Integrated serum metabolomics reveal molecular mechanism of Xietu Hemu prescription on metabolic dysfunction-associated steatotic liver disease-related obesity

Abstract

BACKGROUND

Xietu Hemu prescription (XHP), a Chinese patent formula, is optimized based on the theory of “phlegm-dampness” and has been clinically validated to effectively combat metabolic dysfunction-associated steatotic liver disease (MASLD). It notably reduces visceral fat and body mass index. However, the molecular mechanisms underlying its regulation of lipid metabolism homeostasis remain unexplored.

AIM

To elucidate the mechanisms by which XHP inhibits adipocyte differentiation and maintains lipid metabolism homeostasis.

METHODS

The therapeutic efficacy of XHP in metabolic-related disorders was analyzed using HepG2 cells and 3T3-L1 cells, along with transcriptomics to assess gene expression alterations during white adipogenesis. The primary metabolites of XHP were identified through ultra-performance liquid chromatography, and metabolic pathways were examined via serum metabolomics. Network analysis was employed to predict therapeutic targets. The accumulation of lipid droplets and the expression of associated proteins were confirmed using oil red O staining and Western blotting, respectively. Molecular docking was utilized to identify core targets and signaling pathways, which were substantiated through immunofluorescence and siRNA interference.

RESULTS

XHP-containing serum (XHPS) significantly inhibited the transformation of normal HepG2 cells into fatty liver cells. Concurrently, the treatment suppressed the differentiation of 3T3-L1 cells, reduced lipid droplet accumulation and total cholesterol/triglyceride levels, and downregulated the expression of PPARγ, C/EBPα, and FABP4. Through transcriptomics and network pharmacological intersectionality analyses, 24 core targets were identified, predominantly enriched in the AMPK signaling pathway. Molecular docking validated the strong binding affinity of XHP metabolites to targets such as leptin (-11.3 kcal/mol) and ADIPOQ (-9.4 kcal/mol). ELISA results indicated that XHPS augmented leptin autocrine secretion, thereby activating the AMPK signaling pathway (P < 0.05). Conversely, LEPR knockdown negated this effect (P < 0.05).

CONCLUSION

XHP effectively inhibits adipogenesis and enhances lipid metabolism homeostasis through the LEP/AMPK/PPARγ pathway, presenting a promising multi-target therapeutic strategy for MASLD by mitigating lipotoxicity.

Key Words: Xietu Hemu prescription; Metabolic dysfunction-associated steatotic liver disease; Obesity; Lipotoxicity; Leptin

Core Tip: Xietu Hemu prescription, a traditional Chinese medicine formula, ameliorates metabolic dysfunction-associated steatotic liver disease-related obesity by inhibiting adipogenesis. Integrating serum metabolomics, transcriptomics, and network analysis, this study identified the LEP/AMPK/PPARγ signaling axis as the central mechanism. The formula promotes leptin autocrine secretion, activates AMPK phosphorylation, and subsequently downregulates PPARγ and SREBPs, thereby reducing lipid accumulation and improving metabolic homeostasis, offering a novel multi-target therapeutic strategy.

INTRODUCTION

The development of metabolic dysfunction-associated steatotic liver disease (MASLD) is fundamentally linked to disruptions in lipid metabolism. When the body is unable to store surplus lipids, they are released into the bloodstream and subsequently accumulate in hepatocytes. This accumulation can lead to lipotoxicity, which damages hepatocytes and triggers the activation of hepatic stellate cells, resulting in hepatic fibrosis[1]. Addressing obesity to mitigate lipotoxicity is recognized as one of the most effective strategies for managing MASLD. Obesity, a chronic global metabolic disorder, has escalated to epidemic proportions over recent decades, posing a significant public health challenge. According to a 2021 report by the World Health Organization, the global population of obese individuals has surpassed 650 million, with figures continuing to rise. In recent decades, shifts in lifestyle and dietary habits have contributed to a sharp increase in obesity rates across both developing and developed nations, imposing a substantial economic burden on healthcare systems[2].

Obesity and MASLD exhibit shared pathophysiological mechanisms, including imbalances in energy metabolism, insulin resistance, chronic inflammation, and dysregulation of lipid metabolism[3-5]. Research indicates that between 70% and 90% of obese individuals develop fatty liver disease, compared to 20%-30% in the general population, with the severity of the disease intensifying as obesity progresses[6,7]. This high rate of comorbidity underscores a complex pathophysiological connection between obesity and MASLD. Weight loss remains the most effective therapeutic intervention, with a reduction of 5%-10% in body weight significantly decreasing hepatic fat accumulation[8]. Therefore, managing obesity is a primary approach in the treatment of MASLD.

The Xietu Hemu prescription (XHP), a patented herbal formula (Chinese patent No. ZL 2022 1 0250245.4, authorized in 2023), was developed by the Department of Endocrinology at Jiangsu Provincial Hospital of Chinese Medicine to address obesity and MASLD. Early clinical research on XHP has shown its potential in reducing the fatty liver index, MASLD fibrosis score, body mass index, and visceral fat area[9]. The composition of XHP includes 14 botanical drugs, with Codonopsis Radix, Hedysarum Multijugum Maxim, and Atractylodes Macrocephala Koidz as the principal metabolites. The additional 11 botanical drugs are Bupleuri Radix, Ligustri Lucidi Fructus, Citrus Reticulata, Citri Reticulatae Pericarpium Viride, Arum Ternatum Thunb, Scutellariae Radix, Folium Nelumbinis, Alismatis Rhizoma, Sedi Herba, Crataegi Folium, and Fructus Liquidambaris.

Currently, the United States Food and Drug Administration (FDA) has not approved any pharmacological agents that are highly effective for treating obesity. Most Western medications that are clinically used for obesity demonstrate limited efficacy or present unacceptable side effects. In contrast, prescriptions from traditional Chinese medicine (TCM), which adopt a multi-target, cocktail-like therapeutic approach, have attracted increasing interest. According to TCM theory, the primary pathogenesis of obesity is attributed to the accumulation of “phlegm and dampness”. XHP was developed specifically to target this underlying syndrome. It combines the spleen-fortifying and dampness-resolving properties of Shenling Baizhu San, as originally documented in the Tai Ping Hui Min He Ji Ju Fang during the Song dynasty, with the liver-soothing and qi-regulating effects of Chaihu Shugan San from the Jing Yue Quan Shu of the Ming dynasty. XHP is tailored to address the combined pathological pattern of “phlegm-dampness accumulation and liver stagnation with spleen deficiency”, which typifies the TCM syndrome commonly observed in obese patients with MASLD. The therapeutic reliability and syndrome-specific efficacy of XHP have been corroborated through prolonged clinical use. Given that obesity serves both as a precursor and an overt expression of MASLD, elucidating the molecular mechanisms and therapeutic targets of XHP in the treatment of obesity is critically important for understanding MASLD.

