Sibe Biotechnology Co., Ltd. (Beijing) provided 70 SPF-grade male Sprague-Dawley (SD) rats, weighing (200 $\pm$ 20) g, with license number SCXK (Beijing) 2019-0010. All rats were housed in a feeding room at a temperature of 22–25 °C and humidity of 50%–65%, with free access to food and water. This study was approved by the Biomedical Ethics Committee of the Haikou Affiliated Hospital of Central South University XiangYa School of Medicine (i.e., Haikou people’s Hospital).

Danshen was purchased from Guangdong Yifang Pharmaceutical Co., Ltd. (Foshan, Guangdong, China), Masson (Macklin-M774677), and paraformaldehyde (Macklin-P885233) were purchased from Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China); $\beta$-actin Antibody (1:1000, AF7018, Affinity, Cincinnati, OH, USA), JAK2 Antibody (1:1000, AF6022, Affinity, Cincinnati, OH, USA), Phospho-JAK2 Antibody (1:1000, AF3024, Affinity, Cincinnati, OH, USA), STAT3 Antibody (1:1000, AF6294, Affinity, Cincinnati, OH, USA), Phospho-STAT3Antibody (1:1000, AF3293, Affinity, Cincinnati, OH, USA), TGF-$\beta$1 Antibody (1:1000, AF1027, Affinity, Cincinnati, OH, USA), IL-6 Antibody (1:1000, DF6087, Affinity, Cincinnati, OH, USA) and Goat Anti-Rabbit IgG (H+L) HRP (1:5000, S0001, Affinity, Cincinnati, OH, USA) were purchased from Affinity Biosciences LTD (Shanghai, China); ELISA kits: Serum iron determination kit Fe²⁺ (colorimetric method) A039-1-1, Ferritin SF test kit H129-1-1, Transferrin (TRF) test kit H130-1-1, Type III procollagen (PCIII) reagent kit H212-1-1, Hydroxyproline (Hyp) determination reagent kit (alkaline hydrolysis method) A030-2-1 were purchased from Nanjing Jianchen Bioengineering Institute (Nanjing, Jiangsu, China).

After a 1-week acclimatization period, 70 male SD rats were randomly assigned to four groups using a random number table: a normal control group with 10 rats, a model group with 20 rats, and two Danshen treatment groups with 20 rats each receiving low (3 g$\cdot$kg⁻¹$\cdot$d⁻¹ ) and high (6 g$\cdot$kg⁻¹$\cdot$d⁻¹) doses of Danshen. Except for the normal control group, the other three groups were alternately given a subcutaneous injection of a carbon tetrachloride-vegetable oil mixture on the inner side of the limbs, at a dosage of 0.3 mL/100 g body weight, twice a week. To observe the appearance of the liver and assess the degree of HF, rats were anesthetized via intramuscular injection of Zoletil 50 (Tiletamine - zolazepam composite anesthetic, 50 mg/kg) and sacrificed at 1, 4, and 8 weeks after the initial injection. Euthanasia was performed by cervical dislocation while the rats were under deep anesthesia. The successful establishment of the HF model was confirmed by histopathological examination of liver tissue sections stained with hematoxylin and eosin (HE). On day 9, the Danshen low- and high-dose groups began treatment with daily administration for a continuous period of 4 weeks.

Search for “Danshen” in the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP, https://tcmsp-e.com/), and set the Absorption, Distribution, Metabolism, Excretion (ADME) parameters with OB $\ge$30% and DL $\ge$0.18 as the screening criteria to select the pharmacologically active compounds of Danshen. The TCMSP database was used to identify the corresponding target proteins for these compounds. Next, we used the STRING database (http://www.string-db.org) to filter by species and convert the target protein names to “gene symbol” notation.

Using “HF” and “iron metabolism” as search terms, disease targets were identified using the GeneCards database (https://www.genecards.org/), and Venn diagrams were used to identify the common targets of the two diseases.

