Yangyinghuoxue decoction exerts a treatment effect on hepatic fibrosis by PI3K/AKT pathway in rat model: based on the network pharmacology and molecular docking

Collection of the potential action targets of YYHXD and HF

The active compounds of each herb in YYHXD were retrieved from the TCMSP database (https://old.tcmsp-e.com/tcmsp.php) and published literature. Compounds from the TCMSP database were filtered for oral bioavailability (OB) ≥30% and drug-likeness (DL) ≥0.18. The potential targets of the YYHXD active compounds were identified using SwissTargetPrediction database (http://swisstargetprediction.ch/) and DrugBank database (https://www.drugbank.ca/).

The HF-related targets were obtained from the reported literatures and Online Mendelian Inheritance in Man (OMIM) database (http://www.omim.org/) and Human Phenotype Ontology (HPO) database. The overlapping genes between YYHXD-related targets and HF-related targets were considered potential YYHXD targets for treating HF and subjected to further analysis. A “herb-ingredient-target-disease” network of YYHXD was then established using Cytoscape software (version 3.7.2) based on the drug-disease overlapping targets and their corresponding active ingredients and corresponding herbs. The topological parameters of each node in the “herb-ingredient-target-disease” network were calculated using Cytoscape software. The five compounds with the highest degree score were screened out, which may be the key effective ingredients of YYHXD for treating liver fibrosis.

Protein - protein interaction (PPI) network construction

The protein-protein interaction (PPI) analysis of overlapping targets was performed using the BisoGenet plugin in Cytoscape software. After screening twice with degree ≥2-fold median, the PPI network was constructed. The CytoHubba plugin was utilized to identify the hub genes from the PPI network based on the Maximal Clique Centrality (MCC) algorithm. The top 10 genes with the highest MCC scores were screened out as hub genes, which were considered as key targets for the treatment of hepatic fibrosis with YYHXD. Subsequently, the Molecular Complex Detection (MCODE) plugin in Cytoscape software was used to analyze PPI network modules. The PPI network was divided into several clusters based on the MCODE score. The cluster with the highest MCODE score was considered as the most critical module in the PPI network.

Functional enrichment analysis

The DAVID database (https://david.ncifcrf.gov/tools.jsp) provides dependable algorithms for functional annotation, gene categorization, and identifier translation. To further decipher the potential biological processes and signaling pathways related to YYHXD’s therapeutic effects against liver fibrosis, we conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses utilizing DAVID tools on modularized genes from the highest-ranked protein association module. The GO analysis delineates genes into three domains: Biological Processes (BP), Cellular Components (CC), and Molecular Functions (MF).

Molecular docking verification of the key ingredient and target

The five key active compounds from the “herb-ingredient-target-disease” network were docked against proteins involved in the PI3K pathway. The 3D structures of proteins and ligands were obtained from the PDB (https://www.rcsb.org) and PubChem (https://pubchem.ncbi.nlm.nih.gov/) databases, respectively. Molecular docking simulations were performed in Discovery Studio 2019 to validate the preceding enrichment analysis predictions. Briefly, protein and ligand structures were loaded into Discovery Studio and subjected to LibDock docking with parameters of 100 hotspots and 0.25Å tolerance. The docking program was executed and results visualized using Discovery Studio tools. Binding affinities between active compounds and target proteins were evaluated by LibDock scores.

Reagents and antibodies

CCl4 was purchased from Bodi Chemical Co., Ltd., (Tianjin, China). Olive oil was purchased from Beijing Solarbio Science and Technology Co., Ltd., (Beijing, China). Colchicine and pentobarbital sodium were obtained from Wuhan Servicebio Technology Co., Ltd., (Wuhan, China). Assay kits for aspartate aminotransferase (AST), alanine aminotransferase (ALT), interleukin 6 (IL6), tumor necrosis factor α (TNF-α) kits, hyaluronic acid (HA) and laminin (LN) were purchased from Shanghai Enzyme Research Biotechnology Co., Ltd., (Shanghai, China). Kits for superoxide dismutase (SOD), glutathione peroxidase (GSH-Px) and malondialdehyde (MDA) were purchased from Nanjing Jiancheng Biological Engineering Research Institute (Nanjing, China). The TUNEL Apoptosis Assay Kit was purchased from Thermo Fisher Scientific (Waltham, MA, USA). BCA Protein Assay Kit was purchased from Abcam (Cambridge, MA, USA). The chemiluminescent substrate (ECL kit) was purchased from Affinity (Jiangshu, China). Antibodies against P-PI3K, P-Akt, HSP90, MYC, p53, α-SMA, collagen I and β-actin were from Affinity (Jiangshu, China). Secondary antibodies were from Affinity (Jiangshu, China).