MATERIALS AND METHODS

Reagents

The following primary antibodies were acquired: Anti-C/EBPα (8178), anti-FABP4 (3544), anti-PPARγ (2443), anti-mTOR (2983), and anti-p-mTOR (5536) from CST (United States); anti-GAPDH (10494-1-AP) from Proteintech (Wuhan, China); anti-AMPK (ab32047) from Abcam (United States); and anti-SREBP1/2 (sc-365513/sc-13552; for cytoplasmic protein detection) from Santa Cruz (United States). Other primary antibodies included anti-p-AMPK (AP0116), anti-leptin (A1300), anti-LEPR (A19076), anti-Lamin B1 (A11495), and anti-SREBP1/2 (A15586/A13049; for nuclear protein detection) obtained from Abclonal (Wuhan, China). Secondary antibodies (AS003/AS014) were sourced from Abclonal (Wuhan, China). The ECL developer (BL520) was supplied by Biosharp (Hefei, China), and the protease inhibitor (P001) was from New Cell & Molecular Biotech Co., Ltd (Suzhou, China). The specific reagents for 3T3-L1 preadipocytes, including medium (ZM0089) and serum-free cell freezing solution (CSP042), were procured from Zhongqiao Xinzhou Biotechnology Co., Ltd (Shanghai, China). Bovine insulin (I8040) and Dexamethasone (D4902) were available from Solarbio (Beijing, China) and Sigma (Germany), respectively. 3-Isobutyl-1-methylxanthine (I5879) and oil red O powder (O0625) were also supplied by Sigma (Germany). Additional reagents included fetal bovine serum (FBS), trypsin, and high glucose DMEM from Gibco (United States); CCK8 reagent (HY-K0301) and GW9662 (HY-16578) from MCE (United States); and Silibinin (CAS No. 22888-70-6, B21185) from Yuanye Bio-Technology Co., Ltd (Shanghai, China). Palmitic acid (PA) and oleic acid (OA) were procured from Kunchuang Technology Development Co., Ltd (Xi’an, China), along with other consumables and reagents from Servicebio (Wuhan, China).

XHP preparation

The 14 botanical drugs comprising XHP were sourced from the Department of Traditional Chinese Medicine at Jiangsu Provincial Hospital of Chinese Medicine. The quantities and respective percentages of these botanical drugs used in XHP are detailed in Table 1. These drugs were combined in precise proportions in a ceramic pot, submerged in an adequate volume of water, and subjected to a sequential process of softening for 0.5 hours, followed by heating and decoction for 1 hour. This decoction process was repeated for an additional hour with replenished water. Subsequently, the mixture was filtered to remove solid residues and the excess liquid was evaporated using a rotary evaporator to achieve the desired concentration of XHP for administration to rats. The prepared XHP was then stored at -80 °C, and warmed to 37 °C for use after thawing.

Table 1 The composition and dosage of Xietu Hemu prescription.

TCM name	Latin names	Plant part	Occupied percent (%)	Dosage for adults (g/day/60 kg)	Dosage for rat (g/day/kg)
Dangshen	Codonopsis Radix	Root	7.246	10	1.050
Huangqi	Hedysarum Multijugum Maxim	Root	7.246	10	1.050
Baizhu	Atractylodes Macrocephala Koidz	Stolon	7.246	10	1.050
Chaihu	Bupleuri Radix	Root	4.348	6	0.630
Nüzhenzi	Ligustri Lucidi Fructus	Fruit	10.870	15	1.575
Chenpi	Citrus Reticulata	Pericarp	4.348	6	0.630
Qingpi	Citri Reticulatae Pericarpium Viride	Pericarp	7.246	10	1.050
Banxia	Arum Ternatum Thunb	Tuber	7.246	10	1.050
Huangqin	Scutellariae Radix	Root	4.348	6	0.630
Heye	Folium Nelumbinis	Leaf	7.246	10	1.050
Zexie	Alismatis Rhizoma	Tuber	7.246	10	1.050
Chuipencao	Sedi Herba	All	10.870	15	1.575
Shanzha	Crataegi Folium	Fruit	7.246	10	1.050
Lulutong	Fructus Lipuidambaris	Fruit	7.246	10	1.050
Total	-	-	100.000	138	14.490

TCM: Traditional Chinese medicine.

Drug serum preparation

Male Sprague-Dawley rats, each weighing approximately 200 g (± 10 g), were procured from Jiangsu Qinglongshan Biotechnology Co., Ltd., for the purpose of drug serum preparation. Following the clinical guidelines provided by the Department of Endocrinology at Jiangsu Provincial Hospital of Chinese Medicine, an adult weighing 60 kg requires a daily intake of 138 g of XHP, equivalent to 2.3 g/kg. Accordingly, the dosage for rats was set at 14.49 g/kg, approximately 6.3 times the clinical dose[10]. The rats were administered 10 mL/kg/day of either drinking water or XHP for seven days to harvest the serum, subsequently designated as vehicle serum (VehS) or XHP-containing serum (XHPS). This serum was then inactivated, filtered, and stored in aliquots for further use in cellular experiments.

Ultra-performance liquid chromatography with quadrupole time-of-flight tandem mass spectrometry & serum metabolites identification of XHP

Ultra-performance liquid chromatography (UPLC)-quadrupole time-of-flight tandem mass spectrometry (QTOF/MS/MS) represents a sophisticated analytical method that combines UPLC with QTOF/MS/MS. This technology is primarily utilized for the separation, identification, and structural elucidation of complex biochemical samples.

Sample preparation

Herbal prescription extraction: A small aliquot of the concentrated XHP was desiccated at 45 °C. Following this, 0.1 g of the solid extract was solubilized in 10 mL of 70% methanol and subjected to ultrasonic extraction for 30 minutes, a process that was repeated twice. The resultant supernatants were amalgamated, and 1 mL of this solution was then evaporated to dryness under a stream of nitrogen gas. The residual matter was reconstituted in 1 mL of ultrapure water and subsequently purified using a C18 solid-phase extraction cartridge. This cartridge had been preconditioned with 2 mL of 50% methanol, then loaded with 1 mL of the sample, and rinsed with 1 mL of ultrapure water. Elution was carried out using 1.5-2 mL of methanol. The eluate was centrifuged at 13000 rpm for 15 minutes, and 400 μL of the clear supernatant was again evaporated to dryness under nitrogen.

Serum sample extraction: Pooled serum samples were prepared by combining 30 μL aliquots from each subject in both the XHP-treated group (n = 3, labeled XHPS) and the control group (n = 3, labeled VehS). To precipitate proteins, 0.9 mL of ice-cold methanol was added to each pooled sample, followed by vigorous vortexing and sonication in an ice bath for 10 minutes. The samples were then incubated at 4 °C for 60 minutes and centrifuged at 13000 rpm for 15 minutes. A 400 μL aliquot of the supernatant from each sample was evaporated to dryness under nitrogen. The dried residues were reconstituted in 300 μL of 70% methanol, centrifuged again at 13000 rpm for 15 minutes, and the refined supernatant was subsequently injected into the liquid chromatography quadrupole time-of-flight mass spectrometry system.

Analytical conditions

Chromatographic separation was effected on a Waters Acquity HSS T3 column (2.1 mm × 150 mm, 1.8 μm) maintained at 35 °C. The mobile phase comprised water with 0.1% formic acid and acetonitrile, delivered at a flow rate of 0.3 mL/minute. The volume of each injection was 2 μL, with the elution gradient specified in Supplementary Table 1. Mass spectrometric analysis was executed using a SCIEX Exion LC system coupled to an X500B Q-TOF mass spectrometer (AB Sciex, Foster City, CA, United States) equipped with an electrospray ionization source. Data acquisition occurred in both positive and negative ion modes across a scan range of m/z 100-1250. Detailed parameters for these analyses are provided in Supplementary Table 2.

Cell culture and induction of differentiation

HepG2 cells are frequently used to establish lipid metabolism-related fatty liver diseases, including MASLD and MASH, while 3T3-L1 cells are commonly employed for studies on adipogenesis and obesity. The objectives and methods of the experiment on HepG2 cells and 3T3-L1 cells referred to published researches[11,12].

HepG2 cells, generously provided by Servicebio, were cultured to approximately 80% confluence. Subsequently, these cells underwent a starvation period of 12 hours in DMEM before being subjected to a 24-hour modeling phase with PA (0.5 mmol/L) or PA + OA (PA: 0.25 mmol/L, OA: 0.5 mmol/L) in 5% VehS. Following this, cells were treated with varying concentrations of XHPS or VehS for 24 hours, after which lipid accumulation was assessed using oil red O staining.

