Hispidulin Suppresses Hepatocellular Carcinoma Proliferation by Targeting PI3K-Akt Pathway: A Network Pharmacology and Experimental Validation Study

Abstract

Purpose

As a natural compound present in a variety of plants, hispidulin has an inhibitory effect on tumors.

Methods

In this study, we combined network pharmacology, bioinformatics methods and experimental verification to explore the potential targets of hispidulin in inhibiting hepatocellular carcinoma.

Results

The network pharmacology analysis showed that the pharmacokinetic properties of hispidulin conformed to the Lipinski’s five rules, and it had no hepatotoxicity, carcinogenicity, cytotoxicity, immunotoxicity, mutagenicity, and had the potential to be developed as a drug. The results of network pharmacology and bioinformatics analysis showed that hispidulin may interfere with the occurrence and progression of hepatocellular carcinoma by affecting CDK1, SRC, CCNB1, AURKB and PI3K/AKT signaling pathways. In vitro and in vivo studies, hispidulin inhibited the growth of hepatocellular carcinoma, induced apoptosis and inhibited migration of hepatocellular carcinoma cells and exhibited no significant toxicity to the normal tissues and organs of mice. It downregulated the expression of CDK1, CCNB1, SRC, AURKB protein and attenuated the PI3K/AKT pathway signal. These results suggest that hispidulin can exert anti-HCC effects through the PI3K/AKT pathway.

Conclusion

Hispidulin can inhibit the occurrence and progression of hepatocellular carcinoma through multiple targets, showing good anti-tumor efficacy in vitro and in vivo, and has the potential to develop new therapeutic drugs for hepatocellular carcinoma.

Introduction

Hepatocellular carcinoma, as a malignant tumor, is the sixth most commonly diagnosed cancer and the third leading cause of cancer-related deaths globally.¹ During the early stages, HCC can be treated using modalities such as tumor resection, ablation, and liver transplantation. However, a considerable number of HCC patients are already at an advanced stage at the time of diagnosis and often cannot receive curative treatment.² Sorafenib is currently the leading first-line drug for patients with advanced liver cancer, and its safety and anticancer effects have been successfully verified in the real world.^3,4 However, the clinical efficacy of sorafenib remains limited due to the complex etiology of tumors.⁵ The emergence of new technologies, such as high-throughput sequencing and molecular biology, has enabled increased understanding of the genomic characteristics of liver cancer and broadened the development prospects of new targeted drugs.^6,7

Due to the pharmacological activity of secondary metabolites produced by plants, herbal medicines were the earliest drugs used by humans. Some of these secondary metabolites also have therapeutic effects on cancer; thus, plants are a source of natural and recombinant anticancer drugs.⁸ Hispidulin is a monomethoxy and trihydroxyflavone compound widely distributed in a variety of plants. Synthetic drugs target one target, while herbal drugs target multiple targets related to cancer cell development.⁹ Herbal medicines are much less toxic to normal cells.^10,11 Hispidulin is a monomethoxy and trihydroxy flavonoid compound widely distributed in a variety of plants. It is mainly divided into six families: Asteraceae, Lamiaceae, Water Leaf, Verbenaceae, Bananaceae, and Irisaceae. Research shows that homoplantaginin exhibits anticancer activity against various human cancer cell lines, including glioblastoma,^12,13 renal cell carcinoma,¹⁴ gastric cancer,¹⁵ pancreatic cancer,¹⁶ ovarian cancer,¹⁷ hepatocellular carcinoma,¹⁸ gallbladder cancer¹⁹ and acute myeloid leukemia.²⁰

Traditional drug discovery and development methods are mainly based on animal in vivo experiments and in vitro drug screening. These methods can be costly and labor-intensive. However, advancements in bioinformatics and network pharmacology are improving the efficiency of drug discovery and development. Bioinformatics and network pharmacology are playing an increasingly important role in drug discovery and drug development.²¹ Bioinformatics also plays an important role in the development of anticancer drugs as cancer genome characteristics continue to be discovered.²² Molecular docking can help identify potential biological targets and verify their feasibility as drug targets.²³ It reveals the interaction mechanisms between molecules and their targets. Through early screening and optimization, molecular docking significantly shortens the drug development timeline and reduces the high costs associated with research and development.^24,25 Therefore, the process of drug development assisted by bioinformatics and network pharmacology can shorten the cycle of drug development and reduce the cost of drug development.²⁶ Recent integrative bioinformatics analyses have further identified key hub genes and signaling pathways driving HCC progression, highlighting promising therapeutic targets and underscoring the value of computational approaches in HCC drug discovery.^27–30 In this study, the molecular mechanism of hispidulin in the treatment of HCC was analyzed by bioinformatics methods and verified by experiments.

