381
Dehydrotrametenolic acid	NOS2	-9.8	Trp-461, Ile-462, Phe-476, Trp-463, and Ile-201
16α-hydroxydehydrotrametenolic acid	NOS2	-10.2	Arg-381, Val-465
11-deoxyalisol A	NOS2	-9.1	E377, Y347
Alisol F-24-acetate	GSTM1	-6.4	Trp-7, Leu-59, and Asn-58
Alisol F	TKT	-8.5	Gln-290, Lys-144, Lys-140, and Lys-232
Trans-cinnamaldehyde	ADH4	-6	Phe-325, Tyr-95, and Thr-184
Cinnamic acid	MAOB	-7.5	Gly-12, Gly-13, and Arg-36
4,4’-dihydroxy-2-methoxychalcone	MAOB	-10.1	Thr-196
Liquiritigenin	MAOB	-9.1	Gly-434
6-Shogaol	GSTP1	-5.6	Arg-70, Asp-94, Arg-70, Asn-66, and Gln-64
Dehydrotrametenonic acid	PLA2G4A	-5.6	Asp-99, Arg-59, and Asp-55
Poricoic acid A	PLA2G4A	-4.8	Thr-104, Thr-90, and Arg-59
16α-hydroxydehydrotrametenolic acid	PLA2G4A	-5.8	Asp-55, Lys-58
Betaine	GOT1	-4	Gly-109, Ser-258, and Ala-110
Betaine	GOT2	-3.5	Ser-118, Glu-119

Gly: Glycine; Lys: Lysine; His: Histidine; Ile: Isoleucine; Phe: Phenylalanine; Leu: Leucine; Glu: Glutamate; Val: Valine; Ala: Alanine; Tyr: Tyrosine; Arg: Arginine; Ser: Serine; Gln: Glutamine; Asn: Asparagine; Pro: Proline; Thr: Threonine; Asp: Aspartate; Trp: Tryptophan.

To identify key targets of FZD-derived compounds in MASLD, network pharmacology was employed. First, 1885 disease targets of MASLD were screened from the GeneCards and OMIM databases. In total, 626 targets of drug-derived compounds were predicted via SwissTarget. Next, 183 anti-MASLD targets of FZD-derived compounds were obtained by intersecting these two sets of targets via Venny 2.1.0 (Figure 6A). Subsequently, the KEGG database was employed to extract 590 metabolic pathway related targets from the 17 pathways. Seventeen key targets were subsequently obtained by intersecting the 183 anti-MASLD targets of FZD-derived compounds with the 590 metabolic pathway-related targets via Venny 2.1.0, including PLA2G2A, MAOA, MPO, XDH, MIF, CYP1A2, ALDH2, ACHE, NOS2, GSTM1, TKT, ADH4, MAOB, GSTP1, PLA2G4A, GOT1, and GOT2 (Figure 6B).

Figure 6 Integrated analysis of Fuling Zexie decoction-derived compounds and metabolomics. A: Venn diagram of Fuling Zexie decoction (FZD)-derived compound targets and metabolic dysfunction-associated steatotic liver disease (MASLD) targets; B: Venn diagram of anti-MASLD targets of FZD-derived compounds and FZD-regulated metabolic pathway targets; C: The “active compound-target-pathway-metabolite” network; D: Molecular docking between key components and core targets. FZD: Fuling Zexie decoction group; MASLD: Metabolic dysfunction-associated steatotic liver disease.