The 3T3-L1 preadipocytes, provided by Professor Guang Chen’s laboratory at Tongji Medical College, Huazhong University of Science and Technology, were used within ten passages for all differentiation experiments. Initially, these preadipocytes were maintained in a growth medium supplemented with newborn calf serum. To promote dense cell-to-cell contact, cells were seeded in culture vessels until confluent, followed by an additional two-day culture period. This stage was designated as day 0, marking the transition to a quiescent state. The differentiation induction phase commenced thereafter, during which the medium was replaced with drug serum (VehS or XHPS) derived from rats, supplemented with various inducers or differentiation agents, and continued for 6-10 days, or longer, depending on the experimental requirements.

Cell viability assay

For viability assays, cells were plated in 96-well plates and exposed to different ratios of drug serum. After 72 hours, cell viability was evaluated using the Cell Counting Kit-8 assay, following the manufacturer’s protocol. Absorbance was measured at 450 nm using a PerkinElmer EnSpire microplate reader.

Oil red O staining

Following differentiation induction, 3T3-L1 preadipocytes cultured in 24-well plates were prepared for staining. The medium was carefully aspirated, and cells were washed three times with phosphate-buffered saline (PBS). Cells were then fixed in 10% neutral buffered formalin for 60 minutes. The oil red O working solution was prepared by diluting solution O0625 with deionized water in a 3:2 ratio, followed by filtration and a 10-minute room temperature incubation to eliminate impurities. Approximately 1.5 mL of this working solution was added to each well of a 6-well plate, and the cells were stained for 30-60 minutes. Post-staining, the solution was discarded, and residual dye was removed by a brief rinse with 60% isopropanol, lasting no longer than 10 seconds per well. Cells were washed 2-3 times with PBS and counterstained with hematoxylin for one minute. After a final wash with PBS, the stained cells were visualized and documented using an Olympus CKX41 inverted microscope.

Transcriptomics data analysis

The study utilized 3T3-L1 cells across three distinct passages, each comprising control and model groups. By the sixth day of the modeling process, mature lipid droplets were prominently visible under microscopic examination. The cells were initially rinsed gently with PBS, followed by the addition of Trizol reagent to lyse the cells. The lysates were immediately harvested and preserved at -80 °C. For the assessment of sample quality, polyadenylated mRNA was enriched using magnetic beads coated with oligo (dT). The isolated RNA was fragmented in a fragmentation buffer, and first-strand cDNA synthesis was carried out using random N6 primers. This step was followed by the synthesis of double-stranded DNA, ligation of sticky ends, and adapters to the 3’ ends, along with phosphorylation of the 5’ ends. Polymerase chain reaction (PCR) amplification was then performed, and the PCR products were denatured into single-stranded DNA by heating. Circularization of the single-stranded DNA was achieved using a bridging primer, resulting in the formation of a single-stranded circular DNA library. The libraries underwent stringent quality control measures, and only those that met the quality criteria were selected for sequencing. The analysis of gene expression included evaluating the stability of gene expression and identifying differentially expressed genes (DEGs). Functional enrichment analyses, such as those involving Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, were conducted to elucidate potential mechanisms inferred from the data.

Immunoblotting and analysis

For assays targeting nuclear proteins, samples were prepared utilizing the Nuclear and Cytoplasmic Protein Extraction Kit from KeyGEN (KGB5302), adhering strictly to the manufacturer’s protocol. For assays not concerning nucleoproteins, the following procedures were employed: Protein lysis buffer was fortified with a 1/100 volume of both protease and phosphatase inhibitors to prepare the protein lysis mixture. Following a PBS wash, 100 μL of the lysis mixture was applied to each well to ensure coverage of the cell monolayer. Cells were then dislodged and lysed using a cell sonicator, and the mixture was agitated gently at low temperature for 30 minutes. Post-centrifugation at low temperature, the supernatant was collected, and a quarter volume of protein loading buffer was added. The samples were then heated in boiling water for 10 minutes and were either immediately subjected to electrophoresis or stored. Protein concentration was quantified using a BCA assay kit (Servicebio, G2006) with a microplate reader (PerkinElmer, EnSpire). A standard load of 30 μg of protein per lane was applied to an SDS-PAGE gel. Following electrophoresis, proteins were transferred to a PVDF membrane. The membrane was blocked and subsequently incubated overnight at 4 °C with primary antibodies. The following day, after washing with Tris buffered saline Tween, the membrane was incubated with secondary antibodies at room temperature for 1 hour. Post-washing, the membrane underwent chemiluminescent detection using General Electric’s ImageQuant LAS 4000, and the acquired images were analyzed with ImageJ software.

ELISA assay

The Mouse Leptin ELISA Kit (RK00380) was procured from Abclonal (Wuhan, China). The assay was conducted in accordance with the manufacturer’s protocol, utilizing duplicate wells for each sample. Results were quantified using the established standard curve.

Cellular immunofluorescence

Cells were seeded onto plates with pre-positioned coverslips and cultured until the 3T3-L1 cells adhered. Following treatment with the drug, cells were rinsed with ice-cold PBS and fixed with 4% paraformaldehyde for 15 minutes. Subsequent washes with PBS were performed before cells were permeabilized with 0.3% Triton X-100 for 10 minutes and washed again. Non-specific binding was minimized by incubating the cells with 2% bovine serum albumin at room temperature for one hour. After the removal of the blocking solution, cells were incubated with primary antibodies overnight at 4 °C. The next day, the cells were washed with PBST and incubated with fluorophore-conjugated secondary antibodies or DAPI staining solution for one hour at room temperature in the dark. After additional washes with PBST, the coverslips were mounted on glass slides using an anti-fade mounting medium. Fluorescent images were obtained using an Olympus CKX41 fluorescence microscope.

Construction and transfection of siRNA targeting LEPR

To inhibit LEPR expression in 3T3-L1 cells, siRNA targeting LEPR was utilized, prepared by Nanjing Tsingke Biotechnology Co., Ltd. The sequence number is listed in Supplementary Table 3. Cells were seeded in 6-well plates and left for 24 hours until the cell density reached 70%-80%. For transfection, 125 μL of serum-free Opti-MEM and 100 pmol of siRNA or vehicle were added to a sterile centrifuge tube, while another tube received 125 μL of Opti-MEM and 6 μL of Lipo3.0 transfection reagent (Abbkine, BMU111-CN). Both mixtures were combined and incubated for 15 minutes to form the transfection reagent/siRNA complex. The cells were then cultured in this 250 μL mixture supplemented with 1.5 mL of complete medium for 24 hours. Post-transfection, the efficacy of the siRNA was assessed, and differentiation experiments of 3T3-L1 cells were initiated.

Network analysis

Data collection: Target information for each metabolite in XHPS was sourced from the TCM systems pharmacology[13] database and the PubChem metabolite database[14]. Additional targets were predicted using PharmMapper (https://www.lilab-ecust.cn/pharmmapper/index.html), and this information was integrated. The clinical drug database DrugBank[15] was queried using the keyword “obesity” to retrieve entries related to a body mass index greater than 30 kg/m², including FDA-approved drugs for obesity and their corresponding targets. Disease-related targets were obtained from the disease gene databases DisGeNET[16] and GeneCards[17].

Data transformation: To eliminate invalid data and ensure the accuracy of cross-species gene ID conversions, the String database[18] was employed. Upon selecting the relevant species, the active metabolites of XHP and the associated disease targets for obesity, along with their corresponding clinical drug targets, were uploaded to the String platform. This step facilitated the conversion of all data into standardized gene symbols, thereby ensuring consistency and accuracy for subsequent analyses.

Protein-protein interaction network construction: The protein-protein interaction (PPI) network was constructed using Cytoscape 3.6.0 software[19]. Initially, a table delineating the one-to-one correspondence between XHP-predicted metabolites and their corresponding targets was imported into Cytoscape to establish the metabolite-target network. The PPI network pertinent to XHP’s role in obesity treatment was developed by integrating XHP-putative targets, obesity-related targets, and DEGs derived from the 3T3-L1 adipocyte differentiation experiments. Analysis of the PPI network was performed via the String platform prior to its visualization in Cytoscape 3.6.0. Subsequently, GO and KEGG pathway enrichment analyses were conducted.

