Mechanistic Elucidation of Polysaccharides in Treating MAFLD via Regulation of the Gut Microbiota–Metabolite–Ferroptosis Axis: A Multi-Omics Perspective

Abstract

This study aimed to elucidate the modulatory effects and underlying molecular mechanisms of polysaccharide (TMP) in the context of metabolic dysfunction-associated fatty liver disease (MAFLD). High-performance gel permeation chromatography (HPGPC) analysis indicated a bimodal molecular weight distribution. Monosaccharide composition profiling revealed a predominance of glucose and galactose among other constituents. Scanning electron microscopy (SEM) illustrated a porous, aggregated colloidal microstructure. In a model of MAFLD, TMP intervention significantly attenuated serum levels of TC, TG, and AST, ALT, accompanied by notable histological improvements, including reduced hepatic steatosis and inflammatory cell infiltration. Metagenomic analysis demonstrated that TMP substantially enhanced gut microbial α-diversity, restructured microbial community composition, decreased the Firmicutes/Bacteroidetes ratio, enriched SCFAs-producing genera, and suppressed the excessive proliferation of pro-inflammatory bacterial genera. Integrated proteomic and lipidomic analyses revealed that TMP inhibited hepatic immune-inflammatory responses and ferroptosis pathways, enhanced pathways associated with metabolic homeostasis. Furthermore, TMP modulated hepatic iron metabolism by upregulating the Nrf2/GPx4 antioxidant axis and FPN1 while downregulating TFR1, thereby alleviating oxidative stress and iron overload. These findings demonstrate that TMP exerts therapeutic efficacy through a bidirectional gut-liver regulatory mechanism involving microbial modulation, ferroptosis inhibition, metabolic reprogramming, and activation of antioxidant defenses. This research provides novel insights and molecular targets for the development of natural polysaccharide-based interventions for MAFLD.

1. Introduction

Metabolic dysfunction-associated fatty liver disease (MAFLD) has emerged as the most common chronic liver disorder globally. It has a strong, bidirectional link with the rising incidence of obesity and metabolic syndrome, sharing key pathophysiological mechanisms like insulin resistance, posing a major global health burden. Its potential progression from hepatic steatosis to more severe stages, such as metabolic dysfunction-associated steatohepatitis (MASH), liver fibrosis, and even hepatocellular carcinoma, significantly threatens human health. Recent epidemiological data indicate that the global prevalence of MAFLD has reached approximately 25%, which is particularly high among individuals with obesity and type 2 diabetes.

The pathogenesis of MAFLD is highly complex and multifactorial. Current evidence suggests that disturbances in lipid metabolism, oxidative stress, dysbiosis of the gut microbiota, and ferroptosis are crucial contributors to MAFLD development and progression. − Among these, gut microbiota dysregulation not only directly influences hepatic lipid synthesis and metabolism but also modulates inflammatory cascades and mitochondrial homeostasis, thereby bidirectionally driving the pathophysiological evolution of MAFLD. , Notably, ferroptosisa newly identified form of programmed cell death that is iron-dependenthas been implicated in hepatocyte injury in MAFLD, exerting a pronounced pro-inflammatory effect and playing a critical role in modulating liver pathology. Despite current therapeutic strategies predominantly relying on lifestyle modification combined with pharmacologic interventions such as insulin sensitizers, these approaches often suffer from limited efficacy and considerable adverse effects, highlighting the urgent need for novel, safer treatment modalities. Consequently, natural bioactive compounds with multitarget regulatory properties and low toxicity have attracted significant research interest.

In the context of a flourishing health and wellness industry, health supplements, as a significant category of functional foods, have attracted growing attention due to their dual functions of providing nutritional support and regulating physiological processes, including liver protection. In recent years, multiple studies have demonstrated that various natural polysaccharidesincluding those derived from , inulin, and can exert multifaceted therapeutic effects on the pathological progression of MAFLD by modulating gut microbiota homeostasis and ferroptosis-related pathways, presenting promising clinical potential. −

S. Imai, a wild edible mushroom belonging to the Tricholomataceae family, is indigenous to the grasslands of northern China. Celebrated as a culinary delicacy and known as the premier member of the “Eight Treasures of the Grasslands”, this mushroom is highly valued for its nutritional and therapeutic potential. Previous studies have demonstrated that is rich in bioactive compounds, including polysaccharides, lectins, and polyunsaturated fatty acids, which confer a broad spectrum of biological activities such as immunomodulation, antioxidation, , hypoglycemic effects, gut microbiota regulation, and antitumor properties. , Among these, polysaccharide (TMP) has emerged as the primary focus due to its pronounced biological efficacy. TMP demonstrates robust antioxidant effects ,, and plays a significant role in intestinal digestion, promoting the growth of probiotics. Based on above reports, we hypothesize that TMP improves MAFLD through a bidirectional gut-liver regulatory mechanism that involves microbial modulation, ferroptosis inhibition, metabolic reprogramming, and the activation of antioxidant defenses.

This study elucidated the monosaccharide composition and molecular structural characteristics of TMP. A high-fat diet-induced mouse model of MAFLD was established to evaluate the ameliorative effects of TMP on dyslipidemia, hepatic injury, and histopathological alterations. Furthermore, this study explored TMP’s regulatory effects on gut microbiota composition, along with its systemic modulatory mechanisms at the proteomic and metabolomic levels. The findings indicate that TMP confers hepatoprotective effects by modulating ferroptosis-related pathways in hepatocytes, thereby providing multitiered experimental evidence for its potential application in the prevention and treatment of MAFLD. This study not only offers novel scientific insight into the utility of TMP as a natural intervention for ameliorating MAFLD and its pathological progression but also proposes an innovative strategy for the development of multitarget, synergistic therapeutic approaches for metabolic liver diseases.

2. Materials and Methods

2.1. Materials

was harvested from the Hulunbuir grasslands in China. Biochemical assay kits for triglycerides (TG, 230321201), total cholesterol (TC, 230328201), HDL-C (230403201), LDL-C (230412201), AST (240710101), and ALT (240924101) were purchased from Medical system Biotechnology Co., Ltd. (Ningbo, China). The H&E staining kit (23040567) was purchased from Leica (Wetzlar, Germany). SOD (20240624), GSH (20241129), and MDA (20240419) detection kits came from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). DCFH-DA (D6883) and C11-BODIPY581/591 (SML3717) were derived from Sigma-Aldrich. Antibodies including Nrf2 mouse monoclonal antibody (BF8017), GPx4 antibody (DF6701), ACSL4/FACL4 antibody (DF12141), FTH1 antibody (DF6278) were acquired from Affinity (Ohio). The Pierce BCA Protein Assay Kit (23250) and SuperSignal ECL (34577) were purchased from Thermo Fisher Scientific (Massachusetts). The PrimeScript RT Reagent Kit (RR037A) and TB Green Premix Ex TaqII FAST qPCR Kit (CN830A) were obtained from TAKARA (Dalian, China). Alfa Qubit 1 × dsDNA HS Assay KIT (Finorop, QD001). QSEP400 High-throughput Nucleic Acid and Protein Analysis System (Houze Biology Technology, QSEP400). KAPA HiFi HotStart ReadyMix (Roche, KK2602). NEBNext Ultra II DNA Library Prep Kit (NEB, E7645).Qubit 3.0 Fluorometer (Thermo Fisher Scientific, American). NanoDrop One Microvolume UV–vis Spectrophotometer (Thermo Fisher Scientific,American). Qubit 4.0 Fluorometer (Thermo Fisher Scientific, American). Illumina NovaSeq 6000 Sequencing System (Illumina, American).

