ANIT was obtained from Sigma-Aldarich Corporation (St. Louis, MO, United States). Trans-TSG, with a chemical purity of 99.8%, was bought from Shanghai Yuanye Bio-Technology Co., Ltd. (Shanghai, China). Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) kits were procured from Nanjing Jiancheng Bioengineering Institute (Jiangsu, China). High-performance liquid chromatography (HPLC)-grade methanol, acetonitrile, and isopropanol were supplied by Thermo Fisher Scientific Co., Ltd. (Fair Lawn, NJ, United States), while mass spectrometry (MS)-grade formic acid was acquired from Anaqua Chemicals Supply (Wilmington, DE, United States). Diethyl pyrocarbonate (DEPC) water was provided by Applied Biosystems, Ambion, Life Technologies (Austin, TX, United States). TRIzol total RNA extraction and RNA purification kits were purchased from Shanghai Meiji Biotechnology Co. Ltd., and QIAzol Lysis Reagent was obtained from Qiagen (Germany). Biowest Agarose for agarose gel detection was procured from Biowest (Spain), and Illumina^® stranded mRNA and NovaSeq Reagent Kit were purchased from Illumina (United States). Metabolite and bile acid standards were purchased from Sigma-Aldrich Corporation (St. Louis, MO, United States), and ultrapure water was obtained from Watsons Food and Beverage Co., Ltd. (Guangzhou, China). All primers for reverse transcription quantitative PCR (RT-qPCR) and polyclonal antibodies for western blotting were provided by Affinity Biosciences (Cincinnati, OH, United States).

Due to the instability of cis-TSG in its solid state, it is not commercially available and must be isolated from PM Root extract or obtained by converting its isomer, trans-TSG, using sunlight. In this study, we used trans-TSG previously isolated and purified in our laboratory. The trans-TSG was extracted from PM (voucher number PMR-20160432) purchased from Tongrentang (Beijing, China) and authenticated by Professor Zhang Lijuan of Tianjin University of Traditional Chinese Medicine. It was isolated from the n-butanol extract of a 60% ethanol extract of PM using multiple chromatographic techniques, combined with NMR and MS identification methods, and its structural and molecular weight were verified against literature data. To prepare a 1 mg/mL solution of trans-TSG, 2 g of trans-TSG stored at −80°C was dissolved in 2L of double-deionized water (ddH₂O) in a 4L transparent beaker. The solution was stirred continuously until fully dissolved and then exposed to sunlight for 1 h at room temperature. The production of cis-TSG was confirmed using an Agilent (DAD) HPLC system (Germany) with Waters COSMOSIL C₁₈ packed column (4.6 mm × 250 mm, 5 μm). Isocratic elution with methanol (A, 35%) and water (B, 65%) as mobile phase at a flow rate of 1 mL/min was employed, with a total run time of 6–7 min. Upon achieving 55%–70% conversion, the solution was concentrated using a rotary evaporator. The product was further purified by preparative liquid chromatography at a flow rate of 60 mL/min using the same mobile phase. The sample volume was maintained at 400 mg/mL of ddH₂O, and the detector wavelength was set at 320 nm to collect both cis-TSG and untransformed trans-TSG. Fractions were collected and dried immediately in a dark room to prevent light exposure, which destabilizes cis-TSG. Preparative liquid chromatography was conducted using a GILSON preparative liquid chromatography system (Gilson, Inc., United States), equipped with an NU3000 series UV detector (200–400 nm, dual-wavelength channel), an NP7000 series pump (0–100 mL/min, 10 Mpa), and an HPLCONE C₁₈ column (50 mm × 250 mm, 10 μm) (Suyan, Science and Technologies, Bangalore, India). The purity of cis-TSG was confirmed using Agilent analytical UHPLC, resulting in a total yield of 2.5 g of cis-TSG with 98.5% purity, stored at −80°C for further experiments.

The dosing rationale for stilbene glycosides is based on the maximum safe clinical dose of PM in adults, reported as 12 g/day (Lei et al., 2015). PM contains at least 1% stilbenes (approximately 2.6%) and about 0.2% emodin. According to the American human dose conversion guidelines, the clinical equivalent dose of TSG for mice is approximately 60 mg/kg, while that for emodin is 5 mg/kg (Lei et al., 2015). In this study, we administered a dose of 300 mg/kg, which is five times the clinical equivalent dose of TSG (Lei et al., 2015; Xu et al., 2017).

