Guiqi Yimu Powder (Beijing Edison Biotechnology Co., Ltd., batch number: 2103264) was composed of Angelica sinensis, Astragalus membranaceus, and Leonurus japonicus in a ratio of 2:2:1. Each 1 mL of the extract corresponds to 0.5 g of the raw medicinal herb. During preparation, raw materials underwent selection and quality control to ensure their active ingredients and purity met the required standards. The herbs were then washed, dried, and coarsely ground for solvent extraction. The extract was filtered and concentrated under reduced pressure using a rotary evaporator to remove most of the solvent. Subsequently, the concentrated extract was converted into powder using vacuum freeze-drying technology and was standardized to ensure consistency in active ingredients.

The prepared Guiqi Yimu Powder underwent three rounds of ultra-high performance liquid chromatography–tandem mass spectrometry (UHPLC–MS/MS) analysis. For the sample preparation, the medicinal solution was thawed at 4°C and vortex mixed until uniform. A 1 mL aliquot of the solution was mixed with 3 mL of absolute ethanol, vortexed for 3 min, and subjected to ultrasonication in an ice bath for 10 min. The mixture was left to stand at room temperature for 12 h, followed by centrifugation at 7,000 g for 20 min at 4°C. The supernatant was collected and dried under nitrogen to obtain the refined extract. For mass spectrometry, the dried extract was reconstituted in ultrapure water to a final volume of 1 mL and filtered through a 0.22 μm membrane. Throughout the analysis, samples were kept in a 4°C autosampler. Chromatographic separation was performed using a SHIMADZU LC-30 UHPLC system equipped with an ACQUITY UPLC^® HSS T3 column (2.1 × 100 mm, 1.8 μm) from Waters (Milford, MA, United States). Electrospray ionization (ESI) was employed for both positive (+) and negative (−) ion modes. Post-UPLC separation, mass spectrometric analysis was conducted on a QE Plus mass spectrometer (Thermo Scientific) utilizing a heated electrospray ionization (HESI) source. The total run time for each sample was 15 min. The mass range for the parent ion scan was set to 75–1,050 m/z. Raw data were processed using MSDIAL software for peak alignment, retention time correction, and peak area extraction. Metabolite structures were identified by matching accurate mass values (mass tolerance <0.01 Da) against a TCM database.

In this study, we utilized the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database¹ to identify all the compounds in Guiqi Yimu Powder, which includes Astragalus membranaceus, Leonurus japonicus, and Angelica sinensis (24). Literature searches and database queries were performed based on the chemical components identified from the UHPLC–MS/MS analysis. A compound-target network for Guiqi Yimu Powder was constructed using Cytoscape v3.10.1 software, incorporating the screened compounds and their corresponding target information. To construct a protein–protein interaction (PPI) network, the target genes associated with Guiqi Yimu Powder were uploaded to the STRING database (v12.0) ² to investigate the interactions between target proteins (25). Furthermore, we used ClueGo software to perform a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis on the proteins and genes within the PPI network. These analyses are intended to elucidate the mechanisms by which Guiqi Yimu Powder treats bovine fatty liver at the levels of gene function and signaling pathways (26, 27).

Dairy cow hepatocytes were provided by Professor Yanfeng Ma from Ningxia University and were purchased from BLUEFBIO (Cat# BFN60810761, Shanghai, China). The cell density was adjusted to 1 × 10⁶ cells/mL and seeded into T75 culture flasks. The culture medium consisted of Dulbecco’s Modified Eagle Medium (DMEM, Cat# c11995500BT, Thermo Fisher Scientific, Beijing, China), 10% fetal bovine serum (FBS, Cat# 1099-141, Thermo Fisher Scientific, Australia), and 1% penicillin–streptomycin. The cells were incubated at 37°C in a humidified atmosphere containing 5% CO₂. When the cells reached approximately 80% confluence, the medium was replaced with serum-free DMEM for a further 12 h of incubation.