Graphical visualization: Graphical visualizations including box plots, principal coordinate analysis (PCoA) plots, heatmaps, volcano plots, Venn diagrams, bubble charts, and chord diagrams were generated using the Bioinformatics platform (https://www.bioinformatics.com.cn). Mechanism summary diagrams were produced and exported using FigDraw (https://www.figdraw.com).

Molecular docking

Key proteins and metabolites’ molecular structures were retrieved from the Protein Data Bank (PDB)[20] and PubChem databases, respectively. These structures were optimized with Autodock and Chem3D software, then converted to the “pdbqt” format. Water molecules were removed, and polar hydrogen atoms were appended to the protein structures. The docking box was delineated based on data from the source literature. All structures were subsequently imported into Autodock Vina 1.2.2[21] for molecular docking validation, setting a docking score criterion of ≤ -5.0 to signify successful docking; lower scores indicated greater reliability of the results. The docking outcomes were analyzed using PyMOL software, which facilitated the selection of optimal docking poses based on the ligand-binding regions of the original protein structures. The molecular docking diagrams were subsequently exported. Network analysis and RNA sequencing (RNA-seq) data screening were employed to predict potential targets of XHP for obesity treatment, selecting combinations with the highest predicted efficacy for docking. Protein structures included in this study comprised leptin (PDB ID: 1AX8), PPARG (PDB ID: 2PRG), ADIPOQ (PDB ID: 5 LX9), IGF1 (PDB ID: 6FF3), LIPE (hormone-sensitive type, PDB ID: 8ZVQ), FASN (PDB ID: 7MHD), and SLC2A4 (PDB ID: 7WSM). Settings for the docking center and active pocket were based on structural data or literature references from the corresponding PDB IDs.

Statistical analysis

Statistical analyses were conducted using SPSS software, version 27.0. Data are presented as the mean ± SD. Comparisons among three or more groups were performed using one-way analysis of variance (ANOVA), while pairwise comparisons were executed using Dunnett’s t-test. A P value of less than 0.05 was considered statistically significant.

RESULTS

XHP inhibits MASLD-associated lipid accumulation in the liver and HepG2 cells

In a preceding clinical study involving 163 patients diagnosed with MASLD, treatment with XHP led to improvements in metabolic homeostasis. This was evidenced by reductions in the fatty liver index, MASLD fibrosis score, body mass index, and visceral fat area (Figure 1A).

Figure 1 Xietu Hemu prescription inhibits metabolic dysfunction-associated steatotic liver disease-associated lipid accumulation in vivo and in vitro. A: Xietu Hemu prescription (XHP) clinical treatment inhibits lipid accumulation of the liver and visceral adipose tissue; B: Flow chart of the preparation of XHP-containing serum (XHPS); C and D: Cell activity assay after intervention with different concentrations of XHPS (n = 3); E: Schematic diagram of metabolic dysfunction-associated steatotic liver disease model construction with the HepG2 cell line by XHPS intervention; F-H: Oil red O staining of HepG2 cells after XHPS intervention with different concentrations and statistics of differences in lipid droplet area occupation (n = 3). Data were expressed as mean ± SEM. 400 ×, 25 μm. ^aP < 0.001 vs control; ^bP < 0.001 vs model; ^cP < 0.05 vs low dose of XHP (L-XHP); ^dP < 0.01 vs L-XHP; ^eP < 0.001 vs L-XHP; ^fP < 0.05 vs medium dose of XHP. XHP: Xietu Hemu prescription; XHPS: Xietu Hemu prescription-containing serum; MASLD: Metabolic dysfunction-associated steatotic liver disease; L-XHP: Low dose of Xietu Hemu prescription; M-XHP: Medium dose of Xietu Hemu prescription; H-XHP: High dose of Xietu Hemu prescription; VehS: Vehicle serum; FBS: Fetal bovine serum; PA: Palmitic acid; OA: Oleic acid.

To corroborate the lipid-regulatory effects of XHP in vitro, serum containing the drug was utilized. XHPS and VehS were derived through transabdominal aortic sampling from rats subjected to continuous gavage for one week (Figure 1B). Given the complexity of rat serum compared to FBS, initial adjustments in the concentrations of XHPS and VehS provided to the cells were necessary. HepG2 cells were cultured with varying concentrations of XHPS and VehS for a duration of 72 hours to ascertain an optimal comparison with FBS. The outcomes of the CCK8 assay indicated that concentrations of 5% XHPS and VehS were optimal (Figure 1C). Further experiments to assess the impact of different XHPS concentrations on the differentiation process revealed that all tested doses were deemed safe (Figure 1D). Consequently, for the entirety of the cellular experiments, concentrations of 5% XHPS, 2.5% XHPS + 2.5% VehS, and 1.25% XHPS + 3.75% VehS were employed as high, medium, and low doses of drug-serum, respectively, and were designated as the high dose of XHP (H-XHP) group, medium dose of XHP (M-XHP) group, and low dose of XHP (L-XHP) group. The control, model, and silibinin groups were incubated with 5% VehS throughout the incubation period. The experimental protocol is depicted in Figure 1E.

After a 24-hour induction with PA or PA + OA, a subsequent 24-hour intervention with XHPS mitigated lipid accumulation in HepG2 cells, as evidenced by oil red O staining (Figure 1F-H). Under both PA and PA + OA induction protocols, the model group exhibited the largest lipid droplet area within comparable fields of view relative to other groups. In contrast, lipid droplet areas progressively diminished in the L-XHP, M-XHP, and H-XHP groups, all of which remained significantly smaller than those observed in the model group. This suggests that XHPS treatment alleviated hepatic lipotoxicity in a dose-dependent manner. Silibinin, a standard clinical intervention for MASLD, did not display a statistically significant difference in efficacy when administered at 0.1 mmol/L compared to H-XHPS therapy.

Given the clinical and experimental efficacy of XHP in treating MASLD with noted reductions in lipid accumulation (related to obesity), this study aims to systematically analyze its pharmacological effects.

Differential genes associated with 3T3-L1 cell differentiation at the mRNA level

The process by which 3T3-L1 cells differentiate into mature white adipocytes is depicted in Figure 2A. This study aimed to elucidate the changes in gene expression accompanying the differentiation of 3T3-L1 cells. For this purpose, RNA-seq analysis was conducted on both control and model groups. Following four days of induction with 3-isobutyl-1-methylxanthine, dexamethasone, and insulin (MDI) and a subsequent two days of insulin treatment, significant lipid droplet accumulation was observed in the model group relative to the control, as indicated by oil red O staining. This accumulation suggests successful differentiation into an adipose-like phenotype (Figure 2B). Consequently, RNA-seq analysis was employed to examine the differences in gene expression between the two groups. Figure 2C illustrates that gene expression distribution was remarkably consistent across samples from both groups, indicating stable expression levels across all samples. PCoA employing Bray’s distance algorithm revealed significant mRNA expression differences between the control and model groups, with similar expression patterns within each group (Figure 2D).

Figure 2 RNA sequencing analysis of the differences in 3T3-L1 cell differentiation. A: Schematic diagram of 3T3-L1 cell culture and differentiation process; B: Oil red O staining images in the control group vs model group on day 6; C: Box plot of gene expression of samples (n = 3); D: Principal coordinate analysis plot of gene expression of samples (n = 3); E: The number of up-regulated or down-regulated differential genes between control and model groups (n = 3); F and G: Heatmap (F) and volcano plot (G) of differential gene expression for genes demonstrating significant differences based on the screening criteria of |Log2FC| ≥ 2 and q value ≤ 0.05 (n = 3); H-N: Comparison expression of key genes for lipid droplet accumulation and differentiation in control and model groups. Data were expressed as mean ± SEM. 400 ×, 25 μm. ^aP < 0.05 vs control; ^bP < 0.005 vs control; ^cP < 0.001 vs control. PCoA: Principal coordinate analysis; NCS: Newborn calf serum; FBS: Fetal bovine serum; MDI: 3-isobutyl-1-methylxanthine, dexamethasone, and insulin; TPM: Transcripts per million.