Methods

Identification of Hispidulin Targets in Hepatocellular Carcinoma

The compound name “Hispidulin” was used to search the TCMSP database (https://www.tcmsp-e.com/) to retrieve its associated target proteins, while its canonical SMILES string was submitted to the SwissTargetPrediction server (http://swisstargetprediction.ch/) with the organism set to “Homo sapiens”, and predicted targets with a probability greater than 0.1 were retained to compile the drug target set. Disease targets for hepatocellular carcinoma were obtained by querying GeneCards (https://www.genecards.org/), with genes filtered by a relevance score≥ 10, DisGeNET (https://disgenet.com/) with a gene–disease association score≥0.1, OMIM(https://www.omim.org/), and the Therapeutic Target Database (https://db.idrblab.net/ttd/); all disease targets were merged and deduplicated into a single set. The hispidulin targets and the hepatocellular carcinoma targets were subsequently calibrated using the UniProt Retrieve/ID mapping tool (https://www.uniprot.org/) by retaining only reviewed (Swiss-Prot) entries mapped to “Homo sapiens”, thereby obtaining standardized human target collections. Overlapping drug and disease targets were then identified and displayed by a Venn diagram.

“Drug-Target-Disease” Regulation Network Construction

To analyze the process of systematic intervention of drugs on diseases through multiple targets. The drug-target-disease regulatory network was constructed by matching drugs, targets and diseases. The hispidulin was defined as the drug, the drug-disease intersection target was defined as the target, and the HCC was defined as the disease establishment attribute file, the “drug-target-disease” regulatory network was constructed and visualized using Cytoscape 3.9.0 software.

ADME and Toxicity Analysis of Hispidulin

The SMILES string of hispidulin was downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). The absorption, distribution, metabolism, and excretion (ADME) characteristics of hispidulin were analyzed using the SwissADME online server (http://www.swissadme.ch/). Hepatotoxicity, mutagenicity, immunotoxicity, cytotoxicity, and carcinogenicity of hispidulin were assessed using the ProTox-II online server (https://tox.charite.de/protox3/).

PPI Network Construction and Determination of Core Targets

A protein-protein interaction (PPI) network was constructed to analyze target protein interactions and identify key node proteins. The drug-disease intersection targets were submitted to the STRING database (https://cn.string-db.org/), with the organism set to “Homo sapiens” and the minimum required interaction score set to “highest confidence (0.900)”. This stringent threshold was chosen to minimize false-positive interactions and ensure that only high-confidence protein associations were retained for analysis. Nodes with a connectivity degree of zero at this threshold were removed, as they lack validated interaction partners and do not contribute to the network topology. The resulting network was imported into Cytoscape 3.9.0 software. The “MCC” (Maximal Clique Centrality) algorithm in the “CytoHubba 0.1” plug-in was used to identify core genes, as MCC has been shown to outperform other algorithms for hub protein identification. The core genes were then entered into the “Boxplots” module of the GEPIA database (http://gepia.cancer-pku.cn/), with significant differential expression defined as |log2FC| > 1 and p-value < 0.01. The “LIHC” dataset was selected, and matched normal data were set to “Match TCGA normal and GTEx data” to analyze differential mRNA expression of core genes between normal and tumor tissues. For survival analysis, the core genes were input into the “Survival Analysis” module of GEPIA. Overall survival (OS) was assessed using the median expression cutoff in the LIHC dataset; a log-rank p < 0.05 was considered statistically significant. The exact log2FC and adjusted p-values for all core genes are provided in Supplementary Table S1. Additionally, the core genes were entered into the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/). In the “Tissue” module, “Liver” was selected to obtain protein expression in normal liver tissues; in the “Cancer” module, “Liver cancer” was selected to evaluate protein expression in HCC tissues.

Molecular Docking

The binding capacity of hispidulin with core target proteins was assessed by molecular docking. The 3D structure of hispidulin was retrieved from the PubChem database and energy-minimized using the MM2 force field in ChemBio3D Draw (ChemBioOffice 2019). The crystal structures of the core target proteins were obtained from the PDB database. Water molecules and co-crystallized ligands were removed using PyMOL 2.5.2. Polar hydrogens were added and Gasteiger charges were assigned using AutoDockTools-1.5.6. Molecular docking was performed with AutoDock Vina 1.1.2. The search space was defined as a grid box covering the predicted binding pocket of each target, with the exhaustiveness parameter set to 8 and up to 10 binding modes generated per ligand-protein pair. The binding pose with the lowest docking energy (kcal/mol) was selected as the optimal conformation. Binding energies below −5 kcal/mol were considered indicative of spontaneous binding, and those below −7 kcal/mol were considered indicative of strong binding activity. Results were visualized using PyMOL 2.5.2.

KEGG Enrichment Analysis

KEGG enrichment analyses were performed using the overlapping targets of the drug and disease using the “ClusterProfiler 4.18.2” package in R language (P<0.01 was regarded as significant). The “enrichplot 1.30.3” R package was used to analyze and visualize gene-pathway and pathway-pathway associations.