To further identify the key bioactive constituents and pathways of FZD against MASLD, a compound-target-pathway-metabolite network was constructed using Cytoscape. As shown in Figure 6C, 17 key targets were associated with 24 FZD-derived compounds (16-HA, 11-DA, Bet, AF-24, D10G, t-CA, Liq, 7-HC, 16,23-OB, CA, DLA, PA, AT-II, 8-MA, EA, GA, AF, 6-SG, D-2-MC, dehydrotrametenonic acid, 16α, 25-dihydroxyeburiconic acid,[6]-gingerdione, 1-dehydro-[8]-gingerdione, and 10-gingerdione), 10 metabolic pathways (arginine biosynthesis, arginine and proline metabolism, phenylalanine, tyrosine, and tryptophan biosynthesis, phenylalanine metabolism, tyrosine metabolism, tryptophan metabolism, alanine, aspartate, and glutamate metabolism, glycerophospholipid metabolism, pentose phosphate pathway, and drug metabolism-other enzymes), and 18 metabolites, PC [22:2 (13Z, 16Z)/16:0], serotonin, ornithine, agmatine, N-acetyl-L-glutamate 5-semialdehyde, 4-guanidinobutanoic acid, gentisic acid, vanillylmandelic acid, 3,4-dihydroxyphenylacetic acid, 3,4-dihydroxymandelic acid, phenylpyruvic acid, m-coumaric acid, 6-thioinosine-5’-monophosphate, citrulline, gluconolactone, 3-dehydroshikimate, N-acetyl-1-aspartylglutamic acid, and glucosamine 6-phosphate. These results suggest that 24 FZD-derived compounds might serve as the key active compounds of FZD, acting on 17 key targets, regulating 10 key metabolic pathways, and affecting 18 metabolites to alleviate MASLD. Molecular docking was also performed to assess the binding affinities between the 24 components and their 17 corresponding targets, and the results demonstrated that these FZD-derived compounds formed stable interactions with the target proteins (Table 3 and Figure 6D).

FZD-derived key actives counteracted FFA-induced lipid overload in HepG2 cells

The in vitro anti-hepatic steatosis activity of FCG, comprising 19 key active compounds, reference standards for 16α, 25-dihydroxyeburiconic acid, dehydrotrametenonic acid, 1-dehydro-[8]-gingerdione, 10-gingerdione, and[6]-gingerdione were not available and thus excluded from the compound mixture, was determined. Oil red O staining revealed that FFA stimulation markedly elevated the intracellular lipid content (P < 0.001). Treatment with FCG significantly prevented the accumulation of lipids in the cells (P < 0.001) (Figure 7A and B). Following 24 hours FFA exposure, HepG2 cells exhibited significant increases in TC, TG, ALT, AST, TNF-α, IL-1β, and IL-6 (P < 0.05, P < 0.01, P < 0.001), along with significant decreases in SOD and GSH (P < 0.001). In contrast, FCG treatment significantly reversed the TC, TG, ALT, AST, TNF-α, IL-1β, and IL-6 levels (P < 0.05, P < 0.01, P < 0.001) and raised the SOD and GSH levels (P < 0.05, P < 0.01, P < 0.001) of the cells (Figure 7C-K).

Figure 7 Effect of Fuling Zexie decoction-derived key active compounds on free fatty acid-induced HepG2 cells. A: Oil red O stained sections (scale bar = 50 μm); B: Analysis of relative lipid droplet content; C-K: Total cholesterol (C), triglyceride (D), alanine aminotransferase (E), aspartate aminotransferase (F), superoxide dismutase (G), glutathione (H), tumor necrosis factor-α (I), interleukin (IL)-1β (J), IL-6 (K) levels in the free fatty acid-induced HepG2 cells; L: The expression of NOS2, MAOA and MIF in HepG2 cells after Fuling Zexie decoction (FZD)-derived key compound group treatment; M: The expression of NOS2, MAOA and MIF in the liver of rats with metabolic dysfunction-associated steatotic liver disease after FZD treatment. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. All P values vs the model group. CON: Control group; MOD: Model group; FC: Fenofibrate capsules group; HFCG: High-dose Fuling Zexie decoction-derived key compound group; LFCG: Low-dose Fuling Zexie decoction-derived key compound group; TC: Total cholesterol; TG: Triglyceride; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; SOD: Superoxide dismutase; GSH: Glutathione; TNF: Tumor necrosis factor; HFZD: High-dose Fuling Zexie decoction group; LFZD: Low-dose Fuling Zexie decoction group.