The differential expression analysis identified 1751 genes with increased expression and 1571 genes with decreased expression, using |Log2FC| ≥ 1 and a q value ≤ 0.05 as the inclusion criteria. However, when the threshold for |Log2FC| was set to a minimum of 2, to place a greater emphasis on the multiplicity of differentials, the number of DEGs was reduced to 761 and 602, respectively (Figure 2E). These genes are considered highly differential and warrant further analysis. Figure 2F and G present a heatmap and a volcano plot of differential gene expression based on the aforementioned screening criteria. Among the top 10 differentially up/down-regulated genes, Thrsp, Mrap, Apoe3, Cyp2e1, Retnla, Pnpla3, Ch25h, Sorbs2, and Sfrp2, all of which have been reported to relate directly or indirectly to lipid metabolism, were notably ranked. In Figure 2H-N, key genes implicated in lipid metabolism were further extracted from the RNA-seq raw data, and their comparative transcripts per million values in each sample group were displayed. The results indicated that the model group exhibited significantly higher expression of key adipogenic and lipid storage transcription factors, such as Pparg, Cebpa, Fabp4, Fasn, and Srebp1/2, consistent with the oil red O staining results. Additionally, Klf4, which enhances key adipogenic transcription factors, was expressed at lower levels in the model group, potentially indicating alterations from the dysregulated lipid regulatory system observed in severely obese individuals. The observed changes in Klf4 mRNA levels may be attributable to a negative feedback mechanism regulating lipid droplet synthesis within the intracellular compartment during 3T3-L1 cell differentiation.

XHP inhibits lipid droplet accumulation and synthesis in vitro

The protocol for the differentiation of 3T3-L1 cells treated with XHPS is depicted in Figure 3A. Following a 6-10 days treatment period, the 3T3-L1 cells were stained with oil red O, enabling the visualization of differences in lipid accumulation among various groups (Figure 3B and C). On day 6, XHPS exhibited a dose-dependent inhibition of lipid droplet formation, which, although less potent than GW9662, a PPARγ inhibitor, still demonstrated significant efficacy. As the experimental period progressed, initially small lipid droplets coalesced into larger formations. Notably, by day 10, XHPS continued to effectively inhibit the maturation of 3T3-L1 cells, with minimal differences observed between the H-XHP group and the GW9662 treatment. Given the multi-targeted nature of XHPS as a cocktail therapy, the inhibition observed in the H-XHP group is particularly noteworthy in terms of drug safety and efficacy conversion rates. Microscopic analysis on day 10 further clarified the lipogenic inhibitory effects of XHPS, with statistically significant differences noted (Figure 3D and E). Consequently, all subsequent experiments were conducted with samples harvested following a 10-day differentiation cycle. Corresponding trends were observed in the intracellular levels of total cholesterol (TC) and triglyceride (TG), which paralleled the findings from the oil red O staining (Figure 3F and G). Immunoblotting analysis indicated that the protein expression levels of PPARγ, C/EBPα, and FABP4 were downregulated in the XHPS-treated groups compared to the model group, with the most pronounced downregulation detected in the H-XHP group. Among the treatment groups, the inhibitory effect was significantly stronger in the H-XHP group compared to the M-XHP and L-XHP groups. However, no significant differences were observed between the M-XHP and L-XHP groups (Figure 3H-K). These results collectively suggest that XHP is effective in inhibiting lipid droplet accumulation and lipid synthesis in vitro, highlighting its potential utility in regulating lipid metabolism and adipocyte differentiation.

Figure 3 Xietu Hemu prescription restrains lipid synthesis during the differentiation of 3T3-L1 cells. A: Schematic diagram of the differentiation process of 3T3-L1 cells by Xietu Hemu prescription (XHP)-containing serum intervention; B and C: Photos of oil red O staining of 3T3-L1 cell differentiation on day 6 and 10; D and E: Oil red O staining micro camera and statistics of difference in lipid droplet area occupation on day 10; F and G: Detection of intracellular lipid levels on day 10 (n = 3); H-K: Detection and quantitative statistics of intracellular lipid synthesis-related protein levels on day 10 (n = 3). Data were expressed as mean ± SEM. 200 ×, 50 μm; 400 ×, 25μm. ^aP < 0.01 vs model; ^bP < 0.005 vs model; ^cP < 0.001 vs model; ^dP < 0.05 vs low dose of XHP (L-XHP); ^eP < 0.01 vs L-XHP; ^fP < 0.005 vs L-XHP; ^gP < 0.001 vs L-XHP; ^hP < 0.05 vs medium dose of XHP (M-XHP); ⁱP < 0.005 vs M-XHP. XHPS: Xietu Hemu prescription-containing serum; L-XHP: Low dose of Xietu Hemu prescription; M-XHP: Medium dose of Xietu Hemu prescription; H-XHP: High dose of Xietu Hemu prescription; VehS: Vehicle serum; TC: Total cholesterol; TG: Triglyceride.

Chemical composition analysis of XHP by UPLC-QTOF/MS/MS

Through the application of UPLC-QTOF/MS/MS analytical techniques, our study successfully identified 233 chemical metabolites in XHP. The metabolites comprise 113 flavonoids, 20 phenylpropanoids, 5 polyacetylenes, 23 terpenoids, and 26 triterpenoid saponins, in addition to a diverse array of other metabolite classes including alkaloids, amino acids, organic acids, and iridoids. Figure 4 displays the base peak chromatogram of XHP. The structural features and fragmentation patterns of selected metabolites are depicted in Supplementary Figures 1 and 2.

Figure 4 Xietu Hemu prescription component identification. A and B: Ultra performance liquid chromatography-quadrupole-time of flight mass analyzer-tandem mass spectrometry of Xietu Hemu prescription (XHP); C: The relationship of metabolites and botanical drugs of XHP.

Codonopsis Radix: This botanical drug yielded 41 metabolites across various chemical classes such as phenylpropanoids (e.g., Tangshenoside I), polyacetylenes (e.g., Lobetyolinin), and alkaloids (e.g., Codonopsine), which serve as characteristic markers. Hedysarum Multijugum Maxim: Analysis revealed 35 metabolites, primarily flavonoids and triterpenoid saponins, with Astragaloside IV highlighted as a diagnostic marker. Atractylodes Macrocephala Koidz: The analysis identified 13 metabolites, predominantly polysaccharides and sesquiterpenes (e.g., atractylenolides), indicative of this botanical drug. Citrus Reticulata and Citri Reticulatae Pericarpium Viride: These related samples contained 44 metabolites, predominantly flavonoids with a minor presence of limonoids (e.g., limonin). Bupleuri Radix: The composition analysis indicated 29 metabolites, with saikosaponins (e.g., saikosaponin A) as the primary markers. Scutellariae Radix: This sample was characterized by 24 flavonoid metabolites, particularly flavonoid glucuronides (e.g., baicalin, wogonoside). Ligustrum Lucidum Fructus: The analysis yielded 8 metabolites, notably iridoids (e.g., specnuezhenide) as characteristic markers. Folium Nelumbinis: This sample contained 23 metabolites, including both flavonoids and alkaloids. Alismatis Rhizoma: Identified 12 triterpenic acids, serving as characteristic metabolites. Additional botanical drugs: Fructus Liquidambaris and Arum Ternatum Thunb predominantly contained primary metabolites such as amino acids and nucleosides, lacking distinctive markers. Sedi Herba revealed 15 metabolites primarily comprising flavonoids and monoterpene glycosides. Crataegi Folium yielded 27 metabolites, dominated by flavonoids and phenylpropanoids.