For treatment, the medium was replaced, and cells were divided into groups. The control group received fresh complete medium, while experimental groups were exposed to DMEM with 200, 300, or 400 μmol/L TMAO (Cat# T833724-1 g, Macklin Biochemical Co, Ltd., Shanghai, China), for 12 h to establish the optimal lipid-inducing concentration of TMAO, determined via liver function assays. Subsequently, hepatocytes at stable passage underwent the aforementioned starvation protocol and were randomly assigned into new groups. The lipid accumulation group (Group A) received DMEM with the optimal TMAO concentration. Guiqi Yimu Powder treatment groups (Groups B1–B5) were supplemented with the same TMAO concentration and varying concentrations of Guiqi Yimu Powder (10, 20, 30, 40, 50 g/L). Cells were incubated for an additional 12 h under identical conditions. Post-treatment, cell pellets were harvested, flash-frozen in liquid nitrogen for 30 min, and stored at −80°C for further analysis.

The collected cell pellets were lysed using an ultrasonic homogenizer. The levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and triglycerides (TG) in the samples were measured using biochemical assay kits (microplate method, AST: C010-2-1, ALT: C009-2-1, and TG: A110-1-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions. Subsequently, lipid droplets were stained using an Oil Red O Staining Kit (C0158S, Shanghai Beyotime Biotechnology, Shanghai, China). The collected samples were fixed in 10% neutral formalin, followed by dehydration with a series of ethanol solutions. After rehydration, the samples were stained with freshly prepared Oil Red O solution for 15 min. Excess stain was removed by washing with 60% isopropanol, and counterstaining was performed with hematoxylin. The stained samples were then mounted with glycerin gelatin and observed under a microscope to quantify the lipid droplets.

To ensure the quality of experimental data, we selected three groups of samples for RNA sequencing, with each group consisting of three replicates to enhance the reliability of the results and ensure statistical significance. The RNA-seq services were provided by Shanghai Bioprofile Co, Ltd. (Shanghai, China). Total RNA was extracted from the collected cell samples using Trizol reagent (Cat# CW0580, CWbio, Jiangsu, China). The concentration, quality, and integrity of the RNA were then determined using a NanoDrop spectrophotometer (Thermo Scientific). The constructed libraries were purified using the AMPure XP system (Beckman Coulter, Beverly, CA, United States) and quantified using the Agilent High Sensitivity DNA Kit on the Agilent 2100 Bioanalyzer system. Finally, high-throughput sequencing of the libraries was conducted on the Illumina NovaSeq 6000 platform.

The raw data in FASTQ format were filtered using Cutadapt (v1.15) software, and the filtered reads were aligned to the reference genome using HISAT2 (v2.0.5) (28). For the quantitative analysis of gene expression levels, HTSeq (v0.9.1) was used to count the read numbers mapped to each gene, which served as the initial expression value (29). These expression values were then normalized using the FPKM (Fragments Per Kilobase of transcript per Million mapped reads) method. Principal Component Analysis (PCA) was performed using the Omicshare data processing platform developed by GENE DENovo ³ under default settings. Differentially expressed genes (DEGs) were identified using the DESeq (v1.30.0) R package with the criteria |FC| > 1.5 and p < 0.05 (30). Clustering analysis of gene expression patterns among samples was conducted using the Pheatmap package (v1.0.8) in R, and the results were visualized as heatmaps (31). Finally, the DEGs were annotated using the KEGG database ⁴ (32), and their enrichment was analyzed with the ClusterProfiler package (v3.4.4) (33).

To ensure consistency between the proteomics and transcriptomics datasets, we conducted proteomics analysis on the same set of three samples. The proteomics analysis services were provided by Shanghai Bioprofile Co., Ltd. (Shanghai, China). Proteins were extracted from the cell samples (n = 3) using SDT lysis buffer (4% SDS, 100 mM DTT, 100 mM Tris–HCl, pH 8.0). The samples were boiled for 5 min and then sonicated, followed by another boiling step for 5 min. After centrifugation (16,000 g, 15 min), the supernatants were collected, and the protein concentrations were determined using the BCA Protein Assay Kit (Bio-Rad, United States). For each sample, 200 μg of protein was digested using the FASP method as described by Wisniewski et al. The resulting peptides were desalted using C18 StageTips prior to LC–MS analysis (34). The peptide concentrations were measured at OD280 using a Nanodrop One device.