2.2. Extraction, Purification, and Characterization of TMP

Extraction and purification (Figure ): Fresh samples were collected, cleaned, air-dried, and ground into a powder (approximately 300 mesh). Defatting was carried out using supercritical CO₂ extraction under the following conditions: extraction pressure of 30 MPa, extraction temperature of 35 °C, separation temperature of 50 °C, and extraction duration of 3 h. The defatted mushroom powder was extracted with distilled water at a material-to-liquid ratio of 1:20 and decocted at 95 °C for 3 h. This extraction was repeated twice, and the combined extracts were concentrated using a rotary evaporator. The crude TMP solution (10 mg/mL) was prepared and deproteinized using Sevage reagent at a 4:1 ratio, followed by oscillation for 20 min, standing for 30 min, and centrifugation at 35g for 15 min. The process was repeated 4–5 times until no visible protein layer was observed. The TMP was precipitated with ethanol and sequentially washed with acetone, ether, and ethanol. After centrifugation, the final precipitate was collected.

1.

Process of TMP extraction and purification.

Molecular Weight Determination: A 10 mg sample of TMP was dissolved in a phosphate buffer solution and analyzed by high-performance gel permeation chromatography (HPGPC) to determine its molecular weight and distribution.

Fourier-Transform Infrared (FT-IR) Spectroscopy: 2 mg of TMP and 200 mg of KBr were accurately weighed, thoroughly mixed, ground, and subsequently compressed into pellets. For the blank control, KBr powder was solely compressed into a pellet. Both samples underwent scanning and data recording using a Fourier-transform infrared spectrometer (FT-IR650).

Monosaccharide Composition Analysis: Precisely 10 mg of TMP was placed in an ampule bottle and hydrolyzed with 10 mL of trifluoroacetic acid (TFA, 3 mol/L) at 120 °C for 3 h. The hydrolysate was then transferred to a tube and dried under nitrogen flow. The residue was redissolved in 10 mL of ultrapure water, diluted 10-fold with deionized water, and centrifuged at 13,900g for 5 min. The supernatant was collected for ion chromatography (IC) analysis. The IC was performed using a Dionex Carbopac PA20 column (3 mm × 150 mm). The mobile phases included A: ultrapure water, B: 15 mM NaOH, and C: a mixture of 15 mM NaOH and 100 mM sodium acetate. The flow rate was set at 0.3 mL/min, with an injection volume of 5 μL. The column temperature was maintained at 30 °C, and quantification was performed using an electrochemical detector.

Scanning Electron Microscopy (SEM): Approximately 5 mg of TMP was mounted onto a conductive carbon tape affixed to an SEM stub. The sample was gold-sputtered for approximately 40 s using an ion sputter coater. Subsequently, the sample was transferred to the SEM chamber and observed under an acceleration voltage of 2 kV.

2.3. MAFLD Mouse Model Construction and TMP Intervention

The animal experiment was approved by the Medical Ethics Committee of Inner Mongolia Medical University (Document No.: YKD202502042) and was conducted in accordance with the National Research Council’s Guide for the Care and Use of Laboratory Animals.

Animal Model and Intervention: Fifty 8-week-old SPF-grade male C57BL/6N mice were acclimatized for 2 weeks. Then, they were randomly assigned into five groups (n = 10/group): control, model, and TMP low-, medium-, and high-dose groups. The control group was fed a standard chow diet, while all other groups received a high-fat, high-sugar diet (purchased from Shuangzhi Trading Co., Hohhot, China). From the fourth week onward, mice in the TMP treatment groups were orally administered TMP at doses of 0.4, 0.8, and 1.6 mg/g·day, respectively. The control and model groups received equal volumes of distilled water by gavage. After 12 weeks of continuous administration, all mice were fasted for 12 h (water allowed), anesthetized, and blood was collected via orbital bleeding. The blood was allowed to stand for 2 h. Serum was obtained from centrifugation. Liver tissues were excised; a portion was fixed in 4% paraformaldehyde, and another portion was snap-frozen.

Assessment of Blood Lipid Levels, Liver Function Markers, and Antioxidant Indicators: Levels of blood lipid (TC, TG, HDL-C and LDL-C), liver injury indices (ALT, AST) and antioxidant indicators (SOD, MDA and GSH) were measured using an Indiko automatic biochemical analyzer (Randox Laboratories, Crumlin, U.K.), following the instructions provided with the assay kits.

Hematoxylin and Eosin (H&E) Staining: Liver tissues fixed in 4% paraformaldehyde for 24 h were dehydrated through a graded ethanol series, cleared in xylene, embedded in paraffin, and sectioned into 5 μm slices. The sections were deparaffinized in xylene, rehydrated through decreasing concentrations of ethanol and distilled water, and stained with H&E. Finally, the sections were mounted with neutral resin and scanned using an Aperio CS2 slide scanner (Leica, Wetzlar, Germany).

Immunohistochemical Staining: After tissue sections were prepared using the same method as H&E staining, immunohistochemical staining was performed using the BONDIII automated immunohistochemistry and in situ hybridization staining system (Leica, Wetzlar, Germany). Specifically, paraffin sections were first deparaffinized in xylene I (10 min) and xylene II (10 min), followed by hydration through a gradient of ethanol (100% → 95% → 85% → 70%, 5 min each), and two washes in distilled water. Antigen retrieval was conducted by incubating the slides in 0.01 M sodium citrate buffer (pH 6.0) at 95 °C for 20 min. Endogenous peroxidase activity was blocked by incubation with 3% H₂O₂ solution at room temperature for 10 min. The sections were then incubated with primary and secondary antibodies at 37 °C for 30 min each. DAB substrate mixed with H₂O₂ was applied for 2 min for color development, followed by hematoxylin counterstaining for 2 min and subsequent blueing in tap water. The sections were dehydrated through a gradient of ethanol (70% → 95% → 100%), cleared in xylene, and mounted with neutral resin.

qRT-PCR Analysis: Liver tissue was ground in liquid nitrogen. Total RNA was extracted using TRIzol reagent under RNase-free conditions, following the manufacturer’s instructions. The RNA was reverse-transcribed into cDNA using the PrimeScript RT Reagent Kit (Perfect Real Time). Gene expression was evaluated using the TB Green Premix Ex TaqII FAST qPCR Kit in a LightCycler 480II real-time quantitative PCR system (Roche Diagnostics, Mannheim, Germany). Relative mRNA expression of target genes was determined using the 2^{(‑△△CT)} method, with β-actin as the internal control. The primer sequences are detailed in Table .