Network analysis suggests XHP primarily targets lipid metabolism-related diseases

The literature review reveals a substantial association between XHP and obesity: XHP comprises 14 botanical drugs. In TCM formulations, the compatibility of botanical drugs is essential. Within XHP, Codonopsis Radix, Hedysarum Multijugum Maxim, and Atractylodes macrocephala Koidz function as the “monarch drugs” (Jun), forming the core of the formula. The remaining nine botanical drugs are categorized as “minister” (Chen), “assistant” (Zuo), and “guide” (Shi) drugs, which hierarchically support the monarch drugs to augment therapeutic efficacy. Figure 5A illustrates the proportions of the botanical drugs in XHP, with color coding denoting their roles in the TCM formulation. A search of CNKI (https://www.cnki.net) and PubMed (https://pubmed.ncbi.nlm.nih.gov) regarding the association of botanical drugs in XHP with obesity shows that all are implicated in obesity treatment to varying extents. Figure 5B and Supplementary Table 4 display the number of literature references related to obesity for these botanical drugs, while Figure 5C depicts the frequency of occurrence of XHP metabolites in each herb.

Figure 5 The correlation assessment of Xietu Hemu prescription and obesity. A: Composition of Xietu Hemu prescription (XHP) by “JUN-CHEN-ZUO-SHI”; B: The strength of the correlation between each XHP botanical drug with obesity was obtained from the literature search; C: Number of exclusive metabolites or overlap metabolites of each botanical drug in XHP; D: The structure of representative metabolites in XHP serum; E: Network of XHP serum metabolites with related targets; F: Protein-protein interaction network of the targets in E; G: Model decomposition of F into 7 sub-networks; H-K: Functional analysis of the sub-networks. H and I are the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of cluster 1, while J and K are the KEGG pathway and GO enrichment analyses of cluster 2. XHP: BP: Biological processes; CC: Cellular components; MF: Molecular functions.

Serum metabolites identification of XHP: Based on in vitro UPLC analysis, the extraction of blood entry metabolites was performed. The analysis detected a total of 23 metabolites in rat plasma, with flavonoids being the most absorbed type of metabolites in vivo. The specific distribution of their prototype metabolites is presented in Supplementary Table 5. Nineteen of these metabolites are listed in PubChem and are included in subsequent network pharmacological analyses, and Figure 5D illustrates their structural composition.

Network pharmacological analyses of XHPS: A total of 527 putative targets associated with 19 XHPS metabolites were identified and analyzed using various modular approaches. The relationship between the 19 XHPS metabolites and the 527 putative targets is outlined in Figure 5E. Following the construction of the PPI network and the decomposition using the MCODE model, GO and KEGG pathway enrichment analyses were conducted. The PPI network identifies AKT1, TNF, and ALB as potential hub-targets of XHPS (Figure 5F). The PPI network is subsequently divided into seven clusters (Figure 5G, each containing more than ten nodes). Functional analyses (Figure 5H-K) suggest that cluster 1 may represent the immunological module, involving pathways such as the “IL-17 signaling pathway” and “Th17 cell differentiation”. Meanwhile, cluster 2 appears to be the lipid metabolism module, involving pathways like the “AMPK signaling pathway”, “Insulin signaling pathway”, and “PI3K-AKT signaling pathway”. Collectively, these network pharmacological analyses of XHPS confirm a significant correlation with obesity.

Network intersection analysis suggests XHP regulates white adipose metabolism via the AMPK signaling pathway

Data integration identifies 24 core targets: To further elucidate the signaling pathways by which XHP regulates lipid metabolism, databases related to disease targets, including DisGeNET, GeneCards, and DrugBank, were employed to assemble a collection of obesity-related genes. By intersecting these obesity-related genes with DEGs exhibiting a |Log2FC| of at least 2 from 3T3-L1 cell RNA-seq data and putative targets of XHP, 24 genes were pinpointed as potentially playing a pivotal role in the regulation of lipid metabolism mediated by XHP (Figure 6A). These genes demonstrate intricate upstream and downstream relationships and form a densely interconnected PPI network, with PPARG emerging as the central hub gene within this network (Figure 6B).

Figure 6 Mechanism prediction of Xietu Hemu prescription therapy in obesity hints leptin is the hub-target and the adenosine monophosphate-activated protein kinase signaling pathway might be the core mechanism. A: Intersection results of Xietu Hemu prescription (XHP) putative targets screened from traditional Chinese medicine systems pharmacology (TCMSP) and PubChem, Obesity targets collected from DisGeNET, Genecards, and DurgBank, and RNA sequencing differential genes (|Log2FC| ≥ 2) computed by Dr. Tom platform; B: Protein-protein interaction network of intersecting targets; C and D: Enrichment analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO) for the 24 intersecting targets; E: Classification of the top 20 KEGG Pathways; F and G: The top GO terms and KEGG pathways about their respective mapping genes; H-N: Assessment of the tightness between XHP and the adenosine monophosphate-activated protein kinase signaling pathway by molecular docking. XHPS: Xietu Hemu prescription-containing serum.

Big data analysis predicts the AMPK signaling pathway as the central mechanism of XHP treatment: Functional enrichment analysis was conducted on these 24 core DEGs, with the top GO terms and KEGG pathways being ranked based on their P values and ratios (Figure 6C and D). In the GO enrichment analysis, “cellular response to lipid” was identified as the leading biological process term. Concurrently, the KEGG pathway analyses indicated that the “AMPK signaling pathway” was the highest ranked. These results imply that XHP may predominantly regulate both the formation and breakdown of white adipose tissue, including subcutaneous and visceral fat, primarily through the AMPK signaling pathway, potentially exerting a significant influence on systemic metabolic states.

To further investigate the direct and indirect targets of XHP, the top 20 KEGG pathways were visualized, emphasizing the pathway categories and their associated core genes. The AMPK signaling pathway was noted for containing the highest number of core genes (Figure 6E). Figure 6F and G delineate the relationships among all top GO biological process terms, KEGG pathways, and core genes. Notably, the AMPK signaling pathway implicates key genes such as PPARG, FASN, IGF1, SLC2A4 (also known as GLUT4), LEP, LIPE, and ADIPOQ, all of which require further validation to elucidate their interactions with XHP.

Molecular docking suggests LEP as a candidate core target of XHP on the AMPK signaling pathway: To evaluate the binding affinity and adaptability between candidate targets and their corresponding metabolites in XHP, molecular docking simulations were conducted using Autodock Vina 1.2.2. Based on bioinformatics enrichment analysis, it is postulated that XHP inhibits the differentiation and maturation of white adipose tissue via the AMPK signaling pathway. Consequently, all predicted targets of XHP within this pathway were selected for docking with their corresponding metabolites in the formula. PPARG, identified as the hub target, was associated with the highest number of corresponding metabolites (six), whereas FASN, IGF1, SLC2A4 (also known as GLUT4), LEP, LIPE, and ADIPOQ were each associated with one, one, one, two, two, and one metabolite(s), respectively. A docking score threshold of ≤ -5.0 was employed to define successful docking, resulting in the identification of 14 successful docking pairs. These pairs demonstrated binding through visible hydrogen bonds and robust electrostatic interactions. The detailed docking scores are presented in Supplementary Table 6. Figure 6H-N and Supplementary Table 7 illustrate seven representative docking pairs with highly stable binding conformations. Notably, the pairs “scutellarin to LEP” and “diosmetin to ADIPOQ” exhibited binding energies of -11.3 kcal/mol and -10.5 kcal/mol, respectively, signifying the most stable interactions. These findings suggest that XHP may modulate the AMPK signaling pathway cascade through LEP, potentially suppressing PPARG expression and thereby contributing to obesity treatment.

