PPL was purchased from Shuguang Hospital Affiliated to Shanghai University of Traditional Chinese Medicine (Shanghai, China). Voucher samples were deposited in the Central Laboratory of Shuguang Hospital. Serum biochemical indexes were measured using a full-automatic biochemical analyzer. MDA (Elabscience, E-BC-K025-M, China) and SOD (Elabscience, E-BC-K022-M, China) kits were purchased from Elabscience (Wuhan, China). Liver TC (APPLYGEN, E1026-105, China) and TG (APPLYGEN, E1013-50, China) levels were determined using APPLYGEN assay kits (Beijing, China). Primary antibodies included rabbit anti-PPARγ (CST, #2435, United States of America), rabbit anti-PPARα (Abclonal, A6697, China), mouse anti-CPT1A (Abcam, ab128568, United Kingdom), rabbit anti-CPT2 (Abcam, 181,114, United Kingdom), rabbit anti-SREBP2 (Abcam, ab30682, United Kingdom), mouse anti-HMGCR (Abcam, CL0260, United Kingdom), rabbit anti-CYP7A1 (Abcam, ab234982, United Kingdom), and actin (ABGENT, AP14779b, United States of America).

To obtain an extract of PPL, 200 g PPL was soaked in 2 L distilled water at room temperature (RT) for 1 h and then heat to reflux for another 1 h. The extract was filtered, and the residue was subjected to a secondary extraction. The extracted mixture was concentrated under reduced pressure and further lyophilized to a fine powder to obtain 20.52 g lyophilized powder with a final yield of 10.26%.

Male C57BL/6 mice were purchased from Jiesijie Research Institute (Shanghai, China) and divided into two groups. Mice in Group 1 were orally administered with PPL extract (400 mg/kg, equivalent to the clinical adult dose), and blood was collected from the orbit 1 h after administration. Mice in Group 2 were given saline and the serum was collected as a control group. Both serum samples were inactivated by heating at 56°C for 30 min. The protein precipitation was removed by adding twice the volume of acetonitrile (Aladdin, A298777, China) and the supernatant was transferred into a rotary evaporator (Yarong, RE-3000A, China) for concentration. The pure water was re-dissolved and desalted by a LC-C18 solid phase extraction column (CNW, SBQE-CA0955, China), and transferred into a vacuum centrifugal concentrator (SCIEMTZ, SCIENTZ-ILS, China). The concentrate was weighed, filtered, and sterilized by re-dissolving in pure water and stored at −80 °C until subsequent experiments.

Total proteins were extracted from mice liver tissues sampled from HFD and HFD + PPL groups (n = 3) with 300 µL lysis buffer supplemented with 1 mM Phenylmethanesulfonyl fluoride (PMSF). Samples were centrifuged at 15,000 g for 15 min to remove insoluble particles, and then the precipitate was again excluded. After adding 5 times the volume of cold acetone and precipitating overnight at −20°C, the precipitate was collected by centrifugation at 12,000 g for 15 min at 4 °C. The precipitation was dried at RT for 3 min and then dissolved in SDS-lysis buffer for 2 h. Sample supernatant was collected and centrifuged again to remove the precipitate. The protein concentration was determined according to BCA method and samples were aliquoted to be stored at −80 °C. Base on BCA assay results, 100 μg of protein was extracted for filter-aided sample preparation (FASP) digestion. IAA was added to a final concentration of 50 mM and incubated at RT for 40 min in a dark place, followed by another 1 h incubation at 60 °C. The solution was discarded after centrifugation on a filter at 4 °C for 25 min. After adding 100 µl of 300 mM triethyl ammonium bicarbonate (TEAB), the solution was centrifuged twice for 20 min each time. After washing, TEAB and trypsin were added to the filtration unit and digested overnight for 12 h at 37°C. Finally, EAB was eluted and lyophilized. The raw data was imported into Maxquant for label-free quantitative analysis with a search engine Andromeda.