1. Primer Sequences.

gene	primer sequence
FPN1	downstream primer: 5′-TGTTGTTGTGGCAGGAGAAA-3′
	upstream primer: 5′-AGCTGGTCAATCCTTCTAATGG-3′
TFR1	upstream primer: 5′-AAGTGACGTAGATCCAGAGGG-3′
	downstream primer: 5′-GACAATGGTTCCCCACCAAA-3′
FTH1	upstream primer: 5′-CCATCAACCGCCAGATCAAC-3′
	downstream primer: 5′-GAAACATCATCTCGGTCAAA-3′
β-actin	upstream primer: 5′-ACGCCAACACAGTGCTGTCTG-3′
	downstream primer: 5′-TGCTTGCTGATCCACATCTGCTG-3′

2.4. TBHP-Induced Oxidative Stress Injury Model in HL-7702 Cells and TMP Intervention

Cell Culture: The human normal liver cell line HL-7702 (Catalog number: GNHu6, Cell Bank of the Chinese Academy of Sciences, Shanghai, China) was cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin, at 37 °C in a 5% CO₂ atmosphere.

Simulation of TMP Digestion: TMP was added to artificial saliva (1 U/mL salivary amylase, pH 6.8) at a concentration of 1 g/L and stirred for 5 min at 37 °C. After the reaction, the mixture was heated in a boiling water bath for 5 min to terminate the reaction, followed by centrifugation at 9600g for 5 min to collect the supernatant. The pH of the simulation oral digestion solution was adjusted to 2.0, and pepsin (5 U/mL) was added. After thoroughly mixed, the solution was incubated at 37 °C for 2 h in a constant temperature water bath. Afterward, the pH was adjusted to 7.0. The mixture was boiled for 5 min to terminate the reaction and centrifuged to collect the supernatant. The pH of the simulation gastric digest was adjusted to 6.5, and trypsin was added to a final concentration of 10 mg/mL, along with KH₂PO₄ to 0.1 mol/L. The mixture was incubated in a constant temperature water bath at 37 °C for 2 h. After the pH was adjusted to 7.0, the mixture was boiled for 5 min to terminate the reaction. Upon centrifugation, the supernatant was collected as the TMP simulation digestion solution.

Drug-containing Serum Preparation: Twenty SPF-grade male C57BL/6N mice (16-week-old, 25 ± 2 g) were acclimatized for 7 days under controlled conditions (22 ± 1 °C, 50–60% humidity) and randomly divided into control (distilled water, 10 μL/g) and TMP groups (1.6 mg/g·d TMP suspension, n = 10/group). Daily oral gavage was performed at fixed morning intervals for 7 consecutive days, with body weight monitored every 48 h for dose adjustment. On day 7, mice fasted for 6 h received intraperitoneal anesthesia with 1% pentobarbital sodium (50 mg/kg), followed by retro-orbital venous plexus blood collection using heparinized capillaries 60 min postfinal administration. Whole blood was clotted in sterile separator tubes at 25 °C for 30 min, then centrifuged at 3000g (4 °C, 15 min) to isolate serum supernatant, which was aliquoted into prechilled cryovials and immediately stored at −80 °C.

CCK Assay: HL-7702 cells were seeded in a 96-well culture plate and divided into the blank group, TBHP group, and TMP intervention group, with three replicates for each group. Following cell attachment, the cells were pretreated with TMP simulation digestion liquid for 12 h. Then, the TBHP group and TMP intervention group were treated with 10 μmol/L TBHP for 20 min, followed by two washes with PBS, and continued culturing for 24 h. The blank group received no treatment. Finally, 10 μL of CCK-8 solution was added to each well. The cells were incubated in the incubator for 2 h. The optical density (OD) at 450 nm was measured using a microplate reader.

Flow Cytometry Detection of ROS Levels in Cells: HL-7702 cells were seeded in a 6-well culture plate and divided into the blank group, TBHP group, and TMP simulation digestion liquid intervention group. The treatment method was the same as described in the “CCK Assay” section. After treatment, cells were incubated with 10 μmol/L DCFH-DA for 30 min, washed twice with PBS, digested, and collected. Finally, ROS levels in the cells were detected using a FACSCelesta multicolor flow cytometer (BD, Franklin Lakes, NJ).

Western Blot Analysis: Liver tissue was ground in liquid nitrogen. Proteins were extracted using RIPA lysis buffer containing 1% PMSF and 1% phosphatase inhibitors. Cell disruption was performed on ice using an ultrasonic cell disrupter. The disruption conditions were set to 20W power for continuous 5–10 s, with 10–20 s intervals for 5–10 cycles. Afterward, the samples were centrifuged at 13,900g for 5 min at 4 °C, and the supernatant was collected. The cultured cells were washed with PBS twice. Then, RIPA lysis buffer containing 1% PMSF and 1% phosphatase inhibitors was used for cell lysis, followed by centrifugation to collect the supernatant. The protein concentration was detected using a BCA protein assay kit. Proteins were denatured at 100 °C, separated by SDS-PAGE, and then transferred to a PVDF membrane. The membrane was blocked with 5% nonfat dry milk in TBST buffer for 1 h, followed by overnight incubation with the primary antibody at 4 °C. GAPDH or β-actin was used as the internal control. After five washes with TBST buffer, the membrane was incubated with the secondary antibody at room temperature for 1 h. Following five washes, the membrane was treated with SuperSignal ECL substrate and imaged using a chemiluminescence gel imaging system (Tangen, Beijing, China). Band densities were quantified using ImageJ software.

2.5. Metagenomic Analysis of the of Gut Microbiota

The TMP medium-dose group experimental animals were sacrificed after being fasted for 6 h. They were then thoroughly disinfected with 75% ethanol and fixed on a sterile dissection board. The abdominal cavity was opened along the midline, and the blind colon segments at both ends, 1 cm apart, were ligated. The intestinal segments were gently squeezed with a sterile spatula and the contents were collected into a 2 mL sterile centrifuge tube. The extraction of genomic DNA from gut contents was monitored for integrity and purity using a 1% agarose gel. Concurrently, DNA concentration and purity were evaluated with Qubit 3.0 and Nanodrop One. The quality of the library was assessed through a Qubit 4.0 Fluorescence Quantitation Assay and the Qsep400 High-Throughput Nucleic Acid Protein Analysis System. Ultimately, the library was sequenced on an Illumina NovaSeq 6000 platform, producing paired-end reads of 150 base pairs.The original data were processed using Trimmomatic (version 0.36), yielding Clean Data for subsequent analysis. De novo assembly was conducted using MegaHit software, and statistical analyses were performed on all scaffolds shorter than 500 base pairs. Gene prediction and abundance analysis were carried out using Metagenemark (Version: 3.38), CD-HIT (Version: 4.7), and BBMap software. The Diamond software (available at https://github.com/bbuchfink/diamond/) was utilized to align individual bacterial sequences extracted from the NCBI NR database. Results with an E-value of 1 × 10^–10 were selected for species annotation information determination, employing the LCA algorithm within the MEGAN software system for classification. Based on the LCA annotation results and gene average depths or abundance tables, depth and abundance information for each sample across various taxonomic levels (kingdom, phylum, class, order, family, genus, species) were obtained. Analyses including Krona analysis, relative abundance heatmaps, abundance clustering heatmaps, PCA, and NMS dimensional reduction were all based on the abundance tables at each taxonomic level. Species difference analysis at the genus level was conducted using the Wilcoxon rank-sum test. An ANOSIM analysis was conducted to evaluate intergroup differences, while the LEfSe analysis was employed to identify different species between groups, utilizing the LEfSe software with the default LDA score set to 3 points.Utilizing Diamond software to compare Unigenes within a functional data repository. The functional database chosen for this purpose is the KEGG database (http://www.kegg.jp/kegg/).