XHP inhibits the progression of 3T3-L1 cells into mature white adipocytes over the AMPK signaling pathway through LEP modulation

XHP activates AMPK phosphorylation by promoting LEP release during 3T3-L1 lipogenesis: Based on Figure 3, which demonstrated that XHP effectively inhibited lipogenesis and the synthesis of TC and TG in 3T3-L1 cells on day 10, the experimental conditions described in this section were replicated for mechanism validation. An ELISA assay was initially conducted to measure LEP levels. Since the 3T3-L1 cells in the control group were in an undifferentiated state, leptin secretion was virtually undetectable. Figure 7A indicates that XHP dose-dependently increased leptin secretion compared to the model group, with a significant rise in both intracellular and supernatant LEP levels, and a statistically significant increase in these levels in the H-XHP group compared to the L-XHP group. LEP-associated immunofluorescence staining of 3T3-L1 cell specimens revealed that XHP dose-dependently stimulated LEP protein expression (Figure 7B). Cytoplasmic and nuclear protein samples from 3T3-L1 cells were collected on day 10 for immunoblotting experiments, which clearly demonstrated that H-XHP stimulated the phosphorylation of AMPK but inhibited the phosphorylation of mTOR and the expression of PPARγ in the cytoplasm. Concurrently, H-XHP suppressed the protein expression of SREBP1/2 in the cytoplasm and the protein maturation of the precursor SREBP1/2 in the nucleus (Figure 7C-K).

Figure 7 Xietu Hemu prescription inhibits lipid synthesis of mature 3T3-L1 cells through the leptin signal axis at the protein level of the nucleus and cytoplasm. A: Secretion levels of leptin (LEP) at different time points during high dose of Xietu Hemu prescription (XHP) intervention in 3T3-L1 cell differentiation; B: Immunofluorescence assay for protein expression of LEP after 10 days of XHP intervention in 3T3-L1 cell differentiation at different concentrations; C-H: Detection and quantitative statistics of cytoplasm protein levels on day 10 (n = 3); I-K: Detection and quantitative statistics of nucleus protein levels on day 10 (n = 3). Data were expressed as mean ± SEM. 200 ×, 50 μm; 400 ×, 25 μm; ^aP < 0.05 vs control; ^bP < 0.01 vs control; ^cP < 0.005 vs control; ^dP < 0.001 vs control; ^eP < 0.05 vs model; ^fP < 0.01 vs model; ^gP < 0.005 vs model; ^hP < 0.001 vs model; ⁱP < 0.05 vs low dose of XHP (L-XHP); ^jP < 0.001 vs L-XHP; ^kP < 0.001 vs medium dose of XHP. L-XHP: Low dose of Xietu Hemu prescription; M-XHP: Medium dose of Xietu Hemu prescription; H-XHP: High dose of Xietu Hemu prescription; LEP: Leptin.

Inhibitory effect of H-XHP on lipogenesis in 3T3-L1 cells disappears after disruption of LEP-LEPR binding accompanied by inactivation of AMPK phosphorylation: To ascertain if LEP is indispensable for the effects of XHP treatment, siRNAs were employed to disrupt LEPR expression. Three siRNAs were designed and synthesized; siLEPR1 and siLEPR2 successfully reduced LEPR expression in 3T3-L1 cells by over 80%, effectively inhibiting the binding of LEP to LEPR as demonstrated in Figure 8A and B. Considering the transient nature of gene silencing by siRNA, which typically does not exceed one week, and based on the results from oil red O staining shown in Figure 3, indicating superior inhibition of lipogenesis by H-XHP on day 6, the duration of the siRNA experiment was adjusted to six days.

Figure 8 Inhibition of lipid synthesis by Xietu Hemu prescription following siRNA disruption of leptin receptor in 3T3-L1 cells. A and B: SiRNA disruption of 3T3-L1 cell LEPR efficiency test; C and D: Oil red O staining micro camera and statistics of difference in lipid droplet area occupation on day 10 with/without high dose of Xietu Hemu prescription (H-XHP) or siLEPR intervention; E-J: Detection and quantitative statistics of cytoplasm protein levels on day 10 with/without H-XHP or siLEPR intervention (n = 3); K-M: Detection and quantitative statistics of nucleus protein levels on day 10 with/without H-XHP or siLEPR intervention (n = 3). Data were expressed as mean ± SEM. 200 ×, 50 μm; 400 ×, 25 μm; ^aP < 0.05 vs vehicle; ^bP < 0.05 vs model-vehicle; ^cP < 0.01 vs model-vehicle; ^dP < 0.005 vs model-vehicle; ^eP < 0.001 vs model-vehicle; ^fP < 0.05 vs model-siLEPR1; ^gP < 0.01 vs model-siLEPR1;^hP < 0.001 vs model-siLEPR1; ⁱP < 0.05 vs high dose of Xietu Hemu prescription (H-XHP)-vehicle; ^jP < 0.01 vs H-XHP-vehicle; ^kP < 0.005 vs H-XHP-vehicle; ^lP < 0.001 vs H-XHP-vehicle. H-XHP: High dose of Xietu Hemu prescription.

Figure 8C and D shows that, in the absence of H-XHP (the three model groups), the size of lipid droplets in equivalent fields of view was marginally and not significantly different in the siLEPR1 and siLEPR2 cell lines compared to the vehicle control lines. Conversely, in the H-XHP-treated cells (the three H-XHP groups), lipid droplets were substantially reduced in the vehicle control lines compared to those in siLEPR1 and siLEPR2, suggesting that LEP-LEPR interaction plays a crucial role in H-XHP’s therapeutic efficacy. Subsequent immunoblotting of control, vehicle, and siLEPR1 cell lines revealed significant alterations: A decrease in the ratio of p-AMPK to total AMPK, an increase in the ratio of p-mTOR to total mTOR, and in PPARγ levels in the model-siLEPR1 group compared to the model-vehicle group. While SREBP1/2 increased slightly, this difference was not statistically significant. However, these changes were more pronounced in the H-XHP-Vehicle group compared to the H-XHP-siLEPR1 group, indicating that leptin secretion crucially influences lipidogenic differentiation in 3T3-L1 cells, and that the therapeutic effect of H-XHP is mediated through the interaction of LEP with LEPR, as shown in Figure 8E-J. Further immunoblotting of nuclear proteins showed that the expression trends of SREBP1/2 in the nucleus mirrored those of their cytoplasmic counterparts (Figure 8K-M), suggesting that the inhibitory action of H-XHP on the synthesis of nuclear SREBP1/2 precursors is also mediated by its impact on leptin secretion.

DISCUSSION

The 3T3-L1 cell line, recognized for its established use and widespread acceptance as an in vitro model for adipocyte differentiation, replicates a transcriptional profile highly akin to that of human adipose tissue. This model can be reliably induced to differentiate into white adipocytes, rendering it invaluable for phenotypic and molecular studies of adipogenesis, as well as for the evaluation of anti-obesity agents and the elucidation of their mechanisms of action[22-25]. Notably, comparative studies of 3T3-L1 cells and mouse adipose tissue sections have demonstrated similar caveola densities, suggesting a strong correlation between the 3T3-L1 model and in vivo obesity models[26].

In our investigations, MDI-induced differentiation of 3T3-L1 cells manifested pronounced adipogenic characteristics, as evidenced by morphological alterations and the upregulated expression of lipid metabolism markers such as PPARγ, C/EBPα, FABP4, and SREBP1. These observations affirm the efficacy and reliability of the 3T3-L1 model for assessing the anti-obesity effects of XHP in vitro. Transcriptomic analysis of 3T3-L1 cells identified distinctive gene expression profiles between the control and model groups subsequently. The ten most highly expressed genes in the model group included Thrsp, Map7d2, Fam107a, Mrap, Ang5, Cyp2e1, Retnla, Pnpla3, Apoe3, and Slc22a20, while the control group exhibited elevated expression of genes such as Cxcl5, Sfrp2, Kif26b, Fgf23, Ndst3, Tgtp2, Sorbs2, Ch25h, Cpz, and Inmt. Given that the cells of the control group remained undifferentiated, their gene expression profile may not accurately represent the adipose phenotype of lean individuals. Conversely, the differentiated state of the model group provides a more representative basis for modeling obese phenotypes, thus rendering the exploration of highly expressed genes in the model group particularly pertinent. Several of these DEGs have been previously associated with obesity and lipid metabolism, meriting further exploration and discussion.