Moreover, the protein levels of NOS2, MAOA, and MIF were markedly decreased in FCG-treated cells (P < 0.05, Figure 7L). Similarly, administration of FZD extract to MASLD rats led to a significant downregulation of NOS2, MAOA, and MIF in liver tissue (P < 0.05, P < 0.01, Figure 7M). These results indicate that FCG inhibits the expression of NOS2, MAOA, and MIF, thereby improving hepatocellular steatosis.

DISCUSSION

With rising living standards, the global incidence of MASLD is increasing rapidly. To date, the Food and Drug Administration has approved only two drugs for advanced metabolic associated steatohepatitis (resmetirom and semaglutide), and effective treatments for early- to mid-stage MASLD remain lacking. Therefore, MASLD has been recognized as a significant worldwide public health challenge[16]. Current studies suggest that metabolic disturbances, including fatty infiltration, impaired insulin sensitivity, oxidative stress, inflammatory activation, and gut microbial imbalance, are key factors leading to MASLD[17]. Hepatic lipid accumulation usually manifests as abnormal changes in the indicators of lipid metabolism and liver function, such as TC, TG, ALT, AST, LDL-C, and HDL-C[18]. Hepatic lipid accumulation induces mitochondrial respiratory chain dysfunction, which in turn generates excessive reactive oxygen species (ROS)[19]. The resultant ROS subsequently activate the c-Jun N-terminal kinase (JNK) and nuclear factor kappa-B (NF-κB) inflammatory pathways, leading to enhanced transcription and production of pro-inflammatory mediators, including TNF-α, IL-6, and IL-1β[20]. This ultimately contributes to the amplification of inflammation in MASLD. In addition, SOD and GSH are important antioxidants in the liver, which clear ROS to combat oxidative stress[21]. In this study, the therapeutic efficacy of FZD against MASLD was evaluated using both an FFA-induced HepG2 cell steatosis model and an HFD-fed rat model. In vitro, FZD treatment significantly attenuated the elevated levels of TC, TG, ALT, AST, LDL-C, TNF-α, IL-1β, and IL-6, while concomitantly elevating GSH and SOD in FFA-exposed cells; lipid droplet accumulation was also markedly reduced. In vivo, FZD administration to MASLD rats not only decreased hepatic lipid droplets and restored hepatocellular histomorphology, but also significantly lowered serum and hepatic TC, TG, ALT, and LDL-C, elevated HDL-C, and increased liver SOD and GSH, along with reductions in AST, TNF-α, IL-1β, and IL-6. Collectively, these results demonstrate that FZD alleviates MASLD by ameliorating lipid dysregulation, oxidative stress, and inflammatory responses.

Metabolomics is an analytical technique that profiles metabolites within biological systems to uncover their associations with physiological and pathological states[22]. In this study, following FZD treatment, 226 differential metabolites in MASLD rats were significantly regulated by FZD, which were related to 17 metabolic pathways, mainly including amino acid, lipid, carbohydrate, cofactor, vitamin, terpenoid, polyketide, and carbohydrate metabolism, biosynthesis of other secondary metabolites, and xenobiotic biodegradation and metabolism.

Amino acid metabolic disorders serve as predisposing factors for MASLD[23]. Ornithine, a precursor of L-glutamine, L-arginine and GSH, suppresses lipid peroxidation, improves hepatic microcirculation, and exerts anti-inflammatory and antioxidant effects[24]. Citrulline, synthesized from ornithine, CO₂ and ammonia in the liver, confers protective benefits against MASLD[25,26]. Derived from arginine, agmatine preserves mitochondrial function, regulates fatty acid metabolism, reduces plasma TG/HDL-C ratio, and suppresses hepatic steatosis and adipogenesis in HFD-fed apolipoprotein E-knockout mice[27]. Gentisic acid alleviates diabetic liver damage by reducing fasting blood glucose, normalizing liver enzyme profiles and boosting hepatic antioxidant capacity[28]. A prior study of carnosic acid (CA) against MASLD-triggered dopaminergic neuronal injury reported elevated 3,4-dihydroxyphenylacetic acid in CA-treated rats, revealing its therapeutic potential for MASLD[29]. Glucosamine 6-phosphate initiates the hexosamine biosynthesis pathway, a key driver of insulin resistance[30]. In our HFD-induced MASLD rat model, FZD intervention upregulated ornithine, citrulline, agmatine, gentisic acid and 3,4-dihydroxyphenylacetic acid while downregulating glucosamine 6-phosphate. Collectively, FZD ameliorates MASLD by restoring disrupted amino acid metabolism.