Mass spectrometric analysis was performed on a Q Exactive HFX mass spectrometer using a UPLC system with a BEH Amide column (2.1 mm * 100 mm, 1.7 mm). 25 mmol/L acetate of ammonium and 25 mmol/L hydroxide of ammonium in water (pH = 9.85) A) and acetonitrile B) were respectively used. The auto-sampler was set to 4 °C and the injection volume was 3 L. To obtain MS/MS spectra in information-dependent acquisition mode (IDA), a QE HFX mass spectrometer was utilized. In IDA mode, the acquisition software performed continuous evaluation of the full scan MS spectrum. Accordingly, the electrospray lonization (ESI) source conditions were as follows: 15 Arb sheath gas flow rate, 15 Arb aux gas flow rate, 350 °C capillary temperature, 60,000 full MS/MS resolution, 10/15/60 collision energy in normalized collisional energy (NCE) mode, and 3.6 (positive) or −3.2 (negative) spray voltage. The raw data was converted from XML to mzXML using ProteoWizard, and then the peak data was extracted in R using an in-house program developed by XCMS to align and integrate it. A cut-off of 0.3 was determined in metabolite annotation using an internal MS2 database.

Male C57BL/6 mice were purchased from Jiesijie Research Institute (Shanghai, China). Six-week-old male mice (weight: 20 ± 2 g) were maintained in a standard environment at 22°C–24°C with a 12:12 h light-dark cycle, humidity (65% ± 5%) and ad libitum access to standardized food and tip-filtered water. Male C57BL/6 mice were given HFD (Diets, HF60, China) for 13 weeks to establish a NAFLD pathological model. Mice treated with normal chow diet (NCD) were defined as the control group. Mice fed with HFD + PPL (100, 200, and 400 mg/kg) were defined as the treatment group. Mice treated with physiological saline and PPL (400 mg/kg) were used to verify the toxicity of PPL. All experimental procedures were authorized according to the standards in the Guide for the Care and Use of Laboratory Animals. All animal experiments were approved by the Institutional Animal Care and Use Committee of Shuguang Hospital, Shanghai Chinese traditional medicine University.

Thirty-eight C57BL/6 male mice were randomly divided into 6 groups: 1) Normal chow diet (NCD) group (n = 6); 2) NCD + PPL (400 mg/kg) group (n = 6); 3) High fat diet (HFD) group (n = 7); 4) HFD + PPL (100 mg/kg) group (n = 6); 5) HFD + PPL (200 mg/kg) group (n = 6); 6) HFD + PPL (400 mg/kg) group (n = 7). PPL was dissolved in saline. Both NCD and HFD groups were simultaneously given saline during PPL treatment in HFD + PPL group. Mice were continuously fed HFD for 13 weeks to induce NAFLD model and treated with PPL or saline via oral gavage for another 6 weeks. Mice were fasted overnight, drank water freely, and anesthetized prior to specimen collection. The whole body, liver, and adipose tissues were sampled from each group of mice, weighed, and photographed for documentation. Specimens were divided into three parts, and either frozen in liquid nitrogen for proteomic studies, or stored at −80°C for biochemical analysis, or fixed in paraformaldehyde (Biosharp, BL539A, China) for further morphological and histological analysis of the liver.

Blood specimens were allowed to stand at RT for 30 min and then centrifuged at 4°C for 15 min at 3,000 rpm to obtain serum samples. High-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), serum triglycerides (TG), total cholesterols (TC), alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were determined using kits (Nanjing Jian Cheng Bioengineering Institute, China) according to the manufacturer’s instructions.

Mice liver samples were fixed with 4% paraformaldehyde. White adipose tissues (WAT) and brown adipose tissue (BAT) were fixed with adipose tissue-specific fixed fluid. The liver and adipose tissue were sectioned into 5-µm sections and stained with hematoxylin and eosin (HE) solutions according to standard procedures. Slides were observed under an inverted microscope (Nikon, Ts2R-FL, Japan). Adipocyte diameter and area of white and brown adipocytes were measured using ImageJ software (National Institutes of Health, NIH, United States of America).