2.6. Proteomics Analysis

Liver tissues from the control, model, and treatment groups were collected and quickly ground in liquid nitrogen to preserve sample stability. For proteomics analysis, Proteome Discoverer 2.4 was employed for database searching to identify reliable proteins (Score Sequest HT > 0 and unique peptide ≥ 1). Differential expression analysis was performed based on expression level changes greater than 1.5-fold and a p-value <0.05. GO and KEGG enrichment analyses were conducted on the differentially expressed proteins to explore their biological processes and metabolic pathways.

2.7. Metabolomics Analysis

Metabolomics analysis was carried out on the serum of mice from the control, model, and treatment groups. To 100 μL serum, 20 μL internal standard (l-2-chlorophenylalanine, 0.06 mg/mL in methanol) was added and vortexed for 10 s. Then, 300 μL methanol/acetonitrile (2:1) protein precipitant was added, followed by vortexing for 1 min. The sample was sonicated in ice water for 10 min and left to stand at −20 °C for 30 min. After centrifugation (16,315g, 4 °C, 10 min), 200 μL of the supernatant was dried under nitrogen. The residue was redissolved in 300 μL methanol–water (1:4), vortexed for 30 s, and sonicated for 3 min, followed by standing at −20 °C for 2 h. Upon a second centrifugation (16,315g, 4 °C, 10 min), 150 μL of the supernatant was filtered through a 0.22 μm organic-phase syringe filter and transferred to an LC vial for LC-MS analysis. Quality control (QC) samples were prepared by mixing equal volumes of all samples. An AB ExionLC ultrahigh performance liquid chromatography (UPLC) system coupled with a QE plus high-resolution mass spectrometer was adopted for the analysis. Separation was performed using an ACQUITY UPLC HSS T3 (100 mm × 2.1 mm, 1.8 μm) column. Metabolomics analysis was carried out on the Parson’s Gene Cloud platform (https://www.genescloud.cn).

2.8. Statistical Analysis

All data were statistically analyzed using GraphPad Prism 10 software. The significance of differences between the two groups was assessed using the t test, with p < 0.05 considered statistically significant. Protein expression levels from Western blot experiments were quantified by grayscale analysis using ImageJ software.

3. Results

3.1. Characterization of TMP

HPGPC analysis revealed that TMP primarily consists of two components, with molecular weights of 3713 and 3086 Da, corresponding to relative peak areas of 61.111 and 38.889%, respectively (Figure A). The infrared spectrum showed typical characteristic absorption peaks of polysaccharides. A broad and strong hydroxyl (−OH) stretching vibration peak appeared in the 3600–3200 cm^–1 region. The absorption peak at 2927 cm^–1 corresponded to the asymmetric stretching vibration νas (C–H) of aliphatic C–H bonds. Meanwhile, the peak at 1637 cm^–1 could be attributed to the stretching vibration of bending carbonyl (CO) or olefinic (CC) double bonds, indicating the presence of corresponding functional groups in the sample (Figure B). Ion chromatography confirmed that the major monosaccharide components of TMP were glucose, galactose, mannose, fucose, glucosamine hydrochloride, xylose, and glucuronic acid, with molar mass percentages of 64.5, 13.6, 8.5, 3.7, 2.4, 0.9, and 1.1%, respectively (Figure C), with the contents of 243.73, 47.55, 32.79, 7.09, 3.39, 5.64, and 6.65 μg/mg, respectively (Table S1). SEM observations revealed that the TMP sample exhibited aggregated clusters or plate-like structures with a loose porous surface and attached spherical particles (Figure D–F).

2.

Characterization of TMP. (A) HPGPC chromatogram of TMP. The retention time of TMP is inversely correlated with molecular weight, and the peak area correlates with the relative content of the corresponding molecular fraction. (B) FT-IR spectrum of TMP. The characteristic absorption peaks at 3200–3600, 2927, 1637, and 1000–1200 cm^–1 correspond to hydroxyl (O–H) stretching vibrations, alkyl (C–H) stretching, carbonyl (CO) or olefinic (CC) double bonds, and glycosidic (C–O–C) bond stretching, respectively. (C) Ion chromatography of TMP. The above ion chromatogram illustrates 16 standard monosaccharides, while the one below depicts the composition of TMP monosaccharide. (D–F) Imaging diagrams of TMP using SEM at various magnification levels.

3.2. Effect of TMP on High-Fat Diet-Induced MAFLD Mice

The effect of TMP on high-fat diet-induced MAFLD mice is shown in Figure . Compared to the control group, the model group exhibited significantly elevated serum TC, TG, LDL-C and HDL-C levels (p < 0.05). This trend was evidently reversed by TMP intervention, with blood lipid levels restored to nearly normal ranges (Figure A–D). Liver function indicators showed that serum ALT and AST levels were notably promoted in the model group (p < 0.05), suggesting hepatocellular injury. After TMP treatment, AST and ALT levels were apparently reduced (Figure E,F). Histopathological analysis of liver tissues revealed widespread steatosis and inflammatory cell infiltration in the model group, indicating MAFLD pathology. In contrast, the TMP intervention group showed relatively intact hepatic lobule structure, as well as significantly mitigated steatosis and inflammatory cell infiltration (Figure G,H). These findings demonstrate the hepatoprotective effect of TMP on high-fat diet-induced MAFLD mice.

3.

Effects of TMP on high-fat diet-induced MAFLD mice. (A) The serum total cholesterol (TC) levels. (B) The serum total triglyceride (TG) levels. (C) The low-density lipoprotein cholesterol (LDL-C) levels; (D) The high-density lipoprotein cholesterol (HDL-C) levels. (E) The serum aspartate aminotransferase (AST) levels. (F) The serum alanine aminotransferase (ALT) levels. (G) H&E staining of liver tissue. The hepatocytes in the model group displayed marked signs of ballooning degeneration, characterized by enlarged and irregular cell shapes when compared to the control group. Moreover, the severity of these pathological features was significantly mitigated in the TMP intervention group. (H) Masson staining of liver tissue. (n = 6;*p < 0.05, ** p < 0.01, *** p < 0.001, ns: not significant).

3.3. Effect of TMP on Gut Microbiota Structure in MAFLD Mice

Analysis based on the observed species and Chao1 indices disclosed that the α diversity of the gut microbiota in the TMP intervention group was evidently higher than that in the model group, indicating that TMP effectively enhanced the species richness and evenness of the gut microbiota and maintained its stability (Figure A,B). The Venn diagram showed 115 shared species between the control and model groups, 57 between the model and TMP intervention groups, and 204 between the control and TMP intervention groups, suggesting significant differences in microbial composition between the model group and the other two groups (Figure C). PCoA displayed a marked separation of the microbiota distribution between the control and model groups, indicating significant differences in microbial composition (Figure D). TMP treatment shifted the gut microbiota composition of the model group toward that of the control group, reflecting a regulatory effect of TMP on the microbiota. Hierarchical clustering analysis further confirmed the PCoA results (Figure E). Upon TMP intervention, the gut microbiota structure of the model group resembled that of the control group, suggesting that TMP may positively influence gut homeostasis by modulating microbial diversity. In conclusion, TMP profoundly regulated the disrupted gut microbiota structure in high-fat diet-induced MAFLD mice, promoting a shift toward the microbial community pattern observed in the normal control group.