Translate the main pharmacological components, targets, and corresponding target genes related to diseases of Danshen into Venny 2.1 (http://bioinfogp.cnb.csic.es/tools/venny/index.html), which are potential target genes for Danshen in the treatment of HF and iron metabolism.

Import the obtained common target genes into the STRING database, select the species “Homo sapiens”, and set the minimum interaction score to the highest confidence level of 0.40. We then obtained an overlapping gene Protein-Protein Interaction (PPI) network diagram. The PPI network diagram was imported into Cytoscape 3.9.1 software (San Francisco, CA, USA), and the maximum clique centrality (MCC) algorithm in the cytoHubba plugin was used to identify the top ten most significant hub genes.

Utilize The DAVID database (http://david.abcc.ncifcrf.gov) was used to perform Gene Ontology (GO) functional analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment for the common targets of Danshen in the treatment of HF and iron metabolism, thereby obtaining biological processes (BP), cellular components (CC), molecular functions (MF), and signaling pathways that are closely associated with the therapeutic effects of Danshen on HF and iron metabolism.

Liver tissues were taken from each group of rats, fixed with formaldehyde, and paraffin sections were prepared. The sections were dewaxed routinely. The sections were stained with hematoxylin for 10 min. The mixture was rinsed with distilled water three times. The sections were stained with Masson’s trichrome and picric acid for 30 min. Differentiate and dehydrate directly with anhydrous alcohol and mount with neutral resin. Images were obtained under a microscope to assess the degree of HF.

Rat liver tissue was used to detect the protein expression of IL-6, JAK, and STAT by western blotting. Lysis buffer was added to extract proteins, homogenized, and centrifuged, and the total protein concentration was measured with the BCA method, boiled, loaded samples, transferred to a membrane, and blocked; diluted mouse-sourced IL-6 (1:500), JAK (1:500), STAT3 (1:500), and $\beta$-actin (1:2000) primary antibodies were incubated overnight at 4 °C; goat anti-mouse secondary antibody (1:2000) was added and incubated at 37 °C for 2 h, washed and added to the developing solution, and developed and photographed in the imaging instrument, using $\beta$-actin as an internal control to calculate the relative expression levels of the proteins.

Orbital venous blood was collected from the rats, centrifuged at 3000 rpm for 10–15 minutes to separate the serum, and then centrifuged again at 3000 rpm for 5 min. Fe²⁺, TRF, and ferritin levels were detected according to the instructions of the ELISA kit.

Graph Pad Prism software (version 9.0, Boston, MA, USA) was used for the data processing. All data were expressed as mean $\pm$ standard deviation (mean $\pm$ SD). For comparisons between two groups, the t-test was used, and for comparisons among multiple groups, one-way Analysis of Variance (ANOVA) was used.

By querying the TCMSP database, we obtained 20 pharmacologically active compounds from Danshen and their corresponding 188 targets. Information regarding the main pharmacologically active compounds is presented in Table 1.

After mapping the targets corresponding to Danshen’s active compounds, targets corresponding to HF, and targets corresponding to iron metabolism using a Venn diagram, 92 common targets were identified (Fig. 2).

Fig. 2.

Venn diagram of targets related to Danshen and HF and iron metabolism. HF, Hepatic fibrosis.

The interaction data of the common targets for Danshen’s treatment of HF and iron metabolism in the STRING database were used to obtain the overlapping gene PPI network diagram (Fig. 3). There were 92 nodes and 1543 edges. The PPI network diagram into Cytoscape 3.9.1 software and the MCC algorithm in the cytoHubba plugin was used to identify the top 10 most significant hub genes (Fig. 4).

Fig. 3.

PPI network of Danshen in the treatment of HF and iron metabolism. PPI, Protein-Protein Interaction.

Fig. 4.

Top 10 hub genes in the treatment of HF and iron metabolism by Danshen.