For liver sections staining, frozen liver sections were performed at optimal cutting temperature, rewarmed at RT, rinsed in distilled water, and stained with Oil Red O stain solution (Sigma, O9755, United States of America) for 6–10 min. After staining, sections were soaked in 60% isopropyl alcohol and washed with distilled water. Sections were sealed with glycerin gelatin. For cell staining, cells were harvested and fixed with 4% paraformaldehyde, washed twice, oil red O stained for 6–10 min, examined for lipid droplets. After that, samples were rinsed in 60% isopropyl alcohol and washed three times with ddH₂O. Finally, slides were observed under an inverted microscope.

For mice liver staining, liver tissue sections were frozen for 30 min and immersed in PBS for 10 min, and then soaked in BODIPY 493/503 (GLPBIO, GC42959, United States of America) for 30 min, protecting from light during staining. After washing with PBS three times, slides were counter-stained with DAPI (Beyotime Biotechnology, P0131, China). For cell staining, hepatocytes were rinsed three times with PBS and fixed in 4% paraformaldehyde for 15 min. After washing, samples were stained with BODIPY 493/503 for 30 min, and then counter-stained with DAPI for 5 min. Images were observed under an inverted microscope and fluorescence was quantified using an ImageJ software.

For glucose tolerance test (GTT), mice were fasted overnight for 12 h (20:00–08:00) and drank water ad libitum. Fasting blood glucose levels (0 min) were measured before intraperitoneal injection of glucose (1 gkg⁻¹ body weight), and blood was taken from the tail tip to record blood glucose levels (0, 15, 30, 60, and 120 min). The insulin tolerance test (ITT) was performed 3 days after the GTT test. Mice were given a 6 h fast (07:00–13:00) with free access to water, followed by an intraperitoneal injection of insulin (0.75 gkg⁻¹ body weight). Blood glucose concentrations (0, 15, 30, 60, 90, and 120 min) were recorded from the tail tip.

Mouse primary hepatocytes were isolated from C57BL/6J male mice according to a two-step perfusion method as previously described (Huang et al., 2022). Briefly, cells were maintained in DMEM (Gibco, C11885500CP, United States of America) supplemented with 10% FBS (Gibco, 10100147, United States of America) plus 1% penicillin/streptomycin (Gibco, 10378016, United States of America). The immortalized normal mouse hepatocyte cell line (Alpha mouse liver 12, AML12) cell line, was inoculated in 6-well plates (2 × 10⁶ cells per well) at 37°C in a 5% CO₂ atmosphere in DMEM/F12 medium (Meilunbio, MA0214, China) supplemented with 1% penicillin/streptomycin plus 10% FBS. The following day, both hepatocytes were washed twice and replaced with new culture medium. Blank serum metabolites, PPL metabolites (5.10 μg/ml) and 50 μM PA plus 200 μM OA were added 24 h before cell harvesting. Cells were subsequently collected for further analysis.

AML12 cells and mice primary hepatocytes were inoculated in 6-well plates (2×10⁶ cells per well). After overnight incubation, cells were co-incubated with PPL metabolites (5, 10 μg/ml) and PA/OA for another 24 h. Finally, cells were collected and homogenized in lysis-buffer to determine TG content according to instructions of APPLYGEN Assay Kit.

Cells were inoculated in a 96-well plate (5.0×10³ cells/well). Cell viability was determined using a CCK-8 assay kit (Meilunbio, MA0218-1, China). Each well was incubated with 100 μl of medium and 10 μl of CCK-8 reagent for no more than 4 h. The absorbance of each well was determined at 450 nm.