4.

Effects of TMP on the structure of gut microbiota in MAFLD mice (n = 6). (A, B) Comparison of gut microbiota α diversity using Observed species and Chao1 indices. (** p < 0.01.) (C) Venn diagram depicting the distribution of gut microbiota species among the control, model, and TMP intervention groups. (D) Principal coordinates analysis (PCoA) of gut microbiota structure data among the control, model, and TMP intervention groups. (E) Hierarchical clustering analysis at the OTU level.

3.4. Effect of TMP on the Gut Microbiota Composition of MAFLD Mice

The bar chart of species relative abundance and heatmap analysis indicated that, at the phylum level, the ratio of Firmicutes to Bacteroidetes (F/B) was significantly higher in the model group compared to the control group, while it was notably reduced after TMP intervention (Figure A,B). At the genus level, in the model group, the relative abundance of Bacteroides increased approximately 6.3 times, which may be associated with the enhancement of pro-inflammatory factors and exacerbation of inflammation. The relative abundance of Faecalibaculum surged about 112 times, potentially contributing to the maintenance of intestinal barrier function. However, the relative abundance of Clostridium and Eubacterium decreased, which may be related to the increased F/B ratio and energy metabolism imbalance. Moreover, the relative abundance of Roseburia and Pseudoflavonifractor dropped, possibly affecting short-chain fatty acid (SCFA) production. Compared to the model group, TMP intervention led to a dramatic elevation of approximately 438 times in the relative abundance of Allobaculum, making it the dominant genus. This may be attributed to dietary intervention. The metabolic products of Allobaculum might inhibit the growth of pathogenic bacteria. Meanwhile, the relative abundance of Oscillibacter increased, which could promote bile acid metabolism and energy absorption efficiency. Roseburia and Pseudoflavonifractor were partially restored in the TMP intervention group, suggesting intensified SCFA production and ameliorated intestinal barrier function. The relative abundance of Bacteroides substantially diminished in the TMP intervention group, denoting that TMP may inhibit the excessive proliferation of pro-inflammatory bacteria (Figures C–D and S1, and Table S2). These findings demonstrate that TMP intervention may improve the gut dysbiosis associated with MAFLD by promoting the proliferation of beneficial bacteria, suppressing the growth of potential pathogens, and balancing the expression of pro-inflammatory and anti-inflammatory factors.

5.

Effects of TMP on the composition of gut microbiota in MAFLD mice (n = 6). (A) Relative abundance at the phylum level. (B) Heatmap of the top 10 most abundant phyla. (C) Relative abundance at the genus level. (D) Heatmap of the top 10 most abundant genera. (E) LDA histogram based on LEfSe analysis (LDA > 4 is considered a taxon with different features).

LEfSe analysis ascertained the major bacterial community differences among the groups. The control group was significantly enriched with several SCFA-producing Firmicutes, mainly including the Lachnospiraceae and Oscillospiraceae families. Meanwhile, some Actinobacteria and Bacteroidota communities also exhibited significant enrichment. The model group was specifically enriched with Faecalibaculum from the Erysipelotrichaceae family. The TMP intervention group was mainly enriched with Allobaculum from the Erysipelotrichaceae family and Romboutsia from the Peptostreptococcaceae family (Figure E). These differences suggest notable variations in gut microbiota composition and metabolic potential among the treatment groups, which may be related to their physiological functions and metabolic states.

3.5. Proteomics Analysis of TMP Intervention in MAFLD

A total of 293 differentially expressed proteins were identified between the MAFLD model and control groups, including 227 upregulated and 66 downregulated (Figure A). Between the model and TMP intervention groups, 23 differentially expressed proteins were screened, with 10 upregulated and 13 downregulated (Figure B). The Venn diagram depicted that nine proteins were common between the two groups, with six upregulated in the model group but downregulated in the TMP intervention group. These proteins encompass Cd302, Slc39a14, Slc1a2, Slc16a10, Kdelr3, and Shoc1. Among them, Cd302, Slc39a14, Slc16a10, and Shoc1 can promote the progression of MAFLD (Figure C). These results indicate that TMP intervention partially reverses the abnormal expression of MAFLD-related proteins.

6.

Proteomic analysis of TMP intervention in MAFLD (n = 3). (A) Volcano map of differentially expressed proteins in the model group. (B) Volcano map of differentially expressed proteins in the TMP group. (C) Venn Diagram. (D–G) GO enrichment analysis results: (D) GO enrichment of proteins upregulated in the model group. (E) GO enrichment of proteins downregulated in the model group. (F) GO enrichment of proteins downregulated in the TMP group; (G) GO enrichment of proteins upregulated in the TMP group. The enrichment analysis of the model group focused on the top 30 significantly enriched terms, including the top 10 items selected from biological process (BP), cellular component (CC), and molecular function (MF) categories, respectively. In the TMP group, all significantly enriched items in BP, CC, and MF categories were presented. (H–K) KEGG pathway enrichment analysis: (H) KEGG enrichment of upregulated proteins in the high-fat diet group; (I) KEGG enrichment of downregulated proteins in the high-fat diet group. (J) KEGG enrichment of downregulated proteins in the TMP group; (K) KEGG enrichment of upregulated proteins in the TMP group. For the model group, the top 20 enriched pathways were listed, while all enriched pathways were listed for the TMP group. The x-axis (Enrichment Score) represents the enrichment score, and the y-axis denotes the top 20 pathways. Larger bubbles signify a greater number of differentially expressed proteins. The bubble color transitions from red to green to blue to purple, with smaller enrichment p-values indicating higher levels of significance.

GO enrichment analysis revealed that compared to the control group, upregulated proteins in the model group were primarily enriched in biological processes (BP) related to immune response (innate immune response [GO:0045087] and acute phase response [GO:0006953]), oxidative stress and cell damage (respiratory burst [GO:0045730] and superoxide generating NAD­(P)H oxidase activity [GO:0016175]), and cell membrane functions (platelet aggregation [GO:0070527] and blood coagulation [GO:0007596]). At the cellular component (CC) level, these proteins were mainly located in the endoplasmic reticulum (GO:0005783), extracellular matrix (GO:0005615), and NADPH oxidase complex (GO:0043020). Regarding the molecular function (MF) level, they were enriched in metal ion binding (heme-binding [GO:0020037], calcium-binding [GO:0005509]), antioxidant activity (GO:0016209), and cell function regulation (serine-type endopeptidase inhibition [GO:0004867]) (Figure D). For the downregulated proteins in the model group, BP involved lipid metabolism (fatty acid metabolic process [GO:0006631]) and detoxification and exogenous substance metabolism (exogenous substance metabolism [GO:0006805]). CC focused on mitochondria (GO:0005739), endoplasmic reticulum (GO:0005789), and lipid droplets (GO:0005811). MF was implicated in antioxidant functions (glutathione S-transferase activity [GO:0004364]) and metal ion binding (iron-binding [GO:0005506]) (Figure E).