Male C57 BL6 mice (32 in total, weighing 20–25 g, SPF grade), aged 7–8 weeks, were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The mice were acclimatized for 1 week before the experiments. They were housed in an air-conditioned, specific pathogen-free facility under a controlled 12-h light/12-h dark cycle at 25 ± 2°C and 45 ± 5 humidity, with free access to chow and water. Body weights were recorded daily before drug administration. All animal studies were performed in compliance with the National Institutes of Health (NIH) and local guidelines at the Center for Laboratory Animals, Tianjin University of Traditional Chinese Medicine, Tianjin, China (Ethical number: TCM-LAEC202318m45631). The mice were randomly divided into four groups (n = 8). The treatment groups were as follows: (1) Group I: Control (CO) - administered only drug solvents. (2) Group II: Cis-TSG (CI) - administered only cis-TSG. (3) Group III: Model (MO) - administered ANIT on the 5th day. (4) Group IV: Cis-TSG + ANIT (CM) - administered cis-TSG for 7 days and ANIT on the 5th day. Drugs were dissolved in 0.5% sodium carboxymethyl cellulose (CMC-Na) solution and administered orally via gavage once daily for 7 days. Groups CI and CM received 300 mg/kg cis-TSG for seven consecutive days (Days 1–7), while groups CO and MO were administered 10 mL/kg CMC-Na via intragastric administration. On Day 5, 2 hours after the administration of CMC-Na or drugs, a single dose of ANIT (50 mg/kg, p. o. in corn oil) was given to MO and CM groups. The control group was given the vehicle orally. Forty-8 hours after ANIT treatment and 2 hours after the final drug administration on the 7th day, the mice were anesthetized with 65 mg/kg sodium pentobarbital and euthanized in an ultra-clean workbench to harvest all experimental tissues. Blood samples were collected by orbital bleeding into 1.5 mL heparinized tubes. Plasma was separated by centrifuging the blood at 13,200 × g for 10 min at 4°C. Following the collection of blood, the mice underwent dissection to obtain liver and distal ileum tissues. Subsequently, the liver tissue was washed, weighed, and divided into two parts for further analysis. One part was fixed in 4% paraformaldehyde at room temperature and stained with hematoxylin-eosin (H-E) for histopathological analysis. The other part was placed in pre-labeled cryogenic vials according to the mouse number in each group, snap-frozen in liquid nitrogen, and stored at −80°C until further analysis.

Plasma levels of ALT and AST were measured using an automatic spectrophotometry biochemical analyzer (TECAN, Switzerland) with commercially available kits from Nanjing Jiancheng Bioengineering Institute, following the manufacturer’s protocol. Histological analysis was conducted to assess liver tissue morphology using H-E staining. Liver tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned at a thickness of 6 μm. The sections were dehydrated using graded ethanol and xylene solutions. Nuclei were stained with 5% hematoxylin solution for 10 min, followed by rinsing with distilled water for 5 min. The samples were then incubated in 0.1% hydrochloric alcohol for 30 s and counter-stained with eosin solution for 2 min. After additional washing and dehydration, the H&E stained sections were mounted and imaged using a fluorescence microscope.

Blood samples from mice were centrifuged at 13,200 × g for 10 min at 4°C to obtain plasma. Pre-cooled acetonitrile (−20°C) was added to 100 μL of plasma, vortexed for 2–3 min, and centrifuged again for 15 min at 13,200 × g at 4°C to collect the supernatant. For liver tissue analysis, 100 mg of liver tissue was placed into a 2 mL grinding tube with 400 μL of pure water and 2-3 magnetic beads. The tubes were pre-cooled at −20°C for 20–30 min before homogenization at 60 Hz for 120 s. Next, 100 μL of the liver homogenate was transferred to a 1.5 mL centrifuge tube, and 400 μL of pre-cooled acetonitrile was added. The mixture was vortexed for 2–3 min, centrifuged at 13,200 × g for 15 min at 4°C, and the supernatant was collected. Then, in both plasma and liver sample preparation, the supernatant was dried under a liquid nitrogen flow and reconstituted with 100 μL of a 50% methanol-water mixture, vortexed, and centrifuged again at 13,200 × g for 20 min at 4°C. Approximately 70–80 μL of the supernatant was transferred to a new UHPLC vial for metabolomics analysis, and 2 μL of this supernatant was injected. A quality control (QC) sample, about 5–10 μL, was prepared using the same method and used for UHPLC-MS/MS metabolomic analysis. QC samples were injected at regular intervals throughout the analytical process to ensure repeatability (Lin et al., 2021). Data analysis included principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) to analyze differences between groups. Differential metabolites and variable importance in projection (VIP) scores were determined, followed by KEGG, network, and pathway analyses based on the metabolite comparison data. Bile acids and selected lipids were further analyzed using targeted metabolomics and lipidomics, respectively.