Preparation of prescriptions’ extract

The YYHXD is composed of 14 crude drug materials, including Radix Paeoniae Alba, Radix Bupleuri, tangerine peel, radix paeoniae rubra, Ligusticum chuanxiong, rhubarb, jujube, Salvia miltiorrhiza, angelica, Scutellaria baicalensis, licorice, Rehmannia glutinosa, Eupatorium adenophorum and Fritillaria thunbergii as a fixed ratio of 4:5:5:5:5:3:3:4:4:4:3:4:3:3. All the herbs were provided by School of Traditional Chinese Medicine, Ningxia Medical University.

The herbal prescription was decocted twice by heating under reflux with 8 volumes of water for 1 hour each time. The decoction was then filtered, the filtrates combined, and concentrated to 1.0 g crude drug per mL for subsequent administration.

Animals and drug administration

A total of 48 Wistar male rats (200 ± 10 g) were obtained from the Laboratory Animal Center of Ningxia Medical University. After obtaining IACUC approval (IACUC-NYLAC-2021-009), the rats were housed with ad libitum access to food and water under standard conditions (22–24°C, 45–50% humidity, 12-h dark/illumination cycle). Following 7 days of acclimatization, the rats were randomly allocated into four groups (n = 12 per group): normal control, CCl4 model, YYHXD treatment, and colchicine positive control. Except for the normal control group, hepatic fibrosis was triggered in the rats by administering CCl₄ subcutaneously (diluted in 40% olive oil) twice per week for 6 weeks. The normal control group received olive oil only. In week 7, the YYHXD and colchicine groups were gavaged with YYHXD (10 mL/kg) and colchicine suspension (0.2 mg/10 mL/kg) respectively for 6 weeks. The control and CCl4 groups were sham-gavaged with distilled water (10 mL/kg). At 24 h after the final gavage, all rats were anesthetized with 3% sodium pentobarbital to collect blood samples from the inferior vena cava and liver tissues. This animal study was conducted in accordance with the research ethics guidelines and animal management regulations of the Chinese Ministry of Health.

Histological analysis

The liver tissues were fixed in 10% neutral buffered formalin for 24 h, followed by paraffin embedding and sectioning at 5 μm thickness. The tissue sections were subjected to hematoxylin and eosin (H&E) staining for histopathological examination and Masson’s trichrome staining for collagen deposition analysis to evaluate liver injury and fibrosis.

Serum biochemical analysis

Serum biomarkers aspartate aminotransferase and alanine aminotransferase were quantified by colorimetric assays to assess hepatocyte damage. As cytoplasmic enzymes, their elevated levels indicate liver injury. Pro-inflammatory cytokines interleukin-6 and tumor necrosis factor-α in serum were measured by ELISA kits to evaluate inflammation status. Serum hyaluronic acid and laminin were determined by ELISA as indicators of liver fibrosis progression. They are extracellular matrix components that accumulate with increasing fibrotic tissues.

Determination of liver oxidative stress biomarkers

The activities of antioxidant enzymes superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px), along with the level of lipid peroxidation marker malondialdehyde (MDA) in liver were assayed colorimetrically using commercial kits following manufacturers’ guidelines.

TUNEL assay

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was performed using the Colorimetric TUNEL Apoptosis Assay Kit per manufacturer’s protocol. In brief, paraffin-embedded liver sections were deparaffinized, digested with protease K (37°C, 20 min), rinsed, and incubated with 50 μL TUNEL reaction mixture (37°C, 60 min, dark). After DAPI counterstaining, fluorescence microscopy was used to evaluate apoptosis.

Immunohistochemical staining

The paraffin-embedded liver tissues were cut into 4–6 mm slices, dewaxed and serially dehydrated in ethanol. The sections were boiled while immersed in citrate buffer for antigen retrieval, then incubated in 3% hydrogen peroxide for 10 minutes to inactivate endogenous peroxidase. The sections were incubated primary antibodies at 4°C overnight. The sections were then incubated with again secondary antibodies at room temperature for 20 min., stained with diaminobenzidine (DAB) and restained with hematoxylin. Positive signal was detected as a brown colour under a light microscope. Immunohistochemistry was used for assess the protein expression enriched in PI3K signaling pathway, including P-PI3K, P-Akt, HSP90, MYC, p53.