Pnpla3 has been well-established as a genetic determinant of fatty liver disease and a heritable risk factor for MASLD and metabolic syndrome. Specific alleles of Pnpla3 are significantly associated with an increased risk of liver-related mortality[27]. Cyp2e1 is a cytochrome P450 enzyme implicated in the decomposition and browning of white fat, as well as in the pathogenesis of fatty liver disease[28,29]. Notably, a deficiency in Cyp2e1 activates the expression of hepatic peri-PPARα target genes, including FGF21, which are secreted by the liver to enhance fat browning and energy expenditure, thereby reducing obesity[28]. Ang5 and Apoe3 both play roles in the regulation of lipoproteins. Thrsp is highly expressed in conditions of obesity and non-alcoholic fatty liver disease. Studies employing Thrsp-deficient mice have highlighted the role of Thrsp in adipogenesis and obesity[30]. However, despite enhanced adipogenic differentiation, Thrsp expression was paradoxically downregulated in obese individuals, indicating a potentially complex role in obesity that merits further investigation[31]. In contrast, the genes Map7d2, Fam107a, Mrap, Retnla, and Slc22a20 have been less extensively studied in the context of obesity. These genes may represent novel targets for future research and potential therapeutic interventions. Together, these DEGs constitute the phenotypic basis of adipocyte differentiation and offer potential molecular targets for XHP-targeted interventions.

XHP, a multi-metabolite, multi-target TCM prescription, has demonstrated considerable therapeutic effects in a 3T3-L1 cell-based obesity model. In contrast, GW9662, a known PPARγ antagonist, also exhibited inhibitory effects on adipogenesis; however, its safety profile and systemic applicability were inferior to those of XHP. Network and transcriptomic analyses of 3T3-L1 cells have identified key signaling pathways involved in lipid metabolism that are modulated by XHP. Among the core targets, PPARG, LEP, FASN, and SLC2A4 were prominently enriched within the AMPK signaling axis. The molecular docking results further confirmed strong binding affinities between the active metabolites in XHP and LEP, suggesting that LEP may serve as a central target of XHP, with the AMPK signaling axis potentially being its primary mechanism of action.

LEP, a hormone secreted by adipocytes, regulates energy metabolism via hypothalamic centers. It has been demonstrated that LEP activates pro-opiomelanocortin neurons, thereby inducing satiety, while simultaneously inhibiting AgRP/NPY neurons, a process that attenuates hunger signals and promotes appetite suppression[32]. Furthermore, evidence indicates that LEP can stimulate thermogenesis in brown adipose tissue through sympathetic nerves[33,34], and it has been shown to downregulate genes associated with adipogenesis (e.g., SREBP1c, FAS) through the JAK2-STAT3 pathway[35]. LEP also enhances insulin sensitivity in peripheral tissues (e.g., liver, muscle) and reduces ectopic lipid deposition[36,37]. Despite elevated serum LEP levels in many obese individuals, its regulatory function is often compromised due to central LEP resistance, which contributes to excessive energy intake and fat accumulation. The mechanisms of LEP resistance primarily include impaired transport across the blood-brain barrier and disrupted signaling through multiple LEPR pathways[38,39], potentially involving negative regulatory molecules such as SOCS3 and PTP1B that inhibit JAK2-STAT3 signaling[40-43]. This resistance results in diminished hypothalamic sensitivity to LEP and an inability to effectively suppress appetite and promote energy expenditure. Clinical studies have revealed that exogenous LEP has limited efficacy in general obese populations but demonstrates significant effects in LEP-deficient individuals[44], highlighting LEP resistance as a critical target for obesity treatment.

In the obese state, tissue AMPK signaling pathway activity tends to decrease, with the inhibitory effect being particularly pronounced in models of high-fat diet-induced obesity[45,46]. Reduced AMPK activity significantly diminishes fatty acid oxidation, which in turn, contributes to the development of lipid accumulation and insulin resistance[47]. In the central nervous system, LEP exerts its appetite-suppressing effects by specifically inhibiting AMPK activity in the hypothalamus[48]. In peripheral tissues, LEP’s modulation of lipid metabolism is also closely linked with AMPK signaling. For instance, in hepatic tissues, LEP and AMPK collaboratively inhibit lipogenesis by synergistically suppressing SREBP1c expression[49,50]. LEP serves as an energy storage signal to centrally regulate appetite and energy expenditure, while AMPK functions as an energy sensor to coordinate lipid metabolism in peripheral tissues. The concurrent dysfunction of LEP resistance and AMPK inhibition leads to lipid accumulation and metabolic abnormalities in obesity. Restoring the activity of these pathways may therefore represent a viable strategy for the intervention of obesity.

During the conclusive experiments for mechanism validation, XHP was observed to significantly enhance leptin secretion levels in both the cytoplasm and culture supernatants. Additionally, the si-LEPR interference assay substantiated that LEP-LEPR binding is essential for XHP’s efficacy, indicating that XHP may impede white adipose production by fostering LEP autocrine secretion and activating AMPK, thereby down-regulating the expression of PPARγ within the adipogenic pathway, including SREBP1/2. XHP modulates energy metabolism and fat deposition across individuals via the LEP and AMPK pathways. The role of XHP in inhibiting leukocyte adipogenesis through the LEP/AMPK/PPARγ pathway provides an experimental basis for the clinical advancement of XHP in the prevention and treatment of obesity and MASLD.

There are also some limitations to the research. Although this paper was based on experiments in vitro that found XHP promotes leptin secretion, investigations into whether XHP improves LEP sensitivity in obese individuals remain unexplored, marking a limitation that will be addressed in forthcoming in vivo experiments. Additionally, given that the LEP/AMPK/PPARγ pathway is implicated in the distribution of visceral and subcutaneous adipose tissue as well as in adipose browning, it is imperative that the anti-obesity effects of XHP be validated in vivo with a focus on these specific aspects. Furthermore, prior clinical studies have demonstrated that XHP improves key obesity-related parameters and established MASLD surrogates, with a trend toward enhanced insulin sensitivity (P = 0.06; n = 262). These XHP-mediated improvements warrant further validation in animal models to determine whether their changes are associated with the LEP/AMPK/PPARγ signaling pathway under XHP stimulation. Finally, given the established link between insulin sensitivity and type 2 diabetes mellitus (T2DM), the potential of XHP as a therapeutic intervention for obesity-related T2DM also merits further investigation.

CONCLUSION

In summary, this integrative multi-omics investigation definitively elucidates the molecular mechanism underpinning XHP’s efficacy against obesity, establishing the LEP/AMPK/PPARγ signaling axis as its central therapeutic pathway. We demonstrate that XHP orchestrates its anti-adipogenic effects not through broad metabolic suppression but by specifically augmenting autocrine leptin secretion, which serves as the critical initiator for AMPK phosphorylation. This activation instigates a downstream cascade that transcriptionally represses the master regulators of adipogenesis, PPARγ and SREBPs, thereby attenuating lipid droplet accumulation and restoring metabolic homeostasis. The unequivocal abolition of XHP’s therapeutic impact upon LEPR knockdown provides conclusive evidence of the pathway’s indispensability, positioning XHP as a sophisticated, multi-targeted therapeutic strategy that counteracts MASLD-related obesity by directly targeting the foundational process of lipotoxicity.