Total RNA was extracted from mice liver tissues or cells using TRIzol reagent (Vazyme, R401-01, China). A total of 2 μg RNA was reversely transcribed using a HiScript^® II Q RT SuperMix cDNA synthesis kit (Vazyme, R233-01, China). RT-qPCR was performed using a ChamQTM Universal SYBR^® qPCR Master Mix (Vazyme, Q711-02, China). The relative expression levels of the genes were determined by SYBR-green-based RT-qPCR. The results were analyzed using the 2^−ΔΔCT method and normalized toβ-actin values. Primer sequences are listed in Supplementary Table S1.

Total proteins were extracted from mice liver tissues or cells homogenized in RIPA Lysis Buffer (Beyotime, P0013B, China) supplanted with a protease inhibitor cocktail (MCE, HY-K0010, United States of America). BCA Protein Assay Reagent (Thermo Fisher Scientific, 23,225, United States of America) was used for protein concentration quantification. Approximately 40 μg of total protein samples were loaded to 10% SDS-PAGE, followed by an electrophoretic transfer to a nitrocellulose membrane. After that, the membrane was blocked with 5% BSA (Sangon biotech, A500023, China) for at RT for 1 h, and incubated at 4°C overnight against primary antibodies. After washing, membranes were incubated with secondary antibodies for 1 h. Finally, an Enhanced chemiluminescence Western blot kit (UUBIO, U10012, China) was used to visualize the target protein.

The pathological and morphological changes of tissue sections were assessed at ×200 magnification. The histopathological grading criteria scores of liver tissue sections include periportal-bridging necrosis, intralobular degeneration, portal inflammation, and fibrosis (Zechini B et al., 2004). The histopathological grading criteria scores of kidney tissue sections include tubular vacuolar degeneration/necrosis, tubular casts, and congestion (Gao et al., 2016). The histopathological grading criteria scores of heart tissue sections include myofibrils degeneration, oedema, subendocardial hemorrhage and the distribution of myocardial damage (Liu et al., 2021).

Data were expressed as the mean ± standard error of the mean (SEM). Statistical analysis was performed using GraphPad Prism software (v 8.4.2). Data were compared between groups using Student’s t-tests and among groups using one-way analysis of variance (ANOVA) tests. p < 0.05 was considered statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001). If a statistically significant change was found (p < 0.05), a post hoc comparison was performed using Bonferroni correction.

The results showed that PPL treatment significantly reversed body weight gain (Figure 1A), but no significant difference was observed in food consumption (Figure 1B). Moreover, PPL significantly reduced both liver weight (Figure 1D) and liver-to-body weight ratio (LW/BW) (Figure 1C). Long-term HFD resulted in liver dysfunction tests, as evidenced by elevated serum ALT and AST (Figure 1E), which were mitigated by PPL administration for 6 weeks. In addition, HE staining and liver histological assessment showed that PPL treatment significantly reversed the HFD-induced hepatocyte ballooning, hepatic steatosis and injury in NAFLD mice (Figure 1J). BODIPY staining further confirmed HFD-induced lipid hyper-accumulation in hepatocytes, which was reversed by PPL administration (Figure 1L). Accordingly, as indicated in Figure 1M-N, hepatic TC and TG levels were significantly elevated in HFD-fed vehicle group, while PPL treatment effectively reduced hepatic lipid content (TC, TG). Consistent with the results of hepatic lipid content measurements, PPL treatment significantly downregulated serum lipid levels in HFD-induced mice (Figures 1F–I). In addition, no significant drug toxicity was observed during PPL (400 mg/kg) treatment based on pathological assessment of the heart, liver, and kidney tissue sections (Figure 1O). Collectively, these results suggested that PPL exerted hypolipidemic activity to prevent weight gain and hinder hepatic steatosis, thereby ameliorating HFD-induced liver damage.