Lipid metabolic disorders are widely recognized as a critical driver of the pathogenesis and progression of MASLD. Phosphatidylcholine (PC), a core molecule in the glycerophospholipid metabolic pathway, exerts a crucial effect on the maintenance of cell membrane structure, signal transduction, and regulation of lipid homeostasis[31]. The synthesis of very LDL is promoted by PC, which in turn facilitates lipid export from the liver and reduces the accumulation of TG[32]. Clinical studies further confirm that PC supplementation in MASLD patients has a positive effect on hepatic steatosis[33]. It is noteworthy that FZD administration increased PC [22:2 (13Z, 16Z)/16:0] levels. These findings suggest that FZD attenuates MASLD through the modulation of lipid metabolism dysregulation.

The biosynthesis of other secondary metabolites is closely associated with MASLD. Pinocembrin is involved in the biosynthesis of flavonoids and improves lipidosis in HepG2 cells by activating the sirtuin 1 (SIRT1)/adenosine 5’-monophosphate-activated protein kinase (AMPK) pathway. Isoliquiritigenin, an intermediate product in the biosynthesis of flavonoids, alleviates MASLD in mice by inhibiting macrophage inflammatory activity through the blockade of spleen tyrosine kinase-induced inflammasome activation[34]. Additionally, it mitigates steatohepatitis and fibrosis in mice by promoting autophagy through the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt)/mammalian target of rapamycin signaling pathway[35]. Naringenin is involved in both flavonoid biosynthesis and isoflavone biosynthesis and possesses significant anti-inflammatory effects.

This anti-inflammatory effect is mediated by suppressing the production of proinflammatory cytokines (notably TNF-α and IL-6), blocking the NF-κB signaling cascade, and attenuating cyclooxygenase-2 expression[36]. A randomized clinical trial evaluated the effects of naringenin therapy in MASLD patients and reported beneficial changes in their lipid profiles[37]. Additionally, a cross-sectional study based on data from 1476 participants revealed that the hepatic steatosis index, fatty liver index, and MASLD prevalence were inversely associated with daidzein intake[38]. In this study, FZD treatment increased the levels of pinocembrin, isoliquiritigenin, naringenin, and daidzein, indicating that FZD ameliorates MASLD by modulating dysregulated secondary metabolite biosynthesis.

Abnormal xenobiotic biodegradation and metabolism drive MASLD progression. 4-hydroxybenzoic acid participates in aminobenzoate and benzoate breakdown. Rats fed polyphenol-enriched diets display elevated hepatic 4-hydroxybenzoic acid alongside improved lipid metabolism and antioxidant capacity, implying this metabolite mediates polyphenol hepatoprotection[39]. Protocatechuic acid, a major anthocyanin metabolite, inhibits lipid uptake and synthesis via peroxisome proliferator-activated receptor gamma (PPARγ) suppression and boosts brown adipose tissue energy expenditure, mitigating hepatic steatosis in HFD-induced MASLD mice[40,41]. Syringic acid, a phenolic intermediate of ammonium benzoate metabolism, exerts hepatoprotective, anti-inflammatory, antidiabetic and anti-tumor activities; it alleviates MASLD in rats by activating the nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase-1 axis to reverse oxidative stress and inflammation[42,43]. Excess carbohydrate intake also fuels MASLD onset. Gluconolactone, a cyclized derivative of gluconic acid in the pentose phosphate pathway, declines in rats with impaired hepatic glucose metabolism[44]. In our study, FZD markedly upregulated hepatic 4-hydroxybenzoic acid, protocatechuic acid, syringic acid and gluconolactone, revealing that FZD relieves MASLD by restoring disrupted xenobiotic biodegradation and metabolic homeostasis.