LC–MS/MS analysis was performed using a Q Exactive Plus mass spectrometer coupled with an Easy 1,200 nLC system (Thermo Fisher Scientific). Mass spectrometry data were analyzed using MaxQuant software (v1.6.0.16) (35), with protein identification conducted against the UniProtKB Rattus norvegicus database. Protein quantification in MaxQuant was carried out using the ratio median normalization method (36). Principal Component Analysis (PCA) of protein expression levels was performed using the Omicshare data processing platform. ⁵ Fold changes (FC) in protein abundance and p-values were calculated based on LFQ values, with selection criteria set at |FC| > 1.5 and p < 0.05. The differentially expressed proteins were annotated and subjected to significant enrichment analysis using the KEGG database (37).

A larger sample size (n = 6) was utilized for metabolomics analysis to capture a broader spectrum of metabolites and ensure sufficient statistical power. These cell samples included the same three samples initially used for transcriptomics and proteomics, along with an additional three samples to enhance the robustness of metabolite detection. The untargeted metabolomics analysis services were provided by Shanghai Bioprofile Co., Ltd. (Shanghai, China). The collected cell samples (n = 6) were mixed with 1 mL of pre-cooled methanol/acetonitrile/water (v/v/v, 2:2:1) and sonicated in an ice-water bath for 1 h. The samples were then incubated at −20°C for 1 h to precipitate proteins. After centrifugation at 14,000 g for 20 min (4°C), 800 μL of the supernatant was transferred to sampling tubes. The vacuum-dried extract was reconstituted in 50% acetonitrile, filtered through a 0.22 μm cellulose acetate membrane, and transferred to 2 mL HPLC vials, then stored at −80°C until analysis.

Metabolomics analysis was performed using a UPLC-ESI-Q-TOF-MS system (UHPLC, Shimadzu Nexera X2 LC-30 AD, Shimadzu, Japan) in combination with a Q-Exactive Plus mass spectrometer (Thermo Scientific, San Jose, United States). Liquid chromatography (LC) separation was conducted with an ACQUITY UPLC^® HSS T3 column (2.1 × 100 mm, 1.8 μm) (Waters, Milford, MA, United States). The flow rate was set to 0.3 mL/min with mobile phases A (0.1% formic acid in water) and B (100% acetonitrile, ACN). The gradient conditions were: 0% B for 2 min, a linear increase to 48% B over 4 min, an increase to 100% B over the next 4 min, held for 2 min, followed by a rapid decrease to 0% B in 0.1 min, with re-equilibration for 3 min. Electrospray ionization (ESI) was performed in both positive and negative ion modes for mass spectrometry data acquisition. Multivariate data analysis and modeling were performed using R (v4.0.3) and relevant R packages. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were used to construct the models (38). OPLS-DA was employed to determine the importance of variables to the projection (VIP), with significant differential metabolites being selected based on VIP ≥ 1 and FC > 1.5 or FC < 0.667, and p < 0.05. Clustering analysis of the selected differential metabolites was conducted using R, and KEGG pathway enrichment analysis was performed using the KEGG database to identify key metabolic pathways. For a systematic study of metabolic changes, overall trends in metabolic pathways were analyzed based on the abundance of differential metabolites. The annotation results of the differential metabolites were used to calculate the DA score, and the top 30 pathways were visualized for further insight.

Based on the KEGG pathway enrichment analysis results, which were significantly associated with the regulatory mechanisms of Guiqi Yimu Powder on fatty liver cells (p < 0.05), key differential genes, proteins, and metabolites were selected. A multi-omics functional network diagram was created using Cytoscape software (39). Correlation heatmaps were constructed using the Pearson algorithm in R (v4.0.3), and correlation networks were also visualized with Cytoscape software. Finally, we integrated the multi-omics data to construct a conceptual model illustrating the potential impact of TMAO on lipid metabolism in bovine hepatic cells.

All data were initially processed using Excel software and subsequently analyzed with GraphPad Prism software (v. 8.0). Analysis of Variance (ANOVA) was performed for the statistical analysis of data results. Pairwise significant differences between groups were evaluated using Duncan’s test. Differences were considered statistically significant when p < 0.05 and highly statistically significant when p < 0.01.

The total ion chromatograms of Guiqi Yimu Powder were shown in Figure 1. By comparing the UHPLC–MS/MS results with the precursor ion mass-to-charge ratios, secondary spectra, or retention times of standard substances or database entries, a total of 348 chemical components were identified in Guiqi Yimu Powder (Supplementary Table S1). These components mainly include flavonoids, phenols, carboxylic acids and their derivatives, steroids and their derivatives, fatty acids and their derivatives, hydroxy acids and their derivatives, indoles and their derivatives, among others.