In this experiment, 100 μL aliquots of plasma previously stored at −80°C were transferred into 1.5 mL Eppendorf tubes. The samples were allowed to thaw at 4°C before analysis. Each sample was supplemented with 2 μg/mL of cholic-2,2,4,4-d4 acid (CA-d4) as an internal standard (IS). Following thorough mixing, the samples were placed on ice for 5 min, and 1 mL of pre-cooled acetonitrile (maintained at −20°C) was added to each 1.5 mL Eppendorf tube. The resultant sample-acetonitrile mixture was then agitated for 30 min and subsequently centrifuged at 13,200 g and 4°C for 10 min. After centrifugation, the resulting supernatant was carefully transferred to a sterile centrifuge tube and desiccated under a stream of liquid nitrogen while eliminating the protein debris. The desiccated samples were then reconstituted by adding 50 μL of a 1:1 aqueous-methanol solution and vortexed for 3–5 min, following which they were centrifuged for 20 min at 13,200 × g and 4°C. Subsequently, 50–70 μL of the resulting supernatant from each centrifuge tube was carefully transferred to a UHPLC vial for targeted metabolomic analysis of bile acids using UHPLC-MS/MS. During the UHPLC-MS/MS analysis, an injection volume of 5 μL was employed. The same procedures were also followed to prepare the ileum tissues.

For liver samples, 50 mg of liver tissue was weighed and put into a 2 mL grinding tube. After adding 250 μL (five-fold of the weight) of pre-cooled water, 2-3 magnetic beads were inserted into each grinding tube. The tubes were then put into the holes of a previously cooled (at −20°C for 20–30 min) homogenizer’s grinding tube holder. The holder is placed in its proper position and tightly tied. The liver tissues were homogenized at 60 Hz for 2 min, and 250 μL of the homogenate liquid was then transferred into a 1.5 mL centrifugal tube. In each Eppendorf tube, 10 μL of the internal standard (2 μg/μL IS, CA-d4) was added, and the mixture was thoroughly vortexed and placed on ice for 5 min. To each tube, 1.25 mL of ice-cooled acetonitrile containing 3% ammonia was added, and the mixture was vortexed for 1 h to ensure thorough mixing. The mixture was then centrifuged for 10 min at 4°C at 13,200 × g. The supernatant was transferred to a new Eppendorf tube, and the precipitate was discarded. After drying the supernatant on a liquid nitrogen flow, 50 μL of aqueous-methanol solution (1:1) was used to re-constitute the samples. Samples were then vortexed (well mixed) and centrifuged for 20 min at 4°C, 13,200 × g. The supernatant volume of 50–70 μL was transferred into a UHPLC-MS/MS vial, and a 5 μL sample of each was used for BA determination during the UHPLC-MS/MS run.

Plasma samples stored at −80°C were thawed at 4°C. A 100 μL aliquot of each plasma sample was transferred into a new Eppendorf tube, and 300 μL of pre-cooled isopropanol was added. The mixture was vortexed for 5 min to ensure even mixing, then incubated at −20°C for 1 h. Following incubation, the samples were centrifuged at 13,200 × g for 20 min at 4 °C, and the supernatant was collected for UPLC-MS analysis. The QC sample was prepared by combining 10 μL of each plasma sample, gently shaking, and then adding isopropanol in a volume three times that of the total plasma sample. To analyze liver tissue, a 50 mg sample of liver tissue was weighed and homogenized with five volumes of pure water and 2-3 magnetic beads using a homogenizer at 60 Hz for 2 min. Following this, a 100 μL aliquot of the resulting homogenate was transferred to a new tube, after which 300 μL of pre-cooled isopropyl alcohol was added. The mixture was vortexed for 5 min, incubated at −20°C for 1 h, and then centrifuged at 13,200 × g for 20 min at 4°C. The supernatant was subsequently transferred to an AB vial for UHPLC-MS/MS lipidomic analysis.