Compared to the model group, the downregulated proteins in the TMP intervention group were enriched in BP, such as gluconeogenesis (GO:0006094) and protein transport (GO:0015031). In terms of the CC level, these proteins were distributed in integral membrane components (GO:0016021), the Golgi apparatus (GO:0005794), and cell junctions (GO:0030054). MF was associated with metal ion binding (GO:0046872) and protein interactions (homotypic protein interaction [GO:0042802]) (Figure F). The upregulated proteins in the TMP intervention group were primarily enriched in CC terms related to the cytoplasm/soluble fractions (GO:0005737/GO:0005829), the endoplasmic reticulum (GO:0005783), and the nucleus (GO:0005634). At the MF level, these proteins were involved in protein interactions (homotypic protein interaction [GO:0042802] and homodimerization activity [GO:0042803]) (Figure G).

KEGG enrichment analysis disclosed that compared to the control group, the upregulated pathways in the model group were significantly enriched in immune responses and inflammation (complement and coagulation cascades [mmu04610], neutrophil extracellular trap formation [mmu04613], leukocyte transendothelial migration [mmu04670]), ferroptosis (mmu04216), phagocytosis and autophagy (mmu04145), and platelet activation (mmu04611), suggesting immune activation, lipid peroxidation, and cellular clearance dysfunction during MAFLD progression (Figure H). The downregulated pathways in the model group were primarily related to metabolic homeostasis, including drug metabolism (mmu00980/00982/00983), lipid metabolism (mmu00071/00062/00590/00591), antioxidant systems (mmu00480), steroid biosynthesis and hormone metabolism (mmu00100/00140), and ketone body metabolism (mmu00072), indicating comprehensive metabolic dysfunction in MAFLD (Figure I).

In contrast with the model group, the TMP intervention group significantly suppressed pathways involved in metabolism and energy balance, including bile secretion (mmu04976) and carbohydrate metabolism (mmu00030/00051/00010), alleviating lipid synthesis stress. Additionally, the ferroptosis pathway was downregulated (mmu04216), improving lipid peroxidation damage. Furthermore, cellular function pathways were activated and repaired, such as ECM-receptor interaction (mmu04512), protein digestion and absorption (mmu04974), and hormone signaling pathways (insulin [mmu04910], AMPK [mmu04152]), promoting liver tissue repair and metabolic homeostasis restoration (Figure J). The upregulated pathways were mainly enriched in metabolic regulation (galactose metabolism [mmu00052], amino sugar metabolism [mmu00520]) and hormone signaling pathways (estrogen signaling pathway [mmu04915]), implying that TMP ameliorates the pathological state of MAFLD by accelerating the clearance of abnormal metabolic products and regulating hormones (Figure K).

3.6. Metabolomics Analysis of TMP Intervention in MAFLD

The differential metabolites analysis (Figure A) identified 155 differential metabolites between the control and model groups, with 105 upregulated and 50 downregulated (Figure B–C). Through TMP intervention, the metabolite profile was reconstructed, with 80 metabolites showing significant differences (30 upregulated and 50 downregulated), indicating a multitarget regulatory effect of TMP on metabolic disorders associated with MAFLD (Figure D,E).

7.

Metabolomic analysis of TMP intervention in MAFLD (n = 3). (A) Bar chart of differential metabolites. (B) Volcanic maps of differentially expressed metabolites between the model and control groups. (C) Heatmaps of top 50 differential metabolites between the model and control groups. (D) Volcano map of differentially expressed metabolites between the TMP and model groups. (E) Heatmaps of top 50 differential metabolites between the TMP and model groups. (F) Pathway enrichment of differential metabolites between the model and control groups. (G) Pathway enrichment of differential metabolites between the TMP and model groups. The y-axis indicates the pathway names, and the x-axis represents the enrichment factor (Rich factor = number of significantly changed metabolites/total number of metabolites in the pathway). A higher Rich factor denotes greater enrichment. Color from green to red represents decreasing p-values. Larger dots stand for more metabolites enriched in the pathway.

The 105 upregulated metabolites in the model group included lipids (62 species), such as phospholipids (e.g., PC, LysoPC), triglycerides, fatty acids, and sphingolipids; amino acids and their derivatives (11 species); steroids (3 species); hormones and signaling molecules (8 species); phytochemicals (7 species); and other compounds (18 species). The 50 downregulated metabolites encompassed lipids (7 species), steroids and hormones (5 species), organic acids and their derivatives (6 species), alkaloids and drugs (10 species), carbohydrates and their derivatives (3 species), and other compounds (10 species) (Table S3).

The 30 upregulated metabolites in the TMP intervention group consisted of lipid compounds (11 species), such as lyso-phosphatidylethanolamine (LysoPE), phosphatidylinositol (PI), and triethyl citrate; amino acids and their derivatives (3 species, 10%), including d-lysine and 5-amino-1-ribosylimidazole-4-carboxamide; organic acids and their derivatives (2 species, 7%), such as 3-oxoglutaric acid; phytochemicals (5 species, 17%), including β-thujone and withaferin A; drugs and their derivatives (3 species, 10%), such as camptothecin and zopliclone; and other compounds (6 species, 20%), including hydroxy pyridine and vitamin D3 derivatives. The 50 downregulated metabolites included lipid compounds (7 species, 14%), such as PI, phosphatidylcholine (PC), TG, and sphingosine-1-phosphate; amino acids and their derivatives (5 species, 10%), including amino acids, peptides, and acetylcarnitine; steroids and hormones (5 species, 10%), such as cholic acid glucuronide, corticosterone, and retinoic acid; organic acids and their derivatives (6 species, 12%), such as DNA and nicotinamide; alkaloids and drugs (10 species, 20%), including JWH 369, pirimicarb, and creatine; carbohydrates and their derivatives (3 species, 6%); and other compounds (10 species, 20%), including phytochemicals (e.g., leflunomide), antioxidants (pseudouridine), and unidentified substances (Table S3).

Metabolic pathway analysis of the differential metabolites showed that compared to the control group, the upregulated metabolites in the model group were significantly enriched in lipid metabolism (C04230), prostaglandin biosynthesis (C20388), amino acid metabolism (C00398), energy metabolism (C02990), choline metabolism (C00114), and bioactive substance metabolism (C06124), suggesting that these upregulated metabolites may accelerate the progression of MAFLD through multiple pathways, such as intensifying inflammation, altering lipid metabolism, and interfering with energy metabolism (Figure D and Table S4). The downregulated metabolites in the model group were significantly enriched in flavonoid biosynthesis (C16305), terpenoid biosynthesis (C04223), glutathione metabolism (C05570), nicotinic acid metabolism (C01004), prostaglandin biosynthesis and metabolism (C14794), sphingolipid and phospholipid metabolism (C04230), fatty acid metabolism (C02571), and long-chain fatty acid metabolism (C08281), demonstrating remarkable imbalance in the antioxidant capacity, energy metabolism, and lipid balance in the mouse liver (Figure F and Table S4).

The downregulated metabolites in the TMP group were mainly enriched in fatty acid metabolism (C01227), inflammation signaling regulation (C20388), lipid metabolism (C04230), amino acid metabolism (C00431), and energy metabolism (C02571), suggesting that TMP may improve lipid metabolism, inflammation responses, and energy metabolism (Figure E and Table S5). The upregulated metabolites were primarily enriched in hydroxy acid metabolism (C02502), phytochemical metabolism (C09904), and amino acid metabolism (C00739), indicating that TMP may improve MAFLD by regulating iron homeostasis, antioxidant capacity, and amino acid metabolism (Figure G and Table S5).