Western blotting

Total protein was extracted from liver tissues using RIPA lysis buffer with protease inhibitors and quantified by BCA assay. Equal amounts of protein were separated by 11% SDS-PAGE, transferred to PVDF membranes, and blocked with 5% nonfat milk. The membranes were incubated overnight at 4°C with primary antibodies, followed by 1 h incubation with secondary antibodies at room temperature. Protein bands were visualized using ECL substrate and a gel imaging system. β-actin served as the loading control. Western blotting was performed to analyze the effects of YYHXD treatment on the protein levels of collagen I, α-SMA, and PI3K pathway components including p-PI3K, p-Akt, HSP90, MYC, and p53.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8.3.0. Quantitative data are expressed as mean ± standard deviation (SD). Comparisons between two experimental groups were analyzed by Student’s t-test. One-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparisons test was used to analyze the differences between multiple experimental groups. P < 0.05 was considered statistically significant.

Data availability

All data, models, and code generated or used during the study appear in the submitted article.

Targets collection

Trough TCMSP database and literatures search, a total of 141 active ingredients of YYHXD were screened out (Table 1). Then, 634 corresponding targets were obtained from SwissTargetPrediction database and Drugbank database. Besides, 1175 related target genes for HF were collected from the online databases and literatures. The Venn diagram showed intersection of drug-related targets and HF-related targets (Figure 1A). From Venn diagram, 69 overlapping genes were identified.

Figure 1. Network pharmacology analysis. (A) Intersection Venn diagram of YYHXD-HF. (B) The “herb - ingredient - target - disease” network. The purple nodes represented herb. The red nodes represented YYHXD-regulated targets, the node with larger size, represented higher degree score in the network. The green nodes represented active ingredients, the node with darker color, represented higher degree score in the network. (C) PPI network construction. (D) Gene cluster and core target analysis of PPI network. (a) PPI network. (b) Identification of hub genes in the PPI network. Nodes were ranked by MCC score, with darker red represents higher the MCC score. (c–f) 4 cluster genes identified in the PPI network. The clusters were ranked by MCODE score. Cluster 1 was the most critical module in the PPI network, with an MCODE score of 29.939, containing 34 nodes. The red nodes in cluster 1 represented hub genes of PPI network. Cluster 2 contained 22 nodes, with an MCODE score of 7.524. Cluster 3 contained 17 nodes, with an MCODE score of 3.875. Cluster 4 with an MCODE score of 3.5, contained 9 nodes. (E) GO annotation analysis of Cluster 1 (top 10 most significantly enriched GO terms in BP, CC and MF). (F) KEGG pathways enrichment analysis of Cluster 1. (a) The enrichment analysis of Cluster 1 (top 10 most significantly enriched KEGG pathways). (b) KEGG pathways. (G) Molecular docking results and visual analysis. (a) Docking visualization of AKT1 and senkyunone. (b) Docking visualization of MYC and mandenol. (c) Docking visualization of TP53 and quercetin1. (d) Docking visualization of HSP90AA1 and senkyunon. (e) The heat map of LibDock Scores between the 5 crucial active molecules and 4 hub targets.

The “herb-ingredient-target-disease” network analysis

The “herb-ingredient-target-disease” network of YYHXD was constructed by integrating the drug-disease overlapping targets and their corresponding active ingredients and herbs. As shown in Figure 1B, the “herb-ingredient-target-disease” network contained 224 nodes (represented 14 herbs, 141 compounds and 69 targets) and 490 edges (indicated interactions between nodes). By performing topology analysis, 5 compounds with highest degree score were screened out from the “herb-ingredient-target-disease” network and sequentially ordered as follows: quercetin, senkyunone, wallichilide, mandenol, myricanone, which may play a key role in the network (Table 2).

PPI network analysis and functional enrichment analysis

Using BisoGenet plugin, the PPI analysis of 69 overlapping targets was performed. After screening twice with degree ≥2-fold median, the PPI network was finally constructed (Figure 1C). The PPI network contained 200 nodes and 2768 edges. The top 10 hub genes identified by CytoHubba using MCC algorithm were: TP53, ACTB, HSP90AA1, MYC, Akt1, CTNNB1, EP300, MDM2, JUN, RPS27A (Figure 1D (a–f)). It suggested that these hub genes may be the key targets for the treatment of hepatic fibrosis with YYHXD.

Using MCODE, 4 modules were chosen from PPI network. Cluster 1 was the module with the highest MCODE score and was therefore considered the most critical module in the PPI network. Cluster 1 is a gene cluster with all hub genes and other genes as the core, containing 34 genes and 494 edges. The module genes in cluster 1 were chosen for subsequent functional enrichment analysis (Figure 1D (a–f)).

