All chemical constituents of Scutellaria barbata were searched in the TCMSP database (https://old.tcmsp-e.com/molecule.php?qn=1925). According to the pharmacokinetic parameters: oral bioavailability (OB) ≥ 30%, and drug-like index (DL) ≥ 0.18, the eligible candidate active ingredients were screened, and the related targets of the candidate active ingredients screened in the TCMSP database were collected. The selected candidate active ingredients were retrieved in Pubchem database (https://pubchem. ncbi. nlm. nih. gov/), and the SDF file with 2D structure was downloaded. Then, the Swiss Target Prediction database (http://www.swisstargetprediction.ch/) was imported to predict the target of active ingredients. Finally, the targets collected from the Swiss Target Prediction and TCMSP databases were merged, and the repetitive values were removed as the targets of candidate active ingredients of Scutellaria barbata.

The collected candidate active ingredient targets and disease targets were crossed as potential targets for the treatment of diseases.

The gene information of potential targets for drug treatment of hepatocellular carcinoma was analyzed using the “Clusterprofiler” package of R software for GO functional enrichment and KEGG pathway analysis. p < 0.05 was used as the screening index (p < 0.01 was considered significant enrichment, and the smaller the p-value was, the more significant the enrichment).

There are many compounds in traditional Chinese medicine, and Scutellaria barbata is orally taken after decoction in traditional clinical use. Therefore, combined with clinical application, the ingredients absorbed by oral administration are the effective ingredients of Scutellaria barbata in the treatment of hepatocellular carcinoma. Ideal oral drug molecules should conform to Lipinski’s five rules of physical and chemical properties; which predicts the drug similarity of a compound with certain biological activity. According to Lipinski’s five rules, drug-like compounds used for oral administration should exhibit the following characteristics: MW ≤ 500; mlogP ≤4.5; number of rotatable bonds ≤5; hydrogen bond acceptors <10; and hydrogen bond donor <5; The Swiss ADME network service platform (http://www.swissadme.ch/)) was used to analyze the ADME characteristics of TCM ingredients interacting with the core target genes according to the ingredient-target interaction information in the “Targets Information” in the “Related Targets” section of TCMSP database. Because the expected action site of the screened compounds is the liver, it is hoped that the active ingredients should not penetrate the blood-brain barrier (BBB permeant) to reduce the adverse reactions caused by the later development of active ingredients. Finally, compounds conforming to the Lipinski's five rules and not penetrating the BBB were selected for further molecular docking with the interacting target proteins.

Rhamnazin (CMEC) and Luteolin (CMEC) dissolution with DMSO. When used, diluted with complete medium to a specific concentration (DMSO <0.05%) drug-containing solution.

The data are presented as the means ± standard deviations (SD). GraphPad Prism software was used to draw graphs. The comparison among groups were tested using one-way analysis of variance (ANOVA). The Student’s t-test was used to analyze the differences between two groups. p < 0.05 was considered statistically significant.

According to the criteria of oral bioavailability (OB) ≥ 30% and drug-like (DL) ≥ 0.18, 29 active ingredients of Scutellaria barbata were screened from the TCMSP database (Table 1). A total of 461 active ingredient-related targets were collected through the TCMSP database and Swiss Target Prediction network service platform.

By using “hepatocellular carcinoma” and “HCC” as keywords, 53 targets were collected in the TTD database, 7917 targets in the GeneCards database, 559 targets in the Disgenet database and 236 targets in the OMIM database. After the repetition was removed, 8238 potential therapeutic targets for HCC were obtained.

A total of 372 potential therapeutic targets for hepatocellular carcinoma were obtained by crossing the targets of active ingredients and potential therapeutic targets of hepatocellular carcinoma. Cytoscape software was used to construct the “Chinese medicine-ingredient—target-disease” regulatory network (Figure 1). The control network has 403 nodes and 2424 edges.

The potential target gene information of drug treatment diseases was imported into the STRING online platform (node with 0 hidden value), and the results were imported into Cytoscape3.9.0 software. The key genes of TOP10 were obtained by the MCC′ algorithm of′ Cytohubba′ plug-in (Figure 2). GEPIA database analysis showed that the mRNA expression levels of key genes CDK1, SRC, CDK4, PCNA and E2F1 in HCC were significantly upregulated compared with those in normal liver tissues (Figure 3). The Human Protein Atlas (HPA) results showed that the CDK4 gene was highly expressed in liver cancer tissues, the CDK1, SRC, and CDK2 genes were moderately expressed in liver cancer tissues, and the PCNA, and RB1 genes were lowly expressed in liver cancer tissues, but not expressed in normal liver tissues, the CCND1 gene was highly expressed in liver cancer tissues and normal liver tissues. The E2F1 gene was moderately expressed in normal liver tissue and highly expressed in liver cancer tissue; The CDKN1A and CDK6 were not expressed in liver cancer tissues or normal liver tissues (Figure 4). The results of GEPIA survival analysis showed that the high expression of the key genes CDK1, CDK2, CDK4, E2F1, and SRC in HCC was correlated with poor prognosis (Log-rank p < 0.05) (Figure 5). Combined with the verification results of key genes, it was found that the mRNA and protein expression levels of CDK1, SRC, CDK4, and E2F1 were overexpressed in HCC tissues compared with those in normal liver tissues, and the high expression in HCC patients was related to poor prognosis. These results revealed that CDK1, SRC, CDK4, and E2F1 may be the core genes of Scutellaria barbata to target in the treatment of HCC.