3.7. TMP Improves Ferroptosis in Hepatocytes

In order to explore the molecular mechanisms by which TMP improves MAFLD, the protective effects of TMP on ferroptosis in hepatocytes were assessed through in vitro experiments. Results for SOD and MDA showed that TMP profoundly promoted the activity of antioxidant enzymes in hepatocytes (Figure A,B). Flow cytometric analysis of ROS and lipid peroxidation presented significantly reduced levels of ROS and lipid peroxidation in the TMP treatment group (Figure C–D). These results suggest that TMP mitigated the oxidative stress response in hepatocytes and effectively inhibited oxidative damage associated with ferroptosis. Western blot analysis disclosed that following TMP treatment, the expression of ACSL4 was downregulated, while the expression of GPx4 was significantly upregulated, implying that TMP may suppress ferroptosis by modulating the expression of antioxidant and iron homeostasis-related proteins (Figure E). qPCR analysis revealed that TMP evidently upregulated the expression of iron regulatory proteins (FPN1 and TFR1) and FTH1, suggesting that TMP may reduce ferroptosis in hepatocytes by regulating the absorption, storage, and efflux mechanisms of iron (Figure F). In conclusion, TMP may exert a protective effect on ferroptosis in hepatocytes by modulating antioxidant enzyme activity, regulating iron homeostasis-related protein expression, and inhibiting ferroptosis biomarkers.

8.

Protective effects of TMP on ferroptosis in hepatocytes. (A) SOD activity assay. (B) MDA content assay. (C) Flow cytometry analysis of ROS levels. MFI: Mean Fluorescence Intensity. (D) Flow cytometry analysis of lipid peroxidation, MFI: Mean Fluorescence Intensity. (E) Western blot analysis of GPx4 and ACSL4 expression in hepatocytes. (F) qPCR analysis of iron regulatory proteins (FPN1, TFR1) and FTH1 mRNA expression. (n = 3, * p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant).

3.8. TMP’s Effects on Ferroptosis in Liver Tissue

In comparison to the model group, the TMP treatment group demonstrated a remarkable decrease in MDA levels. Concurrently, the activities of SOD and the levels of GSH and GPx4 in the liver were significantly elevated. These findings strongly suggest that TMP effectively mitigated oxidative stress (Figure A–D). Western blot analysis revealed that in the TMP group, the expression of ACSL4 was downregulated and the expression of GPx4 was significantly up-regulated (Figure E). Immunohistochemical analysis of ferroptosis markers showed that the expression of GPx4, ACSL4, FTH1 and Nrf2 was significantly elevated, further validating the role of TMP in inhibiting ferroptosis (Figure F). Prussian blue staining was used to measure liver tissue iron content. The results showed that compared to the model group, the TMP treatment group had markedly declined iron deposition, suggesting that TMP may suppress ferroptosis by modulating iron homeostasis (Figure G). In summary, TMP significantly impacts liver tissue ferroptosis by attenuating oxidative damage and regulating the expression of iron-related biomarkers, potentially offering a novel mechanistic insight into liver protection.

9.

Effects of TMP on ferroptosis in liver tissue. (A) The levels of MDA in liver homogenate. (B) The SOD activity in liver homogenate. (C) The levels of GSH in liver homogenate. (D) The levels of GPx4 in liver homogenate. (E) Western blot analysis of GPx4 and ACSL4 expression in liver tissue. (F) The immunohistochemical staining and quantified results of ACSL4, FTH1, GPx4, and Nrf2 in liver sections from the control, model, and TMP treatment groups are presented. The localization and expression levels of the proteins are visualized through brown staining. (G) Prussian blue staining in liver sections from the control, model, and TMP treatment groups. The positive staining appears as blue precipitates, indicating the presence and distribution of iron. (n = 3, *p < 0.05, **p < 0.01, ***c 0.001, ns: not significant).

3.9. The Impact of TMP-Containing Serum on Ferroptosis in Hepatocytes

To explore the effect of TMP metabolites on ferroptosis, we evaluated the effect of MTP-containing serum on ferroptosis in hepatocytes through in vitro cell experiments.The experimental results of SOD, MDA activity and GPx4 content detection showed that the MTP-containing serum significantly increased the activity of antioxidant enzymes in hepatocytes (Figure A–C). Flow cytometry analysis revealed a significant reduction in both reactive oxygen species (ROS) levels and lipid peroxidation in the TMP drug-containing serum treatment group compared to the control group (Figure D,E). All the results collectively demonstrated that the metabolites of TMP alleviated the oxidative stress response in liver cells and effectively suppressed oxidative damage associated with ferroptosis.

10.

Impact of TMP-containing serum on ferroptosis in hepatocytes. (A) The levels of MDA in serum; (B) The SOD activity in serum; (C) The levels of GPx4 in serum; (D) Flow cytometry analysis of ROS levels.MFI: Mean Fluorescence Intensity. (E) Flow cytometry analysis of lipid peroxidation. MFI: Mean Fluorescence Intensity. CS: Control serum, DCS: Drug containing serum. (n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).

3.10. Correlation Analysis

This study investigated the relationship between TMP and gut microbiota, metabolites, and ferroptosis in MAFLD. The gut bacterial genera with a relative abundance of ≥1%, the top 50 significantly different metabolites, and MAFLD-related indicators were selected, and the correlation between gut microbiota and metabolites, as well as oxidative stress indicators associated with ferroptosis, was examined. The results indicated that the Clostridium genus exhibited a significant negative correlation with specific lipid and cholesterol metabolism products in the model group, while Faecalibaculum showed a positive correlation (Figure A). In contrast to the model group, Allobaculum and Phocaeicola contributed to lipid and metabolic regulation in the TMP group (Figure B). Analysis of gut microbiota and MAFLD indicators revealed a negative correlation between Clostridium and liver function indicators AST, as well as the oxidative stress indicator MDA (Figure C), suggesting potential benefits for liver function and antioxidation. In the TMP treatment group, Allobaculum demonstrated a positive correlation with the antioxidant enzyme GPx4, indicating an enhancement of antioxidant defense; Oscillibacter exhibited a negative correlation with MDA (Figure D), suggesting its potential to reduce oxidative stress levels.

11.

Correlation analysis. (A) The correlation between gut microbiota and metabolites in the model group. (B) The correlation between the intestinal flora and metabolites in the TMP group. (C) The correlation between the intestinal flora and the related indicators of oxidative stress ferroptosis in the model group. (D) The correlation between the intestinal microbiota in the TMP group and the related indicators of oxidative stress ferroptosis. (n = 6, *p < 0.05, ** p < 0.01, *** p < 0.001, ns: not significant).

4. Discussion

In the field of treatment for MAFLD, although various drugs and intervention measures have been explored, finding safer, more effective and comprehensive treatment methods remains the current focus and challenge of research. This study focuses on the intervention effect of TMP on MAFLD, deeply explores its multifaceted mechanisms for improving MAFLD, and systematically clarifies the synergistic relationship between TMP and the “gut-liver“ axis as well as the cellular and molecular regulatory network.