Due to the essential role of adipocytes in the regulation of body weight and energy metabolism, increased adipocyte size caused by lipid accumulation inevitably leads to alterations in functions related to energy metabolism in different types of adipocytes. Therefore, in the present study, morphological analysis of WAT and BAT were performed in each group of mice to investigate whether the improvement in metabolic symptoms was partly attributed to a reduction in fat mass. According to WAT and BAT size measurement by HE staining, HFD-induced mice showed increased size and weight of WAT and BAT compared to NCD group, whereas PPL supplementation significantly downregulated fat mass in HFD-induced mice (Supplementary Figure S3A, S3B, S3E, S3F). Furthermore, PPL significantly suppressed gene expression levels of inflammatory factors such as IL-6 and MCP-1. Accordingly, PPL significantly increased gene expression levels of browning markers such as Tmem26, CD137 and Cited1 (Supplementary Figure S3D). In addition, PPL significantly increased the expression of thermogenic markers (PGC1α) and significantly inhibited PPARγ activation, while simultaneously promoting PPARα expression in BAT (Supplementary Figure S3H). GLUT4 is a major insulin-responsive glucose transporter, which is downregulated in adipose tissues in a state of insulin resistance. Here, we found that PPL significantly promoted GLUT4 expression in both WAT and BAT (Supplementary Figure S3D, S3H). Collectively, the effect of PPL in improving glycolipid homeostasis could be partly attributed to improved adipose tissue function.

We next investigated the effects of PPL on the regulation of systemic glucose metabolism and oxidative stress. Levels of fasting blood glucose, fasting insulin, ITTs and GTTs were measured. The results showed that PPL treatment significantly reduced HFD feeding induced higher glucose levels. Accordingly, levels of fasting blood glucose and fasting insulin (Figures 2C,D) were suppressed in PPL-fed mice compared to those detected from HFD-fed mice. In the GTT experiment, mice in the PPL200 mg/kg dose group had significantly lower blood glucose values at 0, 30, 60, and 120 min than those in the HFD-fed vehicle group. Additionally, the blood glucose levels of mice in the PPL400 mg/kg dose group at 0, 15, 30, 60, 90, and 120 min were significantly lower than those in the HFD-fed vehicle group (Figure 2A). In the ITT experiment, mice in the 200 mg/kg dose group showed significant reduction in blood glucose levels at 0, 15, 60, 90 and 120 min compared to the HFD-fed vehicle group, and mice in the 400 mg/kg dose group had significantly lower blood glucose values at 0, 15, 30, 60, 90 and 120 min than those in the HFD-fed vehicle group (Figure 2B), suggesting that PPL restored glucose homeostasis and insulin sensitivity in NAFLD mice. Next, we investigated whether PPL treatment could regulate MDA and SOD levels in NAFLD mice. The result showed that PPL significantly reduced MDA and increased SOD activity in serum and liver of NAFLD mice (Figures 2F–I). These data suggested that PPL significantly ameliorated the insulin resistance and subsequent oxidative stress in HFD-induced NAFLD mice.

We further used a label-free quantitative proteomic profiling to identify the role of PPL at the hepatic tissue proteomic level to explore the key protein and molecular mechanisms underlying PPL protection against NAFLD. The protein expression profiles of liver tissues from mice in HFD and HFD + PPL (400 mg/kg) groups were subjected to by proteomic analysis, and a total of 2,693 plausible proteins were identified. A total of 249 differentially expressed proteins (DEPs) were identified by t-test (p < 0.05) in HFD and HFD + PPL groups, according to ratio analysis of protein expression levels. Of these, 191 were upregulated DEPs (FC ≥ 1.5) and 58 were downregulated DEPs (FC ≤ 0.667) (Figure 3B). KEGG pathway and GO analysis were performed using a DAVID Functional Annotation Tool. GO analysis was categorized into “biological process” (BP), “cellular compartment” (CC), and “molecular function” (MF) (Figures 3D–F). Metabolic pathways, protein processing in endoplasmic reticulum, as well as fatty acid metabolism were mostly enriched in KEGG pathways (Figure 3C), with PPAR been identified as a definitive pathway (Figure 4A). In addition, among the top 20 enrichment analysis in BP, the most affected was FAO, where oxidoreductase activity and mitochondria were identified as the most affected molecular functions and cellular components, respectively. These results suggested that PPL treatment contributed to the improvement of metabolism, fatty acid beta-oxidation, oxidoreductase activity and mitochondrial function in the HFD-fed NAFLD mice.