The DAVID database was used to perform GO functional analysis of the common targets of Danshen in the treatment of HF and iron metabolism. In terms of BP (Biological Process), select the top 10 ranked entries, which mainly involve cellular response to glucocorticoids, cellular response to nicotine, and extrinsic apoptotic signaling pathway in the absence of ligand. For CC (Cellular Component), select the top 10 ranked entries, which mainly include nucleoplasm, nucleus, and receptor complexes. For MF (Molecular Function), select the top 10 ranked entries, which mainly involve nuclear receptor activity, estrogen response element binding, and steroid binding, among others. The above results were input into the Bioinformatics website (http://www.bioinformatics.com.cn), and bubble charts for BP, CC, and MF were generated (see Fig. 5, Table 2).

Fig. 5.

GO functional enrichment bar chart for Danshen for the treatment of HF and iron metabolism. GO, Gene Ontology.

The DAVID database was used to perform KEGG functional enrichment analysis of the common targets of Danshen in the treatment of HF and iron metabolism. The top 20 most significantly enriched pathways (p 0.01) were selected, which mainly included pathways related to cellular senescence, apoptosis, AGE-RAGE, TNF, and IL-17 signaling. The results are shown in Fig. 6 and Table 3.

Fig. 6.

KEGG pathway enrichment analysis bubble chart (the size of the bubbles in the chart represents the number of genes annotated in the corresponding entries, and the color of the bubbles corresponds to the corrected p-value). KEGG, Kyoto Encyclopedia of Genes and Genomes.

Comparison of HF histopathology among the groups: In the normal group, the liver lobule structure was intact, hepatocytes were neatly arranged, and there was no evidence of hepatocyte degeneration, necrosis, inflammatory cell infiltration, or fibrous cord formation. In the model group, the liver lobule structure was significantly disrupted with blurred boundaries, disordered cell cords, swollen hepatocytes, and cytoplasmic steatosis. Compared with the model group, the Danshen high-dose group showed a relatively intact liver lobule structure and cell morphology, with improved HF, reduced fibrous cords, and decreased degrees of cellular edema, necrosis, and steatosis. Compared with the model group, the Danshen low-dose group did not show significant pathological changes in tissue, with multiple pseudo-lobules, and cells exhibited varying degrees of edema, necrosis, inflammatory reactions, and steatosis. See Fig. 7.

Fig. 7.

The Impact of Danshen on liver tissue fibrosis. Effect of Danshen on liver tissue fibrosis: (A) sham group; (B) model group; (C) Danshen low-dose group; and (D) Danshen high-dose group. Scale bar = 100 µm. Percentage of fibrous tissue area (E), hepatic lobule structural integrity score (F). ^#p 0.05 vs. the Sham group; *p 0.05 vs. the Model group. Data are presented as mean $\pm$ standard error of the mean (SEM) (n = 3). p 0.05 indicates statistical significance.

Western blot results showed that compared with the blank group, the expression of IL-6, P-JAK2, and TGF-$\beta$1 proteins in the liver of the model group rats was significantly increased (p 0.05), while the expression of IL-6, P-JAK2, P-STAT-3, and TGF-$\beta$1 in the Danshen intervention group was lower than that in the model group, especially in the high-dose Danshen group, where there was a significant downward trend (p 0.05). High-dose Danshen significantly inhibited the expression of IL-6 and TGF-$\beta$1, and suppressed the phosphorylation of JAK2 and STAT3 (P-JAK2, P-STAT-3) in CCl4-induced HF model rats (p 0.05), whereas the total protein levels of JAK2 and STAT-3 remained unchanged. The effect of low-dose Danshen was not significant (p $>$ 0.05). See Fig. 8.

Fig. 8.

The effect of Danshen on the proteins JAK2, P-JAK2, STAT3, P-STAT3, TGF-$\beta$1, and IL-6 in liver fibrotic tissue. ^#p 0.05, ^#⁢#p 0.01 vs. the Sham group; *p 0.05 vs. the Model group. Data are presented as mean $\pm$ standard error of the mean (SEM) (n = 3). p 0.05 indicates statistical significance. JAK2, Janus kinase 2; P-JAK2, Phosphorylated Janus kinase 2; STAT3, Signal transducer and activator of transcription 3; P-STAT3, Phosphorylated Signal transducer and activator of transcription 3; TGF-$\beta$1, Transforming growth factor beta 1; IL-6, Interleukin-6.