In the context of liver diseases, intestinal dysbacteriosis is implicated in the onset of various pathological alterations, including bile acid metabolism disorder, endotoxemia, and hepatic inflammation. TMP potentially ameliorates these pathological disturbances by modulating the gut microbiota, thereby conferring hepatoprotective effects. Prior research has established that probiotic modulation of the gut microbiota can enhance liver function and lipid metabolism in MAFLD patients, a mechanism akin to the action of TMP observed in this study. Additionally, an imbalance in the F/B ratio has been closely linked to obesity and metabolic syndrome, while the enrichment of SCFA-producing bacteria can bolster intestinal barrier integrity and attenuate endotoxemia, thus alleviating liver inflammation. In this study, TMP significantly altered the gut microbiota profile, leading to a reduction in the F/B ratio and enrichment of SCFA-producing taxa such as Roseburia, a shift that may represent a critical component of TMP’s therapeutic effect in MAFLD. Existing literature supports the notion that SCFA-producing bacteria, such as Roseburia, produce butyric acid and other SCFAs, which modulate the host’s energy metabolism and inhibit excessive lipid accumulation, thus exerting protective effects in the progression of MAFLD. The findings in this study demonstrated that TMP treatment markedly reduced hepatic steatosis in mice, with a significant correlation to the increased relative abundance of SCFA-producing bacteria like Roseburia, underscoring the role of enriched SCFA-producing bacteria in TMP’s hepatoprotective action.

Furthermore, previous studies have highlighted the adverse role of pro-inflammatory bacteria, such as Bacteroides, in the pathogenesis of MAFLD. These microbiota may exacerbate hepatic inflammation by producing inflammatory mediators, such as lipopolysaccharides. In line with this, this study found that TMP effectively suppressed the overproliferation of Bacteroides, contributing to hepatic inflammation mitigation. This observation reinforces the concept that TMP improves MAFLD through modulation of the gut microbiota.

Moreover, this study found that TMP intervention evidently reversed MAFLD-associated changes in the proteome and lipidome and counteracted lipid overload induced by MAFLD, particularly the accumulation of phospholipids and triglycerides. Meanwhile, TMP inhibited hepatocellular ferroptosis by upregulating GPx4, downregulating ACSL4, and modulating iron homeostasis-related proteins such as FPN1 and TFR1. It has been reported that certain natural compounds (e.g., palmitoleic acid, berberine, and ginkgolide B) improve MAFLD by regulating the expression of antioxidant enzymes (such as GPx4) and lipid metabolism-related enzymes (such as ACSL4). − These findings align with the mechanisms of TMP observed in this study, suggesting that TMP may exert hepatoprotective effects via similar pathways.

In terms of lipidomics, previous research has indicated that phospholipid and triglyceride accumulation is a hallmark of MAFLD. , Certain bioactive substances (e.g., highland barley, L-carnitine, and traditional Chinese medicine formulas) or natural products have been shown to improve MAFLD by regulating lipid metabolic pathways and alleviating phospholipid and triglyceride accumulation. − This study expands the understanding in this field by revealing the distinctive role of TMP in modulating lipid metabolism. It not only validates the hypothesis that TMP ameliorates MAFLD via targeting lipid metabolism pathways but also provides a valuable theoretical and practical reference for developing TMP-based therapeutic strategies for MAFLD.

In liver disease intervention, many dietary supplements such as probiotics, antioxidants, and dietary fiber have demonstrated hepatoprotective effects. , However, their mechanisms are relatively limited and usually target only a single pathological aspect of MAFLD. In contrast, TMP presents unique advantages in MAFLD intervention. Probiotics primarily function by modulating the gut microbiota and reducing endotoxemia to alleviate liver inflammation. , However, the effects of probiotics are mostly confined to the gut, with limited direct protective effects on the liver. , Proteomics results revealed that TMP downregulated Cd302 expression in hepatic cells induced by a high-fat diet and inhibited the transformation of macrophages into a pro-inflammatory phenotype, a process potentially linked to gut microbiota alterations. These findings suggest that TMP suppresses intrahepatic inflammation via the gut-liver axis.

Antioxidants in MAFLD treatment primarily protect hepatocytes by reducing oxidative stress. − Compared with conventional antioxidants, TMP shows a more precise targeting capability. By upregulating antioxidant-related proteins such as Nrf2 and regulating iron homeostasis proteins, TMP specifically inhibits ferroptosis and oxidative stress, thereby exerting hepatoprotective effects. In addition, dietary fiber contributes to MAFLD treatment mainly by enhancing gut microbiota, lowering endotoxemia, and mitigating liver inflammation. − However, its effects are generally slow and limited. In comparison, TMP not only regulates gut microbiota structure but also modulates hepatic proteomic and lipidomic changes and inhibits ferroptosis, thereby providing rapid and comprehensive liver protection through multiple mechanisms.

Ferroptosis is an iron-dependent form of programmed cell death driven by lipid peroxidation. It is closely associated with iron metabolism and redox balance. High-fat diets have been proven to induce iron overload and lipid peroxidation in hepatocytes, thereby activating the ferroptosis pathway. This directly contributes to hepatocyte death and fosters an inflammatory microenvironment. Ferroptosis not only causes direct liver damage but also triggers inflammatory responses and promotes fibrosis progression by releasing intracellular components that attract immune cells, thus exacerbating hepatic injury. , Various factors, such as oxidative stress and lipid peroxidation, have been reported to trigger ferroptosis in hepatocytes. , Conversely, some natural bioactive compounds, such as gallic acid, can inhibit these triggers and decline the occurrence of ferroptosis. Notably, ellagic acid and ginkgolide B can advance liver function and lipid metabolism in MAFLD by modulating the Nrf2 signaling pathway and iron homeostasis-related proteins. ,

In this study, TMP demonstrated potent antioxidant activity by upregulating GPx4 expression, thereby hindering ROS accumulation and lipid peroxidation in hepatocytes. Simultaneously, TMP downregulated ACSL4 expression, suppressing the generation of PUFA-CoA, a direct precursor of lipid peroxidation. Furthermore, TMP attenuated the expression of TFR1, thereby limiting the uptake of exogenous iron by cells. These findings suggest that TMP ameliorates MAFLD by inhibiting high-fat diet-induced ferroptosis in hepatocytes.

This study elucidates that TMP exerts “bidirectional gut–liver axis regulation” by modulating the gut microbiota-metabolite-ferroptosis Axis, consequently providing a therapeutic effect on metabolism-associated fatty liver disease (MAFLD). The research presents a novel intervention approach that harnesses natural polysaccharides and identifies potential molecular targets for the prevention and treatment of MAFLD, demonstrating significant clinical application prospects.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.5c05877.

• TMP monosaccharide composition; relative abundance and differential abundance analysis of genus level species; differential metabolites and enrichment analysis of metabolic pathways (PDF)

⊥.

C.M., Y.B., and S.H. contributed equally to this work. C.M., Y.B., S.H.: Original draft, funding acquisition, methodology. H.Z., X.B., Q.B., L.Z., X.Z.: Data curation, formal analysis, methodology. L.B., X.B.: Funding acquisition, writingreview and editing.

The authors declare no competing financial interest.