As previously mentioned, the most important modulator of the BP category, namely, FAO, provided us with a clear direction to further explore the potential hypolipidemic efficacy of PPL. Proteomics techniques were used to analyze the effects of PPL administration on pathway changes in liver tissues of HFD-fed mice. Total three fatty acid metabolism-related proteomes were identified, including “PPAR signaling pathway”, “Fatty acid metabolism” and “Fatty acid degradation” (Figures 4A–C). These three KEGG pathways were intercrossed FAO-associated proteins. Venn diagram analysis results showed that a pair of isozymes CPT1A and CPT2 located in the mitochondrial membrane responsible for the regulation of FAO, were simultaneously involved in the four enrichments analysis described above (Figure 4E). Based on the analysis of liver-specific proteins, we found that PPL treatment significantly inhibited the expression of the transcription factor PPARγ while promoting PPARα expression compared to the HFD group. Furthermore, in assessing the biological function of FAO in liver tissue, we found that PPL treatment significantly promoted the expression of CPT1A and CPT2, compared to the HFD group (Figure 4F). As indicated in Figures 4B–D, due to the elevated protein expression of ACADVL, we next examined the extent of transcription of multiple acyl-CoA dehydrogenases, including Acyl-CoA dehydrogenase very long chain (ACADVL), Acyl-CoA dehydrogenase long chain (ACADL), Acyl-CoA dehydrogenase medium chain (ACADM), Acyl-CoA dehydrogenase short chain (ACADS), Acyl-CoA dehydrogenase short/branched chain (ACADSB), which were responsible for facilitating the FAO dehydrogenation reaction. As shown in Figure 4H, PPL significantly promoted the transcriptional activities of ACADVL, ACADL, ACADM, ACADS and ACADSB, compared to those detected from the model group. In terms of glucose metabolism, PPL reversed the expression of gluconeogenesis-related genes (such as G6PC and PCK1) while increased the level of glucose transporter gene (Glut2). In addition, PPL inhibited the expression of cholesterol synthesis-related proteins (SREBP2 and HMGCR) and increased the expression level of cholesterol metabolism-related protein (CYP7A1) (Supplementary Figure S4). Taken together, these results confirmed that PPL promoted hepatic expression of signals linked with mitochondrial FAO in obese mice.

In vitro analysis of cellular steatosis characterized by lipid accumulation was performed in AML12 cells by incubating cells with OA plus PA. The CCK-8 assay results showed no cytotoxic effects of PPL on AML12 cells at concentrations below 100 μg/ml (Figure 5B). Therefore, 5 μg/ml and 10 μg/ml of PPL metabolites were selected for subsequent in vitro experiments. AML12 cells were cultured in medium containing PA/OA. The Oil Red O staining results showed significant lipid deposition, whereas PPL metabolites attenuated lipid accumulation in a dose-dependent manner (Figures 5A,C). Consistent with the results described above, intracellular TG levels were significantly increased after PA/OA co-stimulation, while PPL metabolites dose-dependently downregulated intracellular TG levels (Figure 5D). Furthermore, we found that PPARγ was activated 24 h post treatment of AML12 cells with PA/OA. In contrast, PPL metabolites treatment inhibited PPARγ activation and simultaneously promoted PPARα expression. Like the results from in vivo assay, the downstream of transcription factor PPARα, CPT1A and CPT2 were significantly upregulated (Figure 5E). These results suggested that alterations in lipid content induced by PPL metabolites treatment in AML12 cell were mediated via an effect on FAO.