Mouse liver tissues were harvested and stored at −80°C. RNA was extracted from the liver tissues, and the quality of the extracted RNA was evaluated using the Agilent 2100 Bioanalyzer (Agilent Technologies, California, United States). Following this assessment, a cDNA library was constructed for transcriptome sequencing, and the analysis was performed as previously described (Kim et al., 2018). Data analysis included the following steps: differential gene expression was visualized using Venn diagrams and heatmaps. KEGG pathway analysis and Gene Ontology (GO) term analyses were conducted using Majorbio (https://www.majorbio.com) to identify enriched pathways and biological processes. Total mRNA was extracted from liver tissue using Trizol reagent, following the manufacturer’s guidelines. cDNA synthesis was later performed using primers and the Novoptotein reverse transcription kit, adhering to the provided protocol. Primers were designed and synthesized by Sangon Biotech. GAPDH was used as the internal control gene for normalization, and the relative mRNA expression of PPARα was assessed. For protein expression analysis of PPARα, 30 mg of liver tissue was weighed and processed following standard procedures. Each sample was homogenized in 300 μL of lysis buffer in a 2 mL grinding tube, crushed on ice, and then centrifuged for 10 min at 13,200 × g at 4°C. Protein concentration was determined using the bicinchoninic acid (BCA) method. Proteins were denatured by adding 4X loading buffer proportionally and heating at 100°C for 10 min using a digital double-heat biological dry bath (Thermo Scientific, Fisher, United States). The denatured proteins were stored at −80°C. Proteins were later separated by SDS-PAGE (10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis) and transferred to a polyvinylidene fluoride (PVDF) membrane. The PVDF membrane was blocked and then probed with primary antibodies targeting PPARα (1:1,000; proteintech, PPARA-Antibody-66826-1-lg) and GAPDH (1:5,000). After incubation with primary antibodies, the membrane was treated with HRP-conjugated secondary antibodies. Detection was performed using a chemiluminescence (ECL) kit (EVERBRIGHT, Suzhou, China), and intensity analysis was conducted using ImageJ software (NIH, Bethesda, United States).

Statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA, United States). Data are presented as means ± SD and visualized using bar charts. Statistical significance between the two groups was determined using Student’s t-test. For comparisons among multiple groups, One-way ANOVA was employed. A p-value of less than 0.05 was considered statistically significant.

To investigate whether cis-TSG, a type of stilbene compound in PM, could worsen liver injuries in patients with a history of liver problems who take PM for its therapeutic benefits, we induced a cholestasis model in mice using ANIT. The enzymes ALT and AST are key indicators of liver diseases. To assess whether cis-TSG protects the liver during cell damage and cholestasis, we measured levels of ALT and AST in the plasma. The results showed that, compared to the model group, plasma levels of ALT and AST (U/L) were significantly lower in the cis-TSG treatment (Figure 1E, p ≤ 0.0001), almost similar to their level in the control group. Furthermore, the group administered with cis-TSG only exhibited comparable levels of ALT and AST as the control group. This suggests that cis-TSG does not induce any discernible toxic effects on its own. Upon comparing the CI group with both the CM and MO groups, it becomes evident that the significant reduction of ALT and AST levels in the CM group is attributed solely to the administration of cis-TSG. In Figure 1B, the ANIT-induced group displayed distinct pathomorphological features characterized by visible inflammation indicative of cholestasis. However, treatment with cis-TSG in ANIT-induced mice markedly improved the pathological features, demonstrating similarity to the control group (Figure 1B). Further histological analysis with H&E staining revealed that cis-TSG treatment effectively prevented inflammatory cell infiltration and hepatocyte damage induced by ANIT (Figure 1C). In addition to our findings, we observed changes in bile color (Figure 1B), a characteristic feature of liver cholestasis, and plasma color (Figure 1D), which can indicate the effect of cis-TSG. In comparison to the control group, ANIT treatment resulted in significantly darker bile color, likely attributed to disruptions in bile acid homeostasis caused by ANIT treatment. The perturbed composition of bile acids was restored to near-physiological levels following cis-TSG treatment, as evidenced by the observed change in bile color in the cis-TSG treatment groups as compared to the control and model groups (Figure 1B). These data demonstrate that cis-TSG effectively reduced plasma ALT and AST levels, ameliorated liver injury, and impeded inflammatory responses by inhibiting inflammatory cell infiltration. This suggests that cis-TSG exerts a hepatoprotective effect in ANIT-induced cholestasis.