In the human body, the intestine and liver are directly connected by the portal vein, which enables the gut microbiota to significantly influence liver metabolism. Gut microbiota dysbiosis contributes to the pathogenesis and progression of MASLD. In this study, FZD interventions significantly reduced the Firmicutes/Bacteroidota ratio and the relative abundances of Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter in the intestines of MASLD rats.

The Firmicutes/Bacteroidota ratio is positively correlated with obesity, and thus serves as a marker for intestinal microecological imbalance in obese patients[45]. Unclassified_f_Lachnospiraceae is a group of bacteria associated with obesity and closely linked to the promotion of lipid metabolism disorders. It has been shown that a significant decrease in the abundance of Unclassified_f_Lachnospiraceae may alleviate symptoms in MASLD rats[46,47]. Candidatus Saccharimonas is an opportunistic pathogen whose intestinal abundance is significantly upregulated in patients with gout, obese mice, and MASLD rats[48-50]. Correlated positively with MASLD hepatic fibrosis severity, Turicibacter acts as a potential microbial biomarker for liver fibrosis; it regulates bile-lipid profiles and increases gonadal/epididymal white fat weight in mice[51,52]. Panax japonicus polysaccharide reduces its abundance in MASLD mice, consistent with our findings[53]. We found that the abundances of Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter were significantly positively correlated with serum TC, TG, ALT, and LDL-C, as well as with hepatic AST, TNF-α, IL-1β, and IL-6. Conversely, these taxa exhibited significant negative correlations with HDL-C (both in serum and liver), along with hepatic SOD and GSH. These findings indicate that FZD improves abnormal lipid accumulation, oxidative stress, and inflammatory response by reducing the abundance of Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter, thereby alleviating MASLD.

Gut microbiota compositional changes impact endogenous metabolite profiles. Our correlation analysis indicated that ferulic acid 4-O-sulfate, americanin B, chrysin-7-O-glucuronide, salicyluric acid, and pinocembrin were significantly negatively correlated with Unclassified_f_Lachnospiraceae and Turicibacter.

Additionally, indolelactic acid, ornithine, 3-phenyllactic acid, 4-guanidinobutanoic acid, and protocatechuic acid were significantly inversely associated with Turicibacter. Ferulic acid 4-O sulfate, derived from ferulic acid, protects against oxidative stress in mice and HepG2 cells under iron overload through Nrf2 upregulation and consequent enhancement of antioxidant gene transcription[54]. Americanin B, a lignan compound, increases survival rate and reduces brain injury in lipopolysaccharides-induced sepsis mice, which is achieved by inhibiting pyroptosis and activating the NLRP3 inflammasome[55]. Chrysin-7-O-glucuronide, an intestinal-absorbing compound, inhibits TNF-α-induced tight junction loss in IEC-6 cells and mitigates intestinal mucosal barrier injury in severe acute pancreatitis rats by suppressing NF-κB signaling[56]. Furthermore, it may have anticancer potential by targeting lactate dehydrogenase A[57]. Salicyluric acid is a substance whose level can be restored by aloe-emodin, which is beneficial for treating hyperlipidemic rats[58]. Mechanistic studies indicated that indolelactic acid increased hepatic CYP7A1 and FXR-bile salt export pump levels, stimulating the synthesis and biliary secretion of bile acids, and thus relieving liver steatosis in mouse models on a high-fat and high-sugar diet[59]. 3-Phenyllactic acid alleviates Salmonella Typhimurium-induced colitis through the regulation of microbial community structure, short-chain fatty acids levels, and the inflammatory response[60]. Additionally, ornithine and protocatechuic acid alleviate MASLD in rats. Our findings demonstrate that FZD relieves MASLD by regulating endogenous metabolite levels in a gut microbiota-dependent manner. Notably, 7-imidazo[1,2-a]pyridin-6-ylisoquinolin-1-amine belongs to the differential metabolites strongly correlated with gut microbiota, along with Flavonol 3-O-D-glucoside, PE (16:0/0:0), mayzent, P-cresol glucuronide, fenofibric acid, 2-guanidinobutanoic acid, 4-methoxy-3-(sulfooxy)benzoic acid, tavulin, dihydroferulic acid 4-O-sulfate, [(2S, 3S, 4S, 5S, 6R)-4,5-dihydroxy-2,6-dimethyloxan-3-yl] hydrogen sulfate, P-aminobenzoic acid, daidzein 4’-O-glucuronide, homarine, vinclozolin M2, L-1,2,3,4-tetrahydro-beta-carboline-3-carboxylic acid, protocatechuic acid 3-O-sulfate, 4-guanidinobutanoic acid, cytidine diphosphate diacylglycerol [18:3 (9Z, 12Z, 15Z)/22:6 (4Z, 7Z, 11E, 13Z, 15E, 19Z)-2OH (10S,17)], D-erythroascorbic acid 1’-α-D-xylopyranoside, solerol, stercobilin, methyl (Z)-5-(1-propynyl)-2-thienylacrylate, 2-methylhippuric acid, rubiadin, glutaminylvaline, cefcapene, pyrocatechol sulfate, osmundalactone, N-acetyl-L-glutamate 5-semialdehyde, ethinamate, 2-hydroxynevirapine and 5,6-dihydroxyprostaglandin F1α. None of these metabolites have previously been reported to be associated with MASLD, which warrants further investigation.

