Reagents: Fangchinoline (≥98%, B50673) was ordered from Shanghai Yuanye Bio-Technology (Shanghai, China). N-Acetylcysteine (NAC, HY-B0215) and diethylnitrosamine (DEN, HY-N7434) were sourced from MCE (New Jersey, USA). Commercial ELISA kits for aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) were sourced from Jiangsu Kete Biotechnology (Jiangsu, China). The Nanjing Jiancheng Bioengineering Institute (Nanjing, China) provided the ROS assay kit. ELISA kits for taurine were sourced from Zhonghao Biotechnology (Beijing, China).

Antibodies: The rabbit anti-nuclear factor erythroid 2-related factor 2 (Nrf2, 16396-1-AP), anti-superoxide dismutase 1 (SOD1, 10269-1-AP), anti-β-actin (66009-1-lg) antibodies, anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 10494-1-AP) antibodies, anti-catalase (CAT, 21260-1-AP) and the mouse anti-kelch-like ECH-associated protein 1 (Keap1, 60027-1-Ig), and anti-heme oxygenase-1 (HO-1, 66743-1-Ig) antibodies were sourced from Proteintech (Wuhan, China). The rabbit anti-collagen type I alpha 1 (COL1A1 and A25439), anti-collagen type I alpha 2 (COL1A2 and A5786), anti-α-SMA (66009-1-lg), and anti-cysteine sulfinic acid decarboxylase (CSAD and A13845) antibodies were ordered from ABclonal (Wuhan, China). The rabbit anti-collagen III alpha 1 antibodies (COL3A1 and sR23957) and anti-α-SMA (R380653) were ordered from Zen-Bio (Chengdu, China).

The animal studies strictly complied with the guidelines of the Institutional Animal Ethics Committee of Nanchang University (approval No. SYXK(Gan)2021-0004).

The fixed liver, kidney, and lung tissues underwent sequential processing: dehydration through graded ethanol–xylene solutions, paraffin embedding, and microtome sectioning. Tissue sections were subsequently stained with hematoxylin and eosin (H&E), and histological imaging was performed using a microscope.

Liver sections underwent standard histological processing (deparaffinization and rehydration) prior to Picrosirius red staining. The slices were stained with Picrosirius red, dehydrated using alcohol, and subsequently visualized using a microscope.

For Masson’s trichrome staining, paraffin-processed hepatic tissue was cut into 5-μm slices and conventionally deparaffinized. The sections were incubated with potassium bichromate overnight and stained successively with hematoxylin, Ponceau S acid fuchsin stain, phosphomolybdic acid, and aniline blue. Slides were dehydrated and coverslipped under Permount mounting medium. Photographs were captured with a microscope (Olympus, Tokyo, Japan).

Liver tissues from the designated groups were homogenized, followed by centrifugation. The supernatants and serum were tested for ROS, ALT, AST, ALP, and taurine levels using commercial assay kits as per the manufacturers' protocols.

Liver samples from mice in the blank control and model control groups (three livers per group) were ground, lysed, and centrifuged, and the supernatant was collected. Protein quantification with the BCA assay preceded sample normalization for equal loading. The protein solution was hydrolyzed by protease and desalted by using a C18 column. Samples were separated using a Vanquish Neo UHPLC nanoflow system (Thermo Fisher Scientific). The mobile phases consisted of phase A: 0.1% (v/v) formic acid in water and phase B: 0.1% (v/v) formic acid in acetonitrile (100% acetonitrile). Separation was carried out using a capture-analytical dual-column configuration: trap column: PepMap Neo Trap Cartridge (300 μm × 5 mm, 5 μm); analytical column: Easy-Spray™ PepMap™ Neo UHPLC column (150 μm × 15 cm, 2 μm). The analytical column was maintained at 55 °C. Samples (200 ng) were loaded at a flow rate of 2.5 μL/min. Chromatographically separated samples were analyzed via data-independent acquisition (DIA) using an Orbitrap Astral mass spectrometer (Thermo Fisher Scientific) coupled to the nanoflow Vanquish Neo system. MS parameters included the following: ion mode: positive; MS1 scan range: 380 m/z–980 m/z; MS1 resolution: 240,000 at 200 m/z; normalized AGC target: 500%; maximum injection time: 5 m. The DIA settings included the following: MS2 scans: 299 variable windows; isolation window: 2 Th; HCD collision energy: 25%; normalized AGC target: 500%; maximum injection time: 3 ms.

The proteins in each group were quantified, and the results of the statistical analysis were used to screen out the differentially expressed protein. The principal component analysis (PCA) was performed on the samples to assess both the overall protein differences among sample groups and the magnitude of variation within each group. Proteins with either fold change (FC) ≥ 1.5 or FC ≤ 0.6667 and p-value ≤0.05 were defined as differentially expressed proteins in the two-sample comparison groups, where FC was calculated as the ratio of the average expression in the DEN-treated group to that in the control group. Functional annotation and volcano plot of the identified differential proteins were performed.