To investigate the metabolic profiles associated with cis-TSG treatment, data collected from UHPLC-MS in both positive and negative ion modes were further processed using Compound Discoverer 3.0 software and Metaboanalyst 6.0 (https://www.metaboanalyst.ca/). Multivariate statistical analysis was conducted, and the PCA plots for liver (Figure 2A) and plasma (Figure 2B) samples were generated. PCA score plots demonstrated distinct clustering of QC samples, indicating high instrument reproducibility and stability. Notable differentiation was observed between the CO and MO groups, while no clear separation was seen among the CO, CI, and CM groups. This suggests that cis-TSG confers hepatoprotection by altering the metabolite profile to resemble that of the control group, both in the liver and plasma. OPLS-DA score and volcano plots further confirmed the distinct separation between the control and model groups in the liver (Figures 2E, F) and plasma (Figures 2I, J). Similarly, clear differentiation was observed between the model and cis-TSG pretreated group in the liver (Figures 2C, D) and plasma (Figures 2G, H). These results show significant changes in the metabolic profile due to cis-TSG treatment. Potential metabolic biomarkers related to the effects of cis-TSG were screened using pairwise comparisons of OPLS-DA. Significant metabolites were identified based on a VIP threshold (VIP >1.0) and p-value (p < 0.05), comparing CO and MO and MO and CM groups. These significant metabolites were further identified by comparison with standards, internal databases (mzCloud, Chemical Spider, etc.), and online mass spectral databases (HMBD, METLIN, Pubchem) and validated via KEGG identification. The key hepatoprotective metabolites are summarized in Table 1. In the model group, LPCs and BAs were elevated in the liver and plasma, while these metabolites were reduced by cis-TSG treatment. Conversely, the levels of 5-OxoETE, 12,13-DHOME, and 9,10-DHOME were decreased in the MO group but increased in the CM group, showing enhanced fatty acid oxidation effect with cis-TSG treatment. Seven metabolites, including LPC (18:2), LPC (18:3), LPC (20:5), and taurocholic acid, were significantly higher in the plasma of the MO group compared to the CM group, with cis-TSG pre-treatment reversing their increase. Collectively, these results highlight the differential influence of cis-TSG on potential metabolic pathways, underscoring its hepatoprotective effects in ANIT-induced cholestasis.

To understand the metabolic pathways influenced by cis-TSG in alleviating cholestasis, we conducted a detailed metabolic pathway analysis comparing the model group with the cis-TSG-treated group. The analysis revealed significant associations with glycerophospholipid metabolism, sphingolipid metabolism, and primary bile acid biosynthesis pathways (Figures 3A, B). The sphingolipid metabolism pathway involves key metabolites such as sphingosine and sphingosine-1-phosphate. In the glycerophospholipid metabolism pathway, five metabolites, including choline and four lysophosphatidylcholines (LPCs), were identified. A notable number of fatty acids were involved in these pathways, highlighting the role of lipid and fatty acid metabolism in the hepatoprotective effects of cis-TSG (Table 1). Further investigation into these metabolic processes suggested that the modulation of lipid and fatty acid metabolism is central to the protective effects observed with cis-TSG treatment. Based on the plasma and liver metabolic profiling, a potential pathway of metabolic changes was outlined (Figure 3C). These findings underscore the importance of lipid (glycerophospholipid and sphingolipid) metabolism and bile acid biosynthesis in the therapeutic action of cis-TSG against cholestasis. By altering these metabolic pathways, cis-TSG appears to restore metabolic balance, thereby reducing liver damage and inflammation. Collectively, these results provide insight into how cis-TSG modulates critical metabolic pathways to exert its hepatoprotective effects in ANIT-induced cholestasis. The involvement of glycerophospholipid, sphingolipid, and bile acid biosynthesis pathways suggests that cis-TSG targets lipid and bile acid metabolism, reinforcing its potential as a therapeutic agent for liver diseases.