The gene information of potential targets for drug treatment of hepatocellular carcinoma was used for GO and KEGG enrichment analysis using the “ClusterProfiler” package of R software. A total of 3056 GO enrichment entries were obtained, including biological processes (BP, 2717), cell components (CC, 91), and molecular functions (MF, 248). The main enriched biological processes included response to xenobiotic stimuli, wound healing, response to oxidative stress; the main enriched cell components were plasma membrane raft, vesicle lumen, organelle outer membrane. The main enriched molecular functions included nuclear receptor activity, ligand-activated transcription factor activity, transmembrane receptor protein tyrosine kinase activity, etc. The KEGG pathway analysis was enriched to 180 enrichment results. The results of KEGG pathway enrichment analysis showed that genes were mainly enriched in the PI3K-Akt signaling pathway, IL-17 signaling pathway, TNF signaling pathway, apoptosis and other pathways in addition to various cancer pathways. The 10 pathways with the smallest p values of biological processes, cell components and molecular functions in GO enrichment analysis and the p values of KEGG enrichment analysis were the highest based on analysis using R language software (Figure 6).

According to the “ingredient-target” interaction information of “Targets Information” in the “Related Targets” section of the TCMSP database, one ingredient interacting with CDK1, SRC, and E2F1, nine ingredients interacting with CDK1 and SRC, and one ingredient interacting with CDK4 and CDK1 in Scutellaria barbata was found. There were five ingredients that only interacted with SRC and one ingredient that only interacted with CDK1. There are 17 ingredients interacted with core gene targets in Scutellaria barbata. The Swiss ADME network service platform was used to screen 12 compounds that conformed to the five rules of Lipinski and did not penetrate the BBB (Table 2).

AutoDock Vina software was used for molecular docking of the core gene target protein and the interaction of traditional Chinese medicine ingredients. It is generally believed that the lower the binding energy of the ligand and receptor, the better the binding force. When the binding energy is less than 0 kcal/mol, the two can spontaneously combine. When the binding energy is less than −5 kcal/mol, the binding force is better. When the binding energy is less than −7 kcal/mol, the binding configuration has strong activity. According to the molecular docking results, the binding energy of all the core active ingredients with the core target protein was ≤ −5 kcal/mol (Table 3). The docking modes of the first six compounds with the lowest Vina score were visualized (Figure 7).

After adding different concentrations of drugs to treat HepG2 cells and Hep3B cells, CCK-8 kit was used to detect cell activity. The experimental results showed that specific concentrations of Rhamnazin and Luteolin could inhibit the proliferation of HepG2 cells and Hep3B cells (Figure 8).

Previous studies have shown that inhibiting the expression of core target genes can inhibit the proliferation of cancer cells and induce apoptosis of cancer cells. According to the results of survival analysis, the high expression of these genes in hepatocellular carcinoma is related to poor prognosis. The molecular docking results of the above core targets and their corresponding active ingredients show good active configurations. Combined with previous studies on the role of core genes in the occurrence and progression of cancer and the pharmacological effects of active ingredients, it is speculated that the mechanism of Scutellaria barbata in the treatment of hepatocellular carcinoma may be through Rhamnazin and Hispidulin etc. Active ingredients such as baicalein inhibit the expression of core genes, regulate the cell cycle, inhibit the proliferation and migration of cancer cells, and induce apoptosis.

The PI3K-Akt signaling pathway plays an important role in the results of KEGG enrichment analysis. The PI3K-Akt signaling pathway is closely related to cell proliferation, cell cycle progression and apoptosis (Noorolyai et al., 2019). Inhibition of HCC progression occurs through inhibition of the PI3K-Akt signaling pathway (Liu et al., 2021). Combined with the results of network pharmacology analysis and previous studies, it is speculated that Scutellaria barbata may block the PI3K-Akt signaling pathway by active components such as wogonin, hispidulin and eriodictyol etc to inhibit the proliferation and migration of cancer cells, induce apoptosis of cancer cells, and play a role in the treatment of hepatocellular carcinoma.

Combined with the results of network pharmacology analysis and previous studies, it is speculated that Scutellaria barbata may block the PI3K-Akt signaling pathway by active ingredients such as wogonin, hispidulin and hriodictyol etc to inhibit the proliferation and migration of cancer cells, induce apoptosis of cancer cells, and play a role in the treatment of hepatocellular carcinoma.

In this study, through the network pharmacology method to determine the core target, further screening in liver cancer tissue and normal liver tissue showed that the mRNA and protein expression levels were different, and abnormal gene expression and poor prognosis-related genes were identified in hepatocellular carcinoma patients for follow-up analysis. In addition to the clinical characteristics of Scutellaria barbata, used mainly for oral administration for the treatment of hepatocellular carcinoma in this study, the AMDE characteristics of the active ingredients of Scutellaria barbata were analyzed, and were in accordance with the principle of oral administration and not through the BBB and were screened to reduce the adverse reactions after later development. Finally, the further screened genes and the required active ingredients were combined with network pharmacology analysis and previous research results to explore the mechanism of Scutellaria barbata in the treatment of hepatocellular carcinoma and improve the accuracy and scientificity of the analysis results. The results of this study are mainly based on the network pharmacology method. Currently, some active ingredients screened have no pharmacological effects and are limited to the database, so there may be some differences in the actual action process of Scutellaria barbata.