Kaempferol (KPF) (catalog no. K812226) was purchased from Macklin (Shanghai, China). Assay kits for alanine aminotransferase (ALT; catalog no. C009-1-1), aspartate aminotransferase (AST; catalog no. C010-1-1), triglyceride (TG; catalog no. A110-1-1), total cholesterol (TC; catalog no. A111-1-1), high-density lipoprotein cholesterol (HDL-C; catalog no. A112-1-1), low-density lipoprotein cholesterol (LDL-C; catalog no. A113-1-1), malondialdehyde (MDA; catalog no. A003-1-2), superoxide dismutase (SOD; catalog no. A001-3-2), and glutathione peroxidase (GSH-Px; catalog no. A005-1-2) were obtained from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The Oil Red O staining kit was sourced from Beyotime (Shanghai, China). Primary antibodies included anti-HO-1 (ab305290, Abcam, USA), anti-Nrf2 (SC-365949, Santa Cruz, USA), and HRP-conjugated β-Actin (HRP-66009, Proteintech, Wuhan, China). The secondary antibody, HRP-conjugated Goat anti-rabbit IgG, was purchased from Abcam (ab288151).

Male C57BL/6 mice (6–8 weeks old, specific pathogen-free) were purchased from Shanghai Slack Laboratory Animal Co., Ltd. (Production License SCXK(HU)2022-0004) and housed under controlled conditions (temperature 18–26°C, relative humidity 40–60%), with a normal light-dark cycle and free access to food and water. To adapt the animals, all mice were first given a Lieber-DeCarli-based liquid diet for one week, during which the last three days involved a standard liquid diet without ethanol. These mice were chosen because the C57BL/6 strain reliably develops pronounced steatosis, oxidative damage, and inflammatory responses under Lieber-DeCarli or NIAAA chronic-plus-binge ethanol feeding, making it a standard model for experimental AFLD (5, 12).

After adaptation, the mice were randomly divided into three groups (n=12 per group). The CTRL group was maintained on the standard liquid diet and received daily vehicle gavage. The AFLD group received an ethanol-containing liquid diet plus vehicle gavage. The KPF group received the same ethanol-containing liquid diet supplemented with kaempferol solution (4 mg/mL, 0.1 mL/20 g body weight) by gavage. This dosage was selected based on prior studies demonstrating hepatoprotective effects against alcoholic liver injury in mice without evident toxicity (8, 12), and was further validated through our own preliminary experiments confirming biological efficacy and safety at this dose. The Lieber-DeCarli liquid diet was prepared according to the manufacturer’s instructions (Dyets, USA). For example, the 1% (v/v) ethanol formula per 1 kg diet contained 132.18 g L10016A base powder, 71.68 g maltodextrin, 10 g anhydrous ethanol, and 786.14 g deionized water. Diets with 2%, 3%, and 4% ethanol were produced by proportionally increasing the ethanol volume (20, 30, and 40 g, respectively) while reducing water to maintain equal total volume and identical nutrient composition. During the feeding period, ethanol concentration in the liquid diet was gradually increased, starting at 1% on Day 1, 2% on Day 3, 3% on Day 4, and 4% from Day 6 to Day 15. On Day 16, to simulate binge drinking (13), the AFLD and KPF groups received a single high dose of ethanol (5 g/kg, 31.5% v/v) by gavage, while the CTRL group received an equal volume of maltodextrin. Twenty-four hours later, the mice were fasted, anesthetized with tribromoethanol, and blood was collected via cardiac puncture. Plasma and liver samples were harvested and stored at −80°C for further analyses.

All procedures were reviewed and approved by the Experimental Animal Ethics Committee of Xiamen Medical College (Approval Number 20240207015, February 7, 2024) and were conducted in compliance with the ARRIVE guidelines to ensure animal welfare.

Plasma levels of ALT and AST were measured to evaluate liver injury. Lipid metabolism was assessed by quantifying plasma levels of TG, TC, HDL-C, and LDL-C. Oxidative stress markers, including MDA, SOD, and GSH-Px, were analyzed in liver tissues following manufacturer protocols.