Similarly, we examined the hypolipidemic effect of PPL metabolites on mice primary hepatocytes. CCK-8 analysis results (Figure 6B) indicated that PPL metabolites (5, 10, 50 μg/mL) showed no significant effect on the cell viability of mice primary hepatocytes. Accordingly, 5 μg/ml and 10 μg/ml were selected as the optimum dosing concentrations for subsequent experiments. As indicated in Figures 6A–C, PPL metabolites (5, 10 μg/ml) significantly reversed PA/OA-induced cellular lipid accumulation, which was consistent with the results in Figure 6D that PPL (5, 10 μg/ml) reduced the PA/OA-induced elevated cellular TG levels in mice primary hepatocytes. To further investigate the effect of PPL metabolites on FAO and to identify potential molecular functions involved, we tested whether PPL metabolites significantly reversed the activation of PPARγ, a transcription factor responsible for fat storage. Meanwhile, PPL metabolites significantly promoted the activity of PPARα, a nuclear receptor responsible for lipolysis. Compared to the model group, downstream proteins CPT1A and CPT2 were also upregulated (Figure 6E). These results demonstrated that PPL metabolites alleviated lipid accumulation through affecting FAO in mice primary hepatocytes.

ETO is a fatty acid oxidation inhibitor that inhibits the activity of CPT1A enzyme through binding to its catalytic site and thus has been widely used to validate the molecular function of the hypolipidemic activity of PPL metabolites in vitro. The CCK-8 assay showed that PPL metabolites (5, 10 μg/ml) + ETO (50 μM) did not affect the cell viability of hepatocytes (Figures 7B, F). Accordingly, PPL (10 μg/ml) + ETO (50 μM) was used for the following experiments. To assess lipid content more accurately, an alternative lipid dye BODIPY 493/503, which has a higher sensitivity for lipid droplets, was applied. The BODIPY staining demonstrated that PPL metabolites attenuated lipid droplets in a dose-dependent manner, whereas ETO (50 μM) reversed lipid-lowering phenotype of PPL metabolites (10 μg/ml) compared to PA/OA-treated AML12 cells (Figure 7A). Consistent with the results of fluorescence staining of AML12 cells, the anti-lipid capacity of PPL (10 μg/ml) was blocked by co-incubation with ETO in OA/PA-treated mice primary hepatocytes (Figures 7E,G,H). These results suggested that PPL metabolites blocked lipid accumulation through affecting FAO in AML12 cells and mice primary hepatocytes.

To explore the material basis of PPL for ameliorating HFD-induced NAFLD, UPLC/QE-HFX analysis was used. The results showed that total 29 compounds were identified from PPL metabolites (Supplementary Figure S1 and S2). Of these, 12 compounds (Quercetin, Kaempferol, Genistein, Baicalin, Phloretin, Hydroxygenkwanin, Baicalein, Galangin, Apigenin, Apigenin 7-O-glucuronide, M1 and M2) were flavonoids; 1 compound (Emodin) was quinones, four compounds (4-Hydroxybenzoic acid, 4-hydroxybenzaldehyde, Para-cresol, and Protocatechualdehyde) were phenols, threecompounds (Chlorogenic acid, Praeruptorin A and Demethylwedelolactone) were phenylpropanoids, four compounds (3-O-Acetyldiosgenin, Cholic acid, Ecliptasaponin A and Deoxycholic Acid) were terpenoids, one compound (Abscisic acid) was sesquiterpenoids, two compounds (Guanine and Nicotinic acid) were alkaloids, two compounds (Isatin and Fumaric acid) were derivatives, and two compounds (Melibiose and Isobutyl 4-hydroxybenzoate) were others (Supplementary Table S1 and Supplementary Table S2). Taken together, it was suggested that the main component of PPL metabolites were flavonoids, which might play a major pharmacological activity in anti-NAFLD efficacy of PPL.