Notably, among the 348 compounds, 7 major components were identified: Daidzein, Kaempferol, Adenine, Myrcene, Calycosin, Formononetin, and Stachydrine. These 7 compounds are the primary constituents of Angelica sinensis, Astragalus membranaceus, and Leonurus japonicus, as marked in Figure 1. Consequently, it can be inferred that these compounds reflect the major extract components of the three medicinal plants present in Guiqi Yimu Powder.

All the chemical components present in Astragalus membranaceus, Leonurus japonicus, and Angelica sinensis within Guiqi Yimu Powder were retrieved using the TCMSP database. We screened and identified seven active components: four from Astragalus membranaceus, two from Angelica sinensis, and two from Leonurus japonicus, where kaempferol is a shared component of both Astragalus membranaceus and Leonurus japonicus. These active components correspond to 106 potential target genes. A regulatory network illustrating these interactions was then constructed (Figure 2A), providing a comprehensive visualization of the interactions between Guiqi Yimu Powder’s active components and their corresponding targets.

Gene data related to the pathogenesis of fatty liver disease were collected from the GeneCards and DisGeNet databases. After removing redundant targets, 11,571 disease-related targets were identified. By intersecting the 106 target genes of Guiqi Yimu Powder’s active components with these disease-related genes, we identified 43 potential therapeutic targets for the powder. To further understand the mechanism of action of Guiqi Yimu Powder, we used STRING software to construct a PPI network for these intersected genes (Figure 2B). The PPI network revealed extensive interactions among these genes’ proteins. By analyzing the network’s topology, we identified several core proteins occupying key positions within the network, such as IL6, TNF, PPARG, CASP3, and BCL2. These proteins were likely to play crucial roles in Guiqi Yimu Powder’s treatment of fatty liver. Figure 2C showed the major signaling pathways identified through KEGG enrichment analysis (p < 0.05). The genes were found to be involved in multiple pathways, including lipid metabolism pathways like the Regulation of lipolysis in adipocytes and the Insulin resistance, inflammation-related pathways like the IL-17 signaling pathway and the NF-κB signaling pathway, and cell proliferation and apoptosis pathways such as the Apoptosis pathway and Longevity regulating pathway. These pathways are critical in the development and progression of fatty liver. Through the construction and analysis of the target PPI network for Guiqi Yimu Powder, we uncovered its multi-target and multi-pathway synergistic mechanisms in treating fatty liver. These findings provide a theoretical basis for its effective clinical application.

In this study, we investigated the effects of various concentrations of TMAO on bovine hepatocytes. We assessed the extent of hepatocyte injury by measuring the intracellular levels of AST, ALT, and TG. The experimental results indicated that TMAO elevated the levels of AST, ALT, and TG at all tested concentrations. Notably, AST levels significantly increased at concentrations of 300 μmol/L and 400 μmol/L (p < 0.05, Figure 3A). Similarly, ALT and TG levels showed significant increases following treatment with different TMAO concentrations (p < 0.05, Figures 3B,C). These data suggest that TMAO induces hepatocyte injury across all tested concentrations, with the most pronounced increases in cell injury and TG levels observed at a concentration of 400 μmol/L.

Next, we investigated the ameliorative effects of Guiqi Yimu Powder on TMAO-induced hepatocyte injury and lipid droplet accumulation. As shown in Figure 4, TG levels were significantly elevated (p < 0.01) in the group with lipid accumulation induced by TMAO at the optimal concentration compared to the control group. Treatment with Guiqi Yimu Powder at concentrations of 10 g/L, 20 g/L, and 30 g/L significantly reduced TG levels in hepatocytes with fatty liver (p < 0.01). The lowest TG levels were observed in the group treated with 40 g/L Guiqi Yimu Powder, which also showed a significant difference compared to the previous group. Thus, a concentration of 40 g/L of Guiqi Yimu Powder was most effective in reducing TG levels and could be considered the optimal concentration for preventing hepatocyte lipid accumulation.