Primary activated murine HSCs (MIC-iCell-d018) were sourced from iCell Bioscience (Shanghai, China) and maintained in DMEM (Solarbio, Beijing, China) containing 10% FBS (ExCell Bio, Shanghai, China) and penicillin/streptomycin, and it was maintained under standardized conditions (37 °C, 5% CO₂, and humidified atmosphere).

Human hepatic stellate cells (LX-2 cells, SCSP-527) were purchased from the Cell Bank of Chinese Academy of Sciences Committee (Shanghai, China), cultured in a specialized medium (SNLM-206, SUNNCELL, Wuhan, China), and maintained under standardized conditions.

LX-2 cells were treated with 20 ng/mL TGF-β for 24 h or 48 h, followed by incubation with 12.5 μM FAN with or without 8 mM NAC for 24 h. After fixation with 4% paraformaldehyde and permeabilization, cells were incubated overnight at 4 °C with primary antibodies against COL1A2 or α-SMA. Following thorough washing, fluorophore-conjugated secondary antibodies were applied. The nuclei were counterstained with DAPI, and the fluorescence signal distribution was visualized under a fluorescence microscope.

For the analysis of cell morphology, the primary murine HSCs were exposed to 6.25 μM or 12.5 μM FAN with or without 5 mM taurine for 24 h. From among those cells, the group treated with taurine alone was used as the positive control. Cell morphology was observed by light microscopy.

For phalloidin staining, adherent cells on glass substrates were exposed to 6.25 μM or 12.5 μM FAN with or without 5 mM taurine for 24 h. Following fixation with paraformaldehyde and permeabilization, samples were co-stained with phalloidin and DAPI and then examined using a Diaphot fluorescence microscope.

Protein lysates from murine tissues and primary HSCs were resolved via 10%–15% gradient sodium dodecyl sulfate–polyacrylamide gel electrophoresis. After electrophoretic transfer onto membranes, they were blocked and incubated in primary antibodies including COL1A1, COL1A2, COL3A1, α-SMA, CSAD, Keap1, Nrf2, CAT, HO-1, SOD1, GAPDH, or β-actin, and then they were exposed to horseradish peroxidase-conjugated secondary antibodies. Protein band visualization was achieved using a chemiluminescence detection system coupled with the imaging platform. Quantitative analysis of band intensities was performed through pixel density measurements via ImageJ software.

Statistical analyses were performed using SPSS 22.0 and GraphPad Prism 8.0. Data were presented as the mean ± the standard deviation (SD). Multi-group comparisons were analyzed by one-way ANOVA (p < 0.05 was considered significant).

To investigate FAN’s hepatoprotective effects against fibrosis, we employed a DEN-induced mouse model replicating human hepatic fibrogenesis, with the experimental schematic detailed in Figure 1A. FAN significantly alleviated DEN-induced loss of bodyweight and liver index in mice (Figure 1B, C; Supplementary Figure S1A). In addition, FAN significantly alleviated DEN-induced changes of the lung index, thymus index, cardiac index, and renal index, but not the spleen index (Figures 1D–H). Hepatic biomarker panel abnormalities (ALT, AST, and ALP) can be regarded as a significant event of liver disease. DEN administration increased serum or liver ALT, AST, and ALP levels, but FAN could prevent these changes (Figures 1J–N).

Histopathology confirmed that FAN treatment markedly mitigated the hydropic degeneration, swelling, and necrosis of liver cells; desmoplasia; and scattered inflammatory cell infiltration around many veins in the liver (Figure 1I); it also mitigated bronchial luminal narrowing in the lung (Supplementary Figure S1B) and mesangial expansion and diffuse thickening of the glomerular basement membrane (Supplementary Figure S1C). Picrosirius red is an acid dye that binds easily to the basic groups in collagen molecules and is used to detect the deposition of collagen molecules in tissues. Masson’s trichrome staining is also a classic method to visualize the structure of collagen in pathological tissues. At 18 weeks after DEN injection, Picrosirius red staining and Masson’s trichrome staining detected severe collagen accumulation in hepatic tissues of the DEN-treated only group. These changes exhibited significant attenuation following FAN administration (Figures 2A–D). Thus, FAN can alleviate DEN-induced hepatic fibrosis.

Excessive synthesis and accumulation of liver matrix proteins contribute to collagen deposition. DEN administration markedly upregulated hepatic profibrotic markers (COL1A1, COL1A2, and α-SMA) versus that in controls, an effect substantially attenuated by FAN intervention (Figures 2E–H). Activated HSCs can turn into myofibroblasts, which are the main source of extracellular matrix synthesis and secretion. We obtained activated HSCs from the mouse liver. Primary murine activated HSCs have the characteristics of myofibroblasts, showing that the cells tend to be spindle-shaped. Cells were either treated with FAN or left untreated, and their morphology was examined. After FAN administration, the morphology of activated HSCs lost the spindle-shaped appearance, which suggests that HSCs activation was inhibited (Figure 3A). We treated LX-2 cells with TGF-β to establish an activated model of HSCs. The effect of FAN treatment on the extracellular matrix protein levels was assessed using immunofluorescence. It was found that 12.5 μM FAN significantly suppressed TGF-β-induced increases in COL1A2 and α-SMA levels in LX-2 cells (Figures 3B–E). Western blot analysis similarly demonstrated that FAN markedly inhibited the high levels of COL1A1, COL1A2, and α-SMA in activated HSCs (Figures 3F–I). These results further demonstrate that FAN can relieve hepatic fibrosis via decreasing the matrix protein expression in vivo and in vitro.