To investigate the biological characteristics of modular genes in cluster 1, GO annotation and KEGG pathway analyses were carried out. GO analysis revealed the 34 modular genes were enriched in BP terms, 106 CC terms, and 132 MF terms, (P < 0.05 and FDR <0.05). The BP involved in cellular response to oxidative stress, cellular response to chemical stress, response to oxidative stress, cellular response to reactive oxygen species, response to reactive oxygen species, etc., Notably, oxidative stress related biological processes are widely enriched. The CC terms included transcription regulator complex, RNA polymerase II transcription regulator complex, cyclin-dependent protein kinase holoenzyme complex, spindle, serine/threonine protein kinase complex and so on. In terms of MF, the common targets were mostly enriched in DNA-binding transcription factor binding, RNA polymerase II-specific DNA-binding transcription factor binding, ubiquitin-like protein ligase binding, activating transcription factor binding, ubiquitin protein ligase binding and so on. The top 10 terms in the three categories of GO annotation are displayed in Figure 1E.

In addition, the 34 modular genes were enriched in various KEGG pathways, such as JAK-STAT signaling pathway, PI3K-Akt signaling pathway, Cell cycle, FoxO signaling pathway, Apoptosis. The top 10 most significantly enriched KEGG pathways were shown in Figure 1F (a, b) and Table 3. In this study, we focused on the PI3K-Akt signaling pathway, and subsequent molecular docking and animal experiments were carried out to demonstrate that the PI3K-Akt signaling pathway is an important mechanism for the anti-liver fibrosis effect of the YYHXD.

Molecular docking validation

The 5 crucial active molecules in the “herb-ingredient-target-disease” network (quercetin, senkyunone, wallichilide, mandenol, myricanone) and the target proteins which were identified as hub genes and enriched in PI3K-Akt pathway (MYC, TP53, Akt1, HSP90AA1) were selected for molecular docking verification. The molecular docking results showed that the 5 crucial active molecules of YYHXD matched well with the PI3K-Akt pathway related proteins. The binding affinity expressed as LibDock score was shown in the heat map (Figure 1G (e)). Representative molecular docking results for the target proteins (quercetin, senkyunone, wallichilide, mandenol, myricanone) and their most relevant key ligand compounds are shown in Figure 1G (a–d). The molecular docking results suggested that the YYHXD may have a direct impact on the PI3K-Akt signaling pathway.

YYHXD alleviated histopathological changes in liver fibrosis rats

The morphological changes of liver tissues were observed by HE and Masson staining, and the results were shown in Figure 2A. In the control group, the liver tissues showed normal lobular architecture and cellular structure. However, the liver tissues in the CCl₄ model group exhibited lobular structure disarray, swelling and necrosis of hepatocytes, inflammatory cell infiltration, fatty degeneration, and pseudolobule formation. When compared to model group, these changes were less prominent in the YYHXD and colchicine treated groups. The results suggest that YYHXD can reduce the severity of liver fibrosis in CCl₄-induced rats.

Figure 2. Effect of YYHXD on histological changes in CCl4 rats. (A) H&E staining (200X), Masson staining (200X) of rat livers. (B) Effect of YYHXD on level of inflammatory cytokines. Serum activities AST, ALT, IL6, TNF-α, HA and LN content. ^**P < 0.01 compared with the model group.

YYHXD ameliorate serum indicators and hepatic α-SMA, and collagen I content in liver fibrosis rats

Serum ALT and AST were examined to indicate liver injury (Figure 2B). Compared to controls, CCl₄ remarkably increased ALT and AST levels (P < 0.01), indicating severe liver damage. After YYHXD treatment, serum ALT and AST were significantly decreased compared with the model group (P < 0.01). After colchicine treatment, serum AST was significantly decreased compared with the model group (P < 0.01), which no significant change was observed in ALT level. Additionally, Serum ALT and AST levels were significantly lower in the YYHXD group than in the colchicine group (P < 0.01).

To assess inflammation, levels of IL-6 and TNF-α were examined (Figure 2B). CCl₄ markedly upregulated IL-6 and TNF-α compared to controls. Both YYHXD and colchicine significantly inhibited the upregulation (P < 0.01). Furthermore, IL-6 and TNF-α levels were lower in YYHXD versus colchicine (P < 0.01).