In a cholestatic mice model, it was found that ANIT induces a blockage of bile elimination from liver cells to the bile duct, leading to an increase in bile concentration within the liver (Zhang K. et al., 2021). To investigate the impact of cis-TSG on bile acid accumulation in the liver, an analysis of bile acid levels was conducted using a heatmap and t-test analysis in GraphPad. The heatmap (Figure 4A) as well as the bargraphs (Figure 4B) revealed that the model group exhibited the highest concentration of bile acids when compared to the CO, CI, and CM groups, indicating cholestatic liver toxicity within the model group. Additionally, the administration of cis-TSG significantly reduced the accumulation of bile acids in the liver, confirming its hepatoprotective effect against cholestatic liver injury. Furthermore, an analysis of several primary and secondary BAs revealed increased levels in the model group (Supplementary Figure S1). In comparison to the CO group, the levels of 6 primary BAs, including cholic acid (CA), HCA, taurocholic acid (TCA), beta-muricholic acid (β-MCA), tauro-(α/β)-muricholic acid (TMCA), and taurohyocholic acid/tauro-γ-muricholic acid (THCA), significantly increased, while the levels of alpha-muricholic acid (α-MCA) and chenodeoxycholic acid (CDCA) were notably decreased in the livers of cholestatic mice. Following treatment with cis-TSG, the levels of all 6 primary BAs reversed, displaying a significant downward trend. In the distal ileum tissue (Figure 4C)., concentration of bile acids were low in the model group which indicates their blockage and accumulation in the liver. However, cis-TSG restored normal flow of the bile acids and their concentration were comparable to the normal control in the ileum tissue. These results demonstrate that cis-TSG regulates and alleviates ANIT-induced liver injury by significantly reducing the concentration of bile acids in the liver.

Cholestatic liver disease disrupts the absorption and metabolism of lipids (Fu et al., 2019). To understand the impact of cis-TSG on lipid metabolism in cholestatic mice, we analyzed the concentrations of differential lipids using heatmap visualization and t-test analysis in GraphPad. Significant variations in lipid concentrations were observed between the control and model groups. Pre-treatment with cis-TSG reversed these changes, whereas the model group showed elevated levels of various lipids, indicating lipotoxicity. In liver tissues, the concentration of phosphatidic acid (PA) [PA (20:0/18:2) and PA (18:0/18:1)], a crucial precursor of most phospholipids, was significantly higher in the model group compared to the control group (Figure 5A). Treatment with cis-TSG completely reversed this elevation, restoring PA levels to those seen in the control group. Additionally, the concentrations of phosphatidylcholine (PC) and phosphatidylglycerol (PG) were significantly increased in the model group. Cis-TSG treatment effectively reversed these changes, normalizing PC and PG levels. The heatmap demonstrated a similar pattern, showing distinct differences in lipid concentrations between the study groups (Figure 5B). Phosphatidylethanolamines (PEs) such as P-18:0/18:1, O-16:0/18:1 and P-16:0/20:2 showed higher concentrations in the model compared to the control and cis-TSG-treated groups (Figures 5C, D). However, cis-TSG increased the concentrations of phosphatidylinositol (PI) and lysophosphatidylinositol (LPI) to levels similar to the control group (Figure 5A), indicating its role in restoring normal lipid metabolism. The levels of ceramides (CERs) and other damaging sphingolipids, such as dihydroceramides (DCERs), hexosylceramides (HCERs), and lactosylceramide (LCER (22:1), were found to be significantly elevated in the model group, consistent with previous studies (Choi et al., 2021). Furthermore, the total concentration of other sphingolipids, such as Sphingomyelins (SMs), also showed an increase in the model group (Figure 6A), indicating a dysregulation of sphingolipid metabolism in cholestatic liver injury. However, treatment with cis-TSG led to a reduction in the concentration of sphingolipids, thereby mitigating cholestasis-induced lipotoxicity and promoting liver recovery, as evidenced by improved liver health parameters. On the other hand, the amounts of triacylglycerols (TAGs) and diacylglycerols (DAGs) were significantly lower in the model group compared to the control, while these lipids were notably increased in the group treated with cis-TSG, bringing them more in line with the levels observed in the control group (Figure 6B). This may indicate a way to reduce the high concentrations of CERs by diverting them to the biosynthesis of FFAs and then safely storing them as DAGs and TAGs (Choi et al., 2021). These findings suggest that cis-TSG has a protective effect by adjusting lipid levels and reducing lipotoxicity in ANIT-induced cholestasis. These results underscore the potential therapeutic role of cis-TSG in improving cholestatic liver injury through the regulation of sphingolipid metabolism.