Liver tissues were fixed in 4% paraformaldehyde overnight, embedded in paraffin, and sectioned into 4 μm slices for hematoxylin and eosin (H&E) staining. Frozen liver tissues were sectioned into 8 μm slices and stained with Oil Red O to evaluate lipid accumulation. Histological imaging was performed using an optical microscope (Leica, Wetzlar, Germany). All slides were coded and independently evaluated under double-blind conditions by two board-certified pathologists. The locations of hepatocyte hypertrophy and lipid vacuolation were qualitatively identified and annotated by both pathologists. For Oil Red O staining, lipid-positive areas (%) were quantified from three randomly selected 400× fields per section using ImageJ (NIH, USA), and the average values were used for statistical analysis.

Liver tissues were homogenized in RIPA buffer containing protease inhibitors, followed by ultrasonic disruption. After centrifugation at 12,000 rpm for 10 minutes, the supernatant protein concentrations were determined using a BCA protein assay kit. Equal amounts of protein were resolved via SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skim milk for 1 hour. Membranes were incubated overnight at 4°C with specific primary antibodies, followed by secondary antibody incubation at room temperature for 1 hour. Signals were detected using a chemiluminescence imaging system, and band intensities were quantified using ImageJ software (v1.54g).

Plasma samples were thawed on ice to preserve integrity. For each sample, 50 μL of plasma was mixed with 200 μL of methanol, followed by ultrasonic mixing to ensure uniformity. The mixture was incubated at 4°C for 1 hour to precipitate proteins and centrifuged at 15,000 g for 10 minutes at 4°C. The supernatants were carefully collected, dried under nitrogen, and resuspended in 100 μL of acetonitrile-water solution (1:1, v/v) for HPLC-MS/MS analysis.

Metabolomic profiling was conducted using a Waters Xevo G2-XS QToF mass spectrometer coupled with an ACQUITY UPLC BEH C18 column (2.1 mm × 50 mm, 1.8 μm; Waters Corp., Milford, USA). Chromatographic separation was performed at 20°C using a Waters Acquity™ UPLC system. The mobile phase consisted of formic acid aqueous solution (solvent A) and acetonitrile (solvent B) at a flow rate of 0.4 mL/min, with a gradient program of 0–1 min: 10% B; 1-7.5 min: 10% → 65% B; 7.5-10.5 min: 65% → 100% B; and 10.5–12 min: 100% → 10% B. The injection volume was 2 μL.

The mass spectrometer operated in both positive and negative ionization modes, with a scan range of 50–1200 Da. Key parameters included a cone voltage of 40 kV, a capillary voltage of 3.0 kV, a desolvation gas flow rate of 800 L/h at 450°C, and a source temperature of 100°C. Mass calibration was performed using leucine enkephalin, with [M+H]+ at m/z 556.2771 in positive mode and [M-H]− at m/z 554.2615 in negative mode. Data acquisition was performed using MassLynx™ V4.1 software, and preprocessing, including peak alignment and normalization, was conducted with Progenesis QI V2.1 software (Waters Corp., Milford, MA, USA).

Multivariate statistical analyses, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), were conducted using Origin software (Origin Lab, Northampton, USA). Metabolites were selected based on Variable Importance in Projection (VIP) > 1, fold change (FC) > 1.5, and p < 0.05. Metabolite annotation was performed using KEGG (https://www.kegg.jp/) and HMDB (http://www.hmdb.ca/) databases, with pathway enrichment analysis conducted through MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/). Differential metabolites with AUC > 0.8 were further analyzed as potential biomarkers.

Data were expressed as mean ± SEM (or SD, as specified). Statistical comparisons were performed using One-way ANOVA, and if an overall difference was found, Tukey’s post-hoc analysis was applied. Significance was set at p < 0.05. All statistical tests were conducted in GraphPad Prism 9.

In the CTRL group, body weight steadily increased throughout the experimental period, whereas the AFLD and KPF groups, which were fed an alcohol-containing liquid diet, initially experienced weight loss followed by a slight recovery. Notably, the weight gain in the AFLD group was significantly lower than that in the KPF group ( Figure 1A ). Observation of liver appearance revealed reduced steatosis in the KPF group, evidenced by darker coloration compared to the AFLD group, which exhibited the lightest liver color ( Figure 1B ). The liver index rose from 3.94 ± 0.18% in CTRL mice to 5.28 ± 0.23% in the AFLD group (p < 0.0001 vs. CTRL); kaempferol treatment partially reversed this increase, reducing the index to 4.89 ± 0.28% (p < 0.01 vs. AFLD), as shown in Figure 1C .