To further determine the occurrence of ferroptosis, serum Fe²⁺, serum ferritin (SF), TRF, type III procollagen (PcIII), and hydroxyproline (HyP) levels were measured. The results showed that (see Fig. 9): compared with the blank group, the levels of Fe²⁺ and TRF in the serum of the model group were significantly decreased, while the levels of SF, PcIII, and HyP were significantly increased (p 0.05). The Danshen intervention group showed an opposite trend to that of the model group in the levels of Fe²⁺, TRF, SF, PcIII, and HyP, especially in the high-dose Danshen group, where the levels of Fe²⁺ and TRF were significantly increased, and the levels of SF, PcIII, and HyP were markedly decreased (p 0.05). It can be seen that high-dose Danshen can promote an increase in blood Fe²⁺ and TRF concentrations and simultaneously decrease the concentrations of SF, PcIII, and HyP in CCl₄-induced HF model rats.

Fig. 9.

Effect of danshen on blood Fe²⁺, SF, TRF, PcIII, and HyP in HF rats. ^#⁢#p 0.01 vs. the Sham group; *p 0.05, **p 0.01 vs. the Model group. Data are presented as mean $\pm$ standard error of the mean (SEM) (n = 3). p 0.05 indicates statistical significance. SF, ferritin; TRF, transferrin; PcIII, type III procollagen; HyP, hydroxyproline.

At the level of iron metabolism, a significant increase in blood Fe²⁺ and TRF levels indicates the restoration of normal iron utilization and storage in the body.

To determine whether Danshen ameliorates liver fibrosis through the ferroptosis pathway, we measured serum MDA and GSH levels and hepatic GPX4 protein expression (see Fig. 10). MDA, the core oxidative stress product of HF tissue, was significantly elevated in serum, while Danshen attenuated this increase. GSH, the antioxidant stress marker of HF tissue, was significantly decreased, while Danshen ameliorated this decline. Western blot analysis revealed significantly decreased GPX4 protein expression in fibrotic liver tissue, which was reversed by Danshen intervention; high-dose treatment restored GPX4 levels to near-normal. These findings demonstrate that Danshen ameliorates fibrotic liver injury by inhibiting ferroptosis-related pathways and suppressing oxidative stress.

Fig. 10.

Effects of Danshen on serum MDA and GSH levels and hepatic GPX4 protein expression in liver fibrotic tissue. ^#⁢#p 0.01, ^#⁢#⁢#p 0.001 vs. the Sham group; *p 0.05, **p 0.01 vs. the Model group. Data are presented as mean $\pm$ standard error of the mean (SEM) (n = 3). p 0.05 indicates statistical significance. MDA, malondialdehyde; GSH, glutathione; GPX4, glutathione peroxidase 4.

Our computational analysis identified 92 nodes at the intersection of Danshen, hepatic fibrosis, and iron metabolism, with JAK2/STAT3 emerging as top-ranking hub genes (Figs. 2,3,4). The significant enrichment of AGE-RAGE, TNF, and IL-17 pathways (Fig. 6) aligns with established fibrogenic mechanisms, while the prominence of “cellular senescence” and “apoptosis” pathways reflects the pleiotropic nature of Danshen’s bioactive compounds. Notably, tanshinone IIA and cryptotanshinone—key diterpenoids exhibiting OB $\ge$30% and DL $\ge$0.18 (Table 1)-have been previously implicated in STAT3 inhibition and ferroptosis modulation [11, 12]. The concordance between our in silico predictions and experimental data underscores the utility of network pharmacology for hypothesis generation in complex herbal medicine research.