To further elucidate the impact of TMAO on lipid accumulation in hepatocytes, we performed Oil Red O staining with 400 μmol/L TMAO. In the control group, bovine hepatocytes displayed deep blue nuclei with sparse red lipid droplets scattered in the cytoplasm (Figure 5A). However, treatment with 400 μmol/L TMAO resulted in a substantial accumulation of dispersed red lipid droplets in the cytoplasm (Figure 5B). Compared to the TMAO-only group, the Guiqi Yimu Powder treatment group showed a significant reduction in red lipid droplets within the cytoplasm (Figure 5C). In summary, a concentration of 400 μmol/L TMAO caused significant hepatocyte injury and lipid droplet accumulation in bovine hepatocytes In contrast, Guiqi Yimu Powder markedly mitigates TMAO-induced hepatocyte injury and lipid droplet accumulation, demonstrating its potential protective effects.

PCA based on expression levels showed that biological replicates within the same experimental groups clustered tightly together (PC1: 42.3% and PC2: 18.9%), and samples from different experimental groups were clearly separated along PC1, confirming the reliability and stability of the sequencing results (Figure 6A). Differential gene expression analysis identified a total of 1,536 significantly differentially expressed genes (Supplementary Table S2, p < 0.05), with 1,123 genes upregulated and 413 genes downregulated (Figures 6B,C). Cluster analysis was used to generate a heatmap of differentially expressed genes, which demonstrated good reproducibility within groups and clear differences between groups, indicating that Guiqi Yimu Powder treatment significantly altered the gene expression profiles in bovine hepatocytes (Figure 6D). To more clearly illustrate the significant differences between the two comparison groups, a radar chart was created, detailing the expression patterns of the top 20 differentially expressed genes (Figure 6E).

KEGG pathway enrichment analysis of the differentially expressed genes revealed several key pathways related to metabolism, inflammation and oxidative stress, and signal transduction. The analysis showed that metabolism-related pathways were mainly concentrated in steroid biosynthesis, while inflammation and oxidative stress-related pathways were enriched in lipid and atherosclerosis, hepatitis B, and hepatitis C. Signal transduction pathways included key pathways such as the p53 signaling pathway and the MAPK signaling pathway (p < 0.05, Figure 6F).

To elucidate the mechanisms by which Guiqi Yimu Powder affects lipid metabolism in bovine hepatocytes, we employed Label-Free Quantification (LFQ) proteomics to identify proteins within the cells. PCA scatter plots showed a clear separation between the TMAO-treated group and the control group along the first principal component (PC1), indicating that TMAO treatment significantly altered the overall protein expression profiles in bovine hepatocytes (Figure 7A). Subsequent differential expression analysis identified a total of 152 significantly differentially expressed proteins (Supplementary Table S3, p < 0.05), with 107 proteins upregulated and 45 downregulated (Figures 7B,C). A hierarchical clustering heatmap visually demonstrated the expression differences of these proteins between the different treatment groups (Figure 7D).

KEGG enrichment analysis of the differentially expressed proteins revealed that metabolism-related pathways were primarily enriched in metabolic pathways, fatty acid metabolism, and tyrosine metabolism. In contrast, inflammation and oxidative stress-related pathways were mainly enriched in ferroptosis and the complement and coagulation cascades. Pathways related to signal transduction included key pathways such as cell adhesion molecules (CAMs) and the Hippo signaling pathway (p < 0.05, Figure 7E).

To further investigate the metabolic mechanisms by which Guiqi Yimu Powder affects bovine hepatocytes, we performed untargeted metabolomics analysis on samples from the experimental and control groups. PCA revealed significant differences in metabolic profiles between the two groups, with good biological reproducibility within the groups (Figure 8A). The OPLS-DA score plots indicated clear separation between the groups in both positive and negative ion modes. The OPLS-DA model was validated using 200 permutation tests, yielding model quality parameters of R²Y = 0.996 and Q² = 0.882 for positive ion mode, and R²Y = 0.999 and Q² = 0.894 for negative ion mode, confirming the model’s high stability and lack of overfitting (Figures 8B,C). Differential analysis identified 259 significantly different metabolites, with 109 upregulated and 150 downregulated (Figures 8D,E). These differential metabolites were further subjected to cluster analysis (Figure 8F).