To further clarify which molecular mechanism FAN’s anti-fibrotic effect depends on, we used ultra-fast quantitative proteomic techniques to screen for differential proteins between the control and model groups. KOG analysis revealed that DEN could significantly affect taurine and hypotaurine metabolism, unsaturated fat synthesis, and arachidonic acid metabolism (Figure 4A). Using the volcano plot to quickly identify the significantly different proteins in two groups, we found that the enzyme CSAD, a key enzyme in taurine synthesis, was significantly reduced (Figure 4B). The hepatic CSAD level was decreased by DEN, and this value was increased by treatment with FAN (Figure 4C). Taurine levels in the mouse liver were measured using a kit. It was found that the taurine levels were significantly reduced in mice with hepatic fibrosis compared to the control group. FAN treatment effectively reversed this reduction (Figure 4D).

To investigate whether FAN could also inhibit hepatic fibrosis by modulating the synthesis of taurine, we exposed activated primary murine HSCs with or without taurine and examined the morphological modifications of cells and the matrix protein expression levels in cells. Clinical trials demonstrated taurine’s therapeutic potential in chronic liver disease, especially in painful muscle spasms in patients with liver disease (Vidot et al., 2018). The activated primary murine HSCs acquired myofibroblast-like features. On the one hand, we chose taurine as a tool to investigate the pharmacological mechanism of FAN, and on the other hand, we chose taurine as a positive control agent (H et al., 2018; Tapper and Parikh, 2023). The activation of HSCs was inhibited by treatment with FAN plus taurine compared with the FAN monotherapy group (Figures 4E, F). We found that the elevated expressions of COL1A1, COL1A2, and α-SMA in activated HSCs were decreased after FAN treatment. Additional taurine further reduced the expressions of the three proteins following FAN treatment (Figures 4G–J). Similarly, FAN effectively attenuated the TGF-β-induced increases in COL3A1 and α-SMA protein levels in LX-2 cells. Furthermore, compared to FAN treatment alone, the combination treatment with taurine and FAN further enhanced the ability of FAN to reduce the extracellular matrix protein levels in activated LX-2 cells (Figures 4K–M). Therefore, we reveal that FAN can inhibit hepatic fibrosis by increasing taurine content by raising the protein levels of CSAD, a key enzyme of taurine synthesis.

Studies have shown that taurine can act as an antioxidant. We investigated whether FAN inhibits liver fibrosis by regulating oxidative stress. DEN exposure for 18 weeks resulted in increased ROS levels within the hepatic tissue, which were significantly reduced by FAN treatment (Figure 5A). The Nrf2 signaling pathway is the most important antioxidant mediator. FAN administration alleviated the inhibition of the Nrf2 pathway in the liver specimens of rodents with hepatic fibrosis by increasing the Nrf2 expression and its downstream detoxification enzymes’ (HO-1, SOD1, and CAT) protein levels and decreasing the protein level of Keap1 (Figures 5B–G). This suggests that FAN may inhibit hepatic fibrosis by regulating oxidative stress.

NAC, a thiol group-containing water-soluble antioxidant, is commonly used to neutralize ROS. We also used DEN to generate a murine model of hepatic fibrogenesis and administered FAN with or without NAC (Figure 6A). We aimed to evaluate the effect of ROS removal on the anti-fibrotic effect of FAN according to the results of liver fibrosis-related indicators in vivo. The indices of the heart and thymus were significantly changed after treatment with FAN, and this improvement trend was further reinforced after combination treatment with NAC and FAN; however, the synergistic effects of NAC and FAN did not significantly alter the body weight, liver index, spleen, lung, and renal index (Figures 6B–D; Supplementary Figures S2A–D). Based on the same process, the effect of NAC on FAN-mediated ALT, AST, and ALP levels was examined by ELISA. Compared with that in the FAN-treated group, levels of these three markers were more pronouncedly reduced in the co-treated group (Figures 6E–I).

The pharmacological suppression caused by FAN on the DEN-induced histopathologic changes in the liver, including hydropic degeneration, swelling, and necrosis of liver cells and inflammatory cell infiltration, was further enhanced under combination treatment (Figure 6J). Combination treatment with NAC and FAN improved DEN-induced deposition of collagen molecules in the liver specimens of rodents better (Figures 7A, B) and further reduced the expression levels of COL1A1, COL1A2, and α-SMA (Figures 7C–F). Consistent with the findings described above, immunofluorescence results demonstrated that co-treatment with NAC further reduced the FAN-mediated decrease in TGF-β-induced α-SMA protein levels in LX-2 cells (Figures 7G, H).