Compounds that are systemically absorbed following oral or intragastric administration of TCM are likely the bioactive constituents responsible for its in vivo therapeutic effects. In this study, 39 FZD-derived compounds were identified in the serum. The results from “bioactive compound-target-metabolic pathway-metabolite” network show that 24 key FZD-derived compounds act on 17 targets, regulate 10 key metabolic pathways, and affect 18 metabolites, thereby alleviating MASLD. Among these 24 key FZD-derived compounds, Bet, Liq, CA, and GA have been proven to treat MASLD effectively[61-64]. Alternatively, 6-SG alleviates hepatic steatosis in HFD-induced mice through the miR-3066-5p/Grem2 pathway, and mitigates liver inflammatory and fibrotic responses in metabolic dysfunction-associated steatohepatitis mice via inhibition of oxidative stress, cell death, and endoplasmic reticulum stress[65,66]. 6-gingerdione protects 2% ethyl alcohol-induced zebrafish larvae from liver injury by suppressing inflammatory genes and lipogenic genes while upregulating the antioxidant gene[67]. 7-HC markedly reduced plasma triglycerides, FFA, and glucose, along with adipocyte size, white adipose tissue mass, and hepatic phosphatidate phosphohydrolase activity, while significantly elevating plasma adiponectin and hepatic fatty acid β-oxidation activity, thereby ameliorating hypertriglyceridemia and hyperglycemia[68]. Obesity-induced insulin resistance is attenuated by AT-II, which acts as a potent allosteric modulator of diacylglycerol kinase theta, through the stereospecific numbering-1,2-diacylglycerol-protein kinase C epsilon pathway[69]. D10G inhibits toll-like receptor 4 activation, thereby exerting anti-inflammatory and antioxidant effects in cancer prevention[70]. EA was shown to exert anti-diabetic and lipid-lowering effects through the modulation of Akt phosphorylation, gluconeogenesis, and PPARα in streptozotocin-induced diabetic mice[71]. AF and AF-24, as active constituents of AR, may modulate intestinal microecology and cholesterol metabolism, thereby contributing to diabetes prevention[72]. Furthermore, AF exerts anti-inflammatory activities and liver protection by inhibiting mitogen-activated protein kinase, STAT3, and NF-κB activation in acute liver-injured mice[73]. These findings suggest that the key FZD-derived compounds have the potential to treat MASLD.