To evaluate the hepatoprotective effects of cis-TSG, we analyzed the concentrations of several monounsaturated fatty acids (MUFA) and polyunsaturated fatty acids (PUFA) in ANIT-induced cholestatic mice. As shown in Figure 6, the levels of these unsaturated fatty acids were significantly lowered in the model group compared to the control group. However, cis-TSG treatment significantly increased the concentrations of many MUFAs and PUFAs, confirming its protective effect against ANIT-induced cholestasis in mice. Additionally, the concentrations of various phosphatidylinositol (PI) and lysophosphatidylinositol (LPI) lipids were significantly reduced in the model group compared to the control. In the cis-TSG-treated group (CM), these lipid concentrations were almost completely restored to control levels, further indicating the hepatoprotective properties of cis-TSG. These findings collectively demonstrate that cis-TSG exerts a hepatoprotective effect by modulating the concentrations of various lipids, restoring normal lipid metabolism, and reducing lipotoxicity in ANIT-induced cholestasis.

To elucidate the protective mechanisms of cis-TSG in ANIT-induced cholestasis, transcriptomic analysis was performed. In the model group, 3,095 genes were significantly altered, with 1,492 genes upregulated and 1,603 genes downregulated after ANIT administration (Figure 7). In contrast, cis-TSG treatment significantly modulated these gene expressions towards normal levels observed in the control group. The gene coding for sphingomyelin phosphodiesterase 3 (Smpd3) was significantly upregulated in the model group compared to the control group. Treatment with cis-TSG markedly downregulated Smpd3 expression, restoring it to levels comparable to the control group (Figure 7C). Similarly, the gene coding for phosphoserine aminotransferase 1 (Psat1), a key enzyme in serine biosynthesis pathway, showed a similar expression pattern. These results align with previous findings that elevated ceramide synthesis contributes to liver injury (Morigny et al., 2020). Phospholipid phosphatase 2 (Plpp2), an enzyme involved in phosphatidic acid metabolism, was significantly upregulated in the model group. This upregulation was reversed by cis-TSG treatment, aligning with the lipidomics data that showed a reduction in phosphatidic acid (PA) levels in the cis-TSG-treated group (Li et al., 2006). These transcriptomic findings are consistent with metabolomics and lipidomics analyses, suggesting that cis-TSG exerts its hepatoprotective effects by modulating key genes involved in lipid metabolism. The KEGG pathway enrichment analysis further supported these results, highlighting the significant impact of lipid metabolism pathways in the protective effects of cis-TSG (Figure 7D). Overall, these results indicate that cis-TSG modulates critical genes involved in lipid and bile acid metabolism, providing a mechanistic basis for its hepatoprotective effects in ANIT-induced cholestasis.

Based on earlier findings, it is believed that PPARα increases the production of MUFA by influencing the mRNA levels of genes responsible for fatty acid oxidation, synthesis, transport, and the formation of triacylglycerol (TAGs) (Tian et al., 2020). As a result, we theorized that the higher concentrations of MUFAs and TAGs in the group treated with cis-TSG, compared to the model group, could be attributed to the further activation of the PPARα pathway by cis-TSG. To test our hypothesis, we investigated the mRNA and protein expression levels of PPARα. As presented in Figure 8, RT-qPCR and western blot analyses confirmed a significant reduction in the mRNA and protein expression levels of PPARα in liver cells of the model group compared to the control group. However, treatment with cis-TSG notably reversed this reduction, restoring PPARα expression to levels comparable to the control group. These findings support our hypothesis that cis-TSG mitigates liver lipotoxicity through FFA-induced activation of PPARα. This activation likely contributes to the observed hepatoprotective effects in the cis-TSG-treated group.