To further evaluate the anti-fibrosis effect of YYHXD, serum HA and LN were measured (Figure 2B). The result showed that HA and LN levels were significantly increased in the CCl₄ group compared with the normal control group (P < 0.01), and these markers were decreased remarkably in the YYHXD group and positive control group compared with the liver fibrosis model group (P < 0.01).

Furthermore, immunostaining revealed increased expression of α-SMA and collagen I, indicators of liver fibrosis, in CCl4 group compared to controls (P < 0.01). Both YYHXD and positive control significantly decreased α-SMA and collagen I expression versus CCl4 model (P < 0.01), suggesting YYHXD alleviated CCl₄-induced fibrosis (Figure 3A–3C).

Figure 3. (A) a-SMA and collagen I expression in liver tissue were detected to investigate liver fibrosis. (A, B) Immunohistochemical images showed a-SMA and collagen I positive localization and positive expression rate. ^**P < 0.01 compared with the model group. (C) Western blot image showing a-SMA and collagen I expression in liver tissues. ^**P < 0.01 compared with the model group.

YYHXD improved oxidative stress in liver fibrosis rats

GSH, SOD and MDA were measured to assess YYHXD antioxidant effects (Figure 4A). Compared with the normal group, the liver tissue GSH and SOD levels were significantly decreased and the levels of MDA were significantly increased in CCl₄ model group (P < 0.01). Compared with the model group, the GSH and SOD levels were significantly increased and the levels of MDA were significantly decreased in the YYHXD group (P < 0.01). The colchicine group significantly decreased the MDA level (P < 0.01) but colchicine group has little impact on GSH and SOD compared with the model group. Overall, YYHXD can improve the antioxidant capacity, protecting the liver fibrosis induced by CCl₄ in rats.

Figure 4. (A) Antioxidant levels in the treatment group YYHXD. GSH, SOD, MDA. (B, C) YYHXD can attenuate the apoptosis of liver parenchymal cells induced by CCl4 liver fibrosis, ^**P < 0.01 compared with the model group. The result of TUNEL assay was shown in Figure 4B, 4C. The apoptosis rate in YYHXD group and colchicine group decreased significantly, ^****P < 0.0001 compared with the model group.

YYHXD attenuated hepatocytes apoptosis in liver fibrosis rats

The result of TUNEL assay was shown in Figure 4B, 4C. The number of apoptotic liver parenchymal cells in the normal control group was very small, while more apoptotic hepatocytes were found in the liver tissue of the liver fibrosis model group. As expected, YYHXD could attenuated apoptosis of liver parenchymal cells induced by CCl₄, Apoptosis rate in the blank group (0.07 ± 0.03) and the model group (0.79 ± 0.06), its ability to anti-apoptosis of hepatocytes was better than colchicine. It was obvious that the apoptosis rate in YYHXD group (0.19 ± 0.04) and colchicine group (0.48 ± 0.03) decreased significantly (P < 0.01), YYHXD group was significantly stronger than that in colchicine group (P < 0.01).

YYHXD inhibited the PI3K-Akt signaling pathway in liver fibrosis rats

In this study, network pharmacology research and molecular docking experiments suggested that PI3K-Akt signaling pathway was potential pharmacological mechanism of YYHXD treatment for hepatic fibrosis. Thus, immunohistochemistry and Western blot assay were used to assess the effects of YYHXD treatment on the expression levels of the key genes in the pathway (P-PI3K, P-Akt1, HSP90, MYC, P53). As shown by Western blot assays and immunohistochemistry (Figure 5A–5D), the expression of P-PI3K, P-Akt1, HSP90, MYC, P53 proteins was upregulated in the CCl₄ model group compared with the normal group (P < 0.01), while all were significantly attenuated by YYHXD and colchicine (P < 0.01). Moreover, compared with colchicine treated group, fewer expressions of P-PI3K, P-Akt1, HSP90, MYC, p53 proteins were found in YYHXD group by immunohistochemical detection (P < 0.01), and fewer expressions of P-PI3K, HSP90, p53 but not P-Akt1, MYC proteins were found in the YYHXD group compared to the colchicine treated group by Western blot analysis (P < 0.05) (Figure 5E, 5F).

Figure 5. PI3K-AKT signaling pathways and molecular docking targets. (A, C) Immunohistochemical images showing P-PI3K, P-AKT1, HSP90, MYC, P53. Positive localization of protein in liver tissues. (B, D) Immunohistochemical positive expression rate. Expression rate. ^**P < 0.01 compared with the model group. (E, F) Western blot image showing P-PI3K, P-AKT1, HSP90, MYC, P53 expression in liver tissues. ^**P < 0.01 compared with the model group.