The JAK/STAT3 pathway represents a critical convergence point for inflammatory and iron-regulatory signals in hepatic fibrosis [13]. Our data demonstrate that CCl₄-induced fibrosis is characterized by IL-6 overexpression and STAT3 hyperphosphorylation, consistent with HSC activation and hepcidin-mediated iron sequestration [14]. Danshen dose-dependently suppressed IL-6 expression and STAT3 phosphorylation while preserving total STAT3 protein levels (Fig. 8), suggesting targeted modulation rather than non-specific cytotoxicity. This is mechanistically significant because: (1) IL-6 is a master regulator of hepcidin transcription via STAT3 binding to the HAMP promoter [15]; (2) STAT3-driven hepcidin elevation impairs iron export via ferroportin degradation, creating a hypoxic microenvironment that perpetuates fibrogenesis [16]. By interrupting this feedforward loop, Danshen restores iron bioavailability, as evidenced by increased serum Fe²⁺, transferrin (TRF) and decreased ferritin (SF) (Fig. 9). The observed 1.8-fold elevation in TRF is particularly noteworthy, as transferrin not only facilitates iron delivery but also acts as an antioxidant by sequestering free iron that would otherwise catalyze Fenton reactions [17].

Our data reveal that CCl₄-induced fibrosis produces a ferroptosis-permissive environment marked by GPX4 downregulation, GSH depletion, and MDA accumulation (Fig. 10), corroborating recent reports linking iron metabolism dysfunction to lipid peroxidation-driven cell death in chronic liver disease [18, 19]. Danshen’s restoration of GPX4 expression to near-normal levels represent a critical intervention point, as GPX4 is the sole enzyme capable of detoxifying membrane phospholipid hydroperoxides (PLOOHs) [20]. The reciprocal changes in MDA and GSH-core oxidative stress products and the primary GPX4 cofactor, respectively-demonstrate that Danshen reconstitutes the entire GPX4-GSH antioxidant axis. This finding resonates with recent observations that tanshinones can activate Nrf2, the transcriptional master regulator of GPX4 and GCLM (GSH synthesis) [21, 22]. Importantly, our study places ferroptosis within the broader JAK/STAT signaling context: STAT3 activation transcriptionally represses Nrf2, creating a vicious cycle of oxidative stress [23]; Danshen’s dual inhibition of STAT3 and restoration of GPX4-GSH likely disrupts this reciprocal inhibition.

We propose a novel feedback mechanism wherein Danshen-induced normalization of iron metabolism reinforces anti-fibrotic efficacy (Fig. 9). Elevated bioavailable Fe²⁺ satisfies the requirement of prolyl hydroxylases for proper collagen cross-linking, reducing abnormal matrix deposition [24]. Simultaneously, increased intracellular iron may suppress IL-6 transcription via ROS-dependent inhibition of NF-$\kappa$B signaling, or alternatively, promote IL-6 expression through iron-induced oxidative stress, depending on the cellular context and iron concentration, creating a stable anti-inflammatory state [25]. This “iron-inflammation” crosstalk, mediated through the JAK/STAT-GPX4 axis, represents a multi-dimensional therapeutic target that distinguishes Danshen from single-target anti-fibrotics.

While our integrated approach provides compelling evidence, several limitations warrant acknowledgment. First, the mechanistic validation remains at the protein expression level; functional indispensability of the JAK/STAT axis should be confirmed using specific inhibitors (e.g., Ruxolitinib) or hepatocyte-specific Stat3 knockout mice. Second, although we demonstrate GPX4 restoration, direct evidence that Danshen inhibits ferroptosis requires validation with GPX4 pharmacological inhibitors (RSL3/ML162) or conditional Gpx4 knockout models to establish causal dependency. Third, our study focused on male rats; sex-specific differences in iron metabolism and ferroptosis sensitivity necessitate future investigations in female cohorts [26]. Fourth, the network pharmacology analysis relied on databases with inherent annotation biases; experimental validation of additional hub genes (e.g., STAT1, ACVR1) would provide a more comprehensive understanding of Danshen’s polypharmacology. Finally, translating these findings to clinical practice requires multi-center, double-blind trials evaluating Danshen’s effects on serum PCIII, transient elastography, and patient-reported outcomes.