KEGG metabolic pathway enrichment analysis of the 259 differential metabolites was conducted to explore the biological pathways potentially affected by Guiqi Yimu Powder in bovine hepatocytes (Supplementary Table S4). The results indicated that metabolism-related pathways were primarily enriched in glutathione metabolism, fatty acid degradation, choline metabolism, amino acid biosynthesis, as well as the metabolism of glycine, serine, and threonine. Pathways related to inflammation and oxidative stress were mainly enriched in rheumatoid arthritis and various cancer pathways. Signal transduction pathways included key pathways such as the mTOR signaling pathway and the PI3K-Akt signaling pathway (p < 0.05, Figure 8G).

By integrating transcriptomics, proteomics, and metabolomics data, and analyzing the significantly enriched pathways (p < 0.05), we constructed a comprehensive regulatory network. As shown in Figure 9, this network revealed the multi-faceted impact of Guiqi Yimu Powder treatment on TMAO-induced metabolic pathways in bovine fatty liver cells. The results indicated that Guiqi Yimu Powder significantly affected signal transduction pathways, lipid metabolism pathways, and pathways related to inflammation and oxidative stress in bovine hepatocytes. Target genes identified through network pharmacology analysis were mainly concentrated in inflammation pathways, underscoring the critical roles these genes play in the pathological mechanisms. To further explore the interactions among key genes, proteins, and metabolites that showed significant changes upon Pearson correlation analysis was performed on Guiqi Yimu Powder treatment in bovine hepatocytes, and a correlation network diagram was subsequently constructed (Figure 10). The analysis revealed that L-5-Oxoproline showed a highly significant correlation with many key genes and proteins (p < 0.01), which were significantly enriched in signal transduction pathways, lipid metabolism pathways, and inflammation-related pathways (p < 0.05). The significant increase in L-5-Oxoproline represents a crucial metabolic alteration, highlighting its potential regulatory role in signal transduction, lipid metabolism, and the modulation of inflammation and oxidative stress.

On the other hand, the key genes and proteins significantly enriched in these pathways suggested that L-5-Oxoproline acted as a central node mediating the interaction and coordination of multiple metabolic pathways. For instance, significant associations were observed between L-5-Oxoproline and several related genes and proteins: lanosterol synthase (LSS) and squalene synthase (FDFT1), which are involved in lipid metabolism regulation; microsomal glutathione S-transferase 1 (MGST1) and X-linked inhibitor of apoptosis protein (XIAP), which are involved in oxidative stress regulation; interferon alpha/beta receptor 2 (IFNAR2), which is involved in inflammatory responses; and low-density lipoprotein receptor-related protein 6 (LRP6), which is important in energy metabolism. Investigating these targets further can provide deeper insights into the potential mechanisms of Guiqi Yimu Powder in treating bovine fatty liver. Finally, through comprehensive multi-omics data analysis, we elucidated the potential mechanisms by which Guiqi Yimu Powder affects bovine fatty liver cells. Figure 11 illustrated how Guiqi Yimu Powder regulated bovine fatty liver through multiple pathways. Specifically, in TMAO-induced fatty liver formation, oxidative stress and damage lead to glutathione depletion. Guiqi Yimu Powder regulated the development of bovine fatty liver by affecting glutathione metabolism and energy metabolism. By significantly upregulating metabolites such as L-5-Oxoproline, Glutamine, Gamma-Glutamylmethionine, Gamma-Glutamylglutamine, and Glycine, Guiqi Yimu Powder enhances glutathione synthesis, thereby improving cellular antioxidant capacity. Concurrently, L-5-Oxoproline converted to glutamate, which further converts to alpha-ketoglutarate (α-KG, an important intermediate in the TCA cycle) and succinate, promoting the efficiency of the TCA cycle and boosting energy metabolism levels. Additionally, the upregulation of SIRT1 played a crucial role in this mechanism. SIRT1 can upregulate the expression of multiple genes involved in glutathione metabolism, including glutamate-cysteine ligase (GCL), which is the rate-limiting enzyme in glutathione synthesis. SIRT1 also promotes the conversion of glutamate to α-KG, further enhancing the TCA cycle and energy metabolism. Furthermore, the upregulation of the target gene CDK2 was beneficial in this context. CDK2 was involved in cell cycle regulation and can aid in cellular repair and regeneration, contributing to the maintenance and restoration of tissue structural integrity. This integrated mechanism ultimately contributed to the establishment and maintenance of tissue structural integrity, improving the overall health of hepatic cells.