Plasma levels of ALT and AST, key indicators of liver injury, were significantly elevated in the AFLD group compared to the CTRL group. KPF treatment significantly decreased these levels, with AST showing notable improvement (p < 0.05) and ALT showing a pronounced reduction (p < 0.0001) ( Figures 1D, E ). Antioxidant marker analysis in liver tissues revealed that SOD activity, which was significantly reduced in the AFLD group (p < 0.05), increased in the KPF group, although this increase did not reach statistical significance. Similarly, GSH levels were partially restored in the KPF group but remained statistically insignificant. In contrast, MDA levels, a marker of oxidative stress, were significantly elevated in the AFLD group (p < 0.05) and were significantly reduced following KPF treatment (p < 0.05) ( Figures 1F–H ).

Histopathological examination demonstrated normal liver morphology in the CTRL group, characterized by well-organized hepatocytes with no vacuolar degeneration, inflammation, or necrosis. In contrast, the AFLD group exhibited disorganized hepatic architecture, hepatocyte hypertrophy, and extensive lipid vacuolation in the cytoplasm. These pathological changes were notably reversed in the KPF-treated group, with restoration of liver structure and a reduction in lipid vacuolation ( Figure 1I ).

Lipid metabolism was assessed by measuring plasma levels of TC, TG, HDL-C, and LDL-C. KPF treatment significantly reduced TC (p < 0.0001), TG (p < 0.0001), and LDL-C levels (p < 0.01), while HDL-C levels were significantly increased compared to the AFLD group (p < 0.01) ( Figures 2A–D ). Additionally, blood glucose (GLU) levels were analyzed, and although KPF treatment tended to reduce the elevated glucose levels observed in the AFLD group, this effect did not reach statistical significance ( Figure 2E ).

Oil Red O staining of liver sections revealed extensive lipid droplet accumulation in the cytoplasm of hepatocytes in the AFLD group, indicated by numerous and larger red lipid droplets. In contrast, KPF treatment markedly reduced both the number and size of lipid droplets, demonstrating an improvement in hepatic steatosis (p < 0.05) ( Figure 2F ).

Heme oxygenase-1 (HO-1) and nuclear factor erythroid 2-related factor 2 (Nrf2) are critical regulators of antioxidant and anti-inflammatory pathways. As shown in Figure 3 , Nrf2 protein levels were significantly elevated in the AFLD group versus the CTRL group, then dropped to a lower level after KPF treatment (p < 0.05) ( Figure 3B ). In contrast, HO-1 protein expression in the AFLD group was higher than in the CTRL group, but although it trended downward in the KPF group, this change did not reach statistical significance ( Figure 3A ). Collectively, these findings suggest that KPF reduces both Nrf2 and HO-1, although the effect is only statistically significant for Nrf2.

PCA was conducted to visualize the metabolic profiles. In positive-ion mode, the CTRL, AFLD, and KPF groups were distinctly separated, whereas in negative-ion mode, the KPF group partly overlapped with the AFLD group ( Figures 4A, B ). OPLS-DA of 213 ions between the AFLD and KPF groups (p < 0.05, FC > 2, VIP > 1) revealed clear differences in metabolite profiles ( Figures 4C, D ).

A volcano plot identified 50 significantly downregulated metabolites and 79 significantly upregulated metabolites (p < 0.05, FC > 2) in the KPF group compared to the AFLD group ( Figure 4E ). Pathway enrichment analysis of these metabolites highlighted several pathways regulated by KPF, including glycolysis/gluconeogenesis, pentose phosphate pathway, and glutathione metabolism. Multiple lipid metabolism–related pathways—sphingolipid metabolism, glyoxylate and dicarboxylate metabolism, and primary bile acid biosynthesis—were also impacted ( Figure 4F ). By matching significantly altered metabolites with endogenous metabolite databases (AUC > 0.8), we identified 32 differentially regulated endogenous metabolites ( Table 1 ), which are mainly involved in oxidative stress, amino acid metabolism, and lipid metabolism, providing further mechanistic clues to KPF’s protective actions.