NOS2, MAOA, and MIF were the targets associated with the greatest number of FZD-derived components, suggesting that they may serve as key therapeutic targets of FZD. NOS2, an inducible nitric oxide synthase isoform, catalyzes the conversion of L-arginine to nitric oxide and L-citrulline. In inflammatory conditions, it is upregulated and can produce excess nitric oxide, which can cause hepatocellular damage[74]. Furthermore, NOS2 serves as a critical mediator of both inflammation and insulin resistance. In obesity, chronic NOS2 induction disrupts insulin signaling to the PI3K/Akt pathway in metabolic tissues[75]. In T37i brown adipocytes, NOS2 induction has been demonstrated to directly impair mitochondrial respiration, an effect that was completely reversed upon inhibition of NOS2 activity[76]. MAOA is a subtype of monoamine oxidase, which participates in the degradation of dopamine, norepinephrine, tryptamine, and tyramine in most species[77]. It was demonstrated that MASLD is closely associated with upregulated MAOA activity in the liver, and inhibition of MAOA can alleviate liver injury in steatohepatitis mice[78]. Moreover, monoamine oxidase inhibitors have been shown to downregulate the expression of the pro-inflammatory cytokines TNF-α, IL-6, and IL-1β at both the gene and protein levels[79,80]. MIF is a cytokine that plays a critical role in various inflammatory conditions, while also exhibiting endocrine and enzymatic activities[81]. MIF exerts a complex and dual role in MASLD. MIF activates the downstream AMPK/SIRT1 signaling pathway, thereby increasing autophagy and reducing cell apoptosis, and it inhibits activation of the JNK pathway, thereby improving lipid metabolism[82,83]. However, studies have shown that MIF deficiency attenuates HFD-induced insulin resistance in mice by reducing adipose inflammation and hepatic steatosis, whereas elevated MIF levels correlate with a heightened risk of MASLD[84,85].

Notably, the other 14 intersecting targets (PLA2G2A, MPO, XDH, CYP1A2, ALDH2, ACHE, GSTM1, TKT, ADH4, MAOB, GSTP1, PLA2G4A, GOT1, and GOT2) also participate in multiple auxiliary pathways closely related to hepatic metabolism, and they cannot be ignored for a comprehensive understanding of FZD’s multi-target pharmacological characteristics. Our results demonstrated that GOT1 and GOT2 modulate tyrosine metabolism, phenylalanine metabolism, arginine and proline metabolism, alanine, aspartate and glutamate metabolism, phenylalanine, tyrosine and tryptophan biosynthesis, and arginine biosynthesis. ALDH2 participates in arginine and proline metabolism and tryptophan metabolism, whereas MAOB simultaneously regulates arginine and proline metabolism, phenylalanine metabolism, tryptophan metabolism, and tyrosine metabolism. MPO, XDH, GSTM1 and GSTP1 collectively mediate the pathway drug metabolism of other enzymes. PLA2G2A, ACHE and PLA2G4A modulate glycerophospholipid metabolism. Furthermore, TKT regulates the pentose phosphate pathway, CYP1A2 modulates tryptophan metabolism, and ADH4 governs tyrosine metabolism. In the present study, these 14 targets were not validated in the samples. Follow-up systematic experiments will be conducted to detect expression changes in the remaining targets, thereby fully dissecting the multi-gene regulatory network mediated by FZD.

Here, the compound group, comprising 19 active constituents, significantly reduced the protein expression levels of NOS2, MAOA, and MIF in cultured HepG2 cells. In parallel, FZD extract substantially downregulated the expression of these three proteins in the liver tissue of MASLD rats. These results further confirm that the 24 FZD-derived components may be the core pharmacological constituents of FZD against MASLD, acting on 17 targets, regulating 10 metabolic pathways, and modulating 18 endogenous metabolites.

CONCLUSION

FZD alleviated hepatic histopathological changes and lipid overload in MASLD rats, together with modulating lipid metabolism, oxidative stress, and inflammation, an effect likely mediated by regulating 17 pathways, reducing pathogenic microbial abundance (Unclassified_f_Lachnospiraceae, Candidatus Saccharimonas, and Turicibacter), and reversing 226 metabolites. Thirty-nine FZD-derived compounds were detected in the serum of MASLD rats, of which 24 are key effective compounds of FZD against MASLD. These compounds modulate 17 targets, regulate 10 metabolic pathways, and affect 18 metabolites to prevent MASLD. Moreover, the key effective compounds attenuate lipid metabolism disturbances, oxidative stress, and inflammatory responses in fatty HepG2 cells, with their mechanism involving the inhibition of NOS2, MAOA, and MIF expression.