Cell Lines and Culture

Human hepatocellular carcinoma cell lines Hepa1-6 and Huh7 cells, were cultured in Minimum Essential Medium (MEM) supplemented with 10% fetal bovine serum (FBS), streptomycin (100 ng/mL), and penicillin (100 U/mL), and the cells and culture medium were obtained from Zhong Qiao Xin Zhou Biotechnology. The parameters of the incubator were 5% CO2 and 37 °C.

Compound

Hispidulin (MCE) was dissolved in DMSO. When used, the drug-containing solution was diluted with complete medium to a specific concentration (DMSO< 0.05%).

Cell Viability Assay

Hepa1-6 cells and Huh7 cells were cultured overnight in 96-well plates at 1×10⁵ cells/mL (100 μL/well). A CCK-8 kit (ProteinBio) was used to detect cell viability. The cells were treated at drug concentrations of 0 μM, 25 μM, 50 μM, 100 μM, 200 μM, 400 μM for 12 h, 24 h, 72 h, respectively. The original medium was removed, and the cells were washed twice with 100 μL of PBS per well. Then, 100 µL of complete medium (containing 10 µL of CCK-8 reagent) was added to each well and incubated for 2 h, and the enzyme calibration detection OD value at 450 nm wavelength was used.

Colony Formation Assay

Hepa1-6 cells and Huh7 cells were seeded in 6-well plates (1 × 10³/well) and cultured overnight at 37°C and 5% CO2. The medium was removed, and the cells were washed twice with PBS. Hispidulin at concentrations of 0 μM, 50 μM, 100 μM, 200 μM were added respectively. The culture medium was replaced every 3 days. After 14 days, the colonies were fixed with 4% paraformaldehyde for 20–25 min and stained with crystal violet solution for 10 min. Cell colonies in each well that had ≥50 cells per colony were counted after washing with water.

Apoptosis Analysis by Annexin-V FITC/PI Double Staining

Huh7 and Hepa1-6 cells in the logarithmic growth phase were evenly inoculated into 6-well cell culture plates at a density of 2.5 × 10⁶ cells per well. After 12 hours, when the cells were completely adherent, they were treated with hispidulin-containing medium at concentrations of 0 μM, 50 μM, 100 μM, and 200 μM for 24 hours. Apoptotic cells were identified using an Annexin V-FITC Apoptosis Detection kit (Beyotime) with Annexin V and PI staining. The cells were then digested with trypsin without EDTA, centrifuged at 1800 rpm for 5 min, and collected. Subsequently, the cells were resuspended in 100 μL of Binding buffer, and 5 μL of annexin V-FITC and PI stains were added to each tube for 10 min in the dark. Cell apoptosis was detected using flow cytometry, and the experimental results were analyzed with FlowJo software.

Wound‑Healing Assay

The cells were seeded in a 6-well plate. After the cells were fully fused, the cell monolayer was disrupted with a 100 μL pipette tip, and then, serum-free medium with hispidulin concentration of 0 μM, 50 μM, 100 μM, and 200 μM was added. The scratches were imaged at 0 and 48 h after scratching using a 10× objective microscope. The cell migration rate was calculated as the distance of the recovered wound versus that of the original wound.

Western Blot Analysis

Hepa1-6 cells and Huh7 cells were treated with hispidulin at concentration of 0 μM, 50 μM, 100 μM, and 200 μM for 48 h. Total protein was isolated by RIPA lysis solution (Proteinbio) containing a protease inhibitor, and the protein concentration was measured by a BCA protein detection kit. The protein was then separated by 10% SDS‒PAGE and transferred to PVDF membranes. Five percent skimmed milk powder was used to block the membranes for 2 h. Then, the membrane was incubated with GAPDH (abclonal), CDK1 (abclonal), AURKB (abclonal), CCNB1 (abclonal), SRC (abclonal), AKT (abclonal), P-AKT (abclonal), PI3K (abclonal), bax (abclonal), and bcl-2 (abclonal) antibody solution overnight at 4°C. The membranes were incubated with HRP-conjugated goat anti-rabbit IgG for 2 h at room temperature, and the protein bands were visualized by an ECL kit. ImageJ software was used to measure the gray value of Western blot bands. The relative expression level was determined by dividing the gray value of the target protein by that of the internal reference protein.

Animal Study

This study followed the requirements of the ARRIVE guidelines to ensure the transparency and repeatability of animal experiments, and was approved by the Animal Ethics Committee. Three- to five-week-old BALB/c-nu male mice were purchased from the Zhejiang Weitong Lihua Experimental Animal Technology Co., Ltd. and divided into four groups according to treatment: control, 4 mg/kg, 8 mg/kg, 16 mg/kg. A total of 1 × 10⁷ H22 cells in 200 μL of PBS were implanted subcutaneously into the mice. The mice were divided into four groups and were administered doses of 0 mg/kg, 4 mg/kg, 8 mg/kg, and 16 mg/kg via intragastric administration every two days. After 21 days, the mice were sacrificed by cervical dislocation for hematoxylin and eosin (HE) staining.

Histopathologic Examination (H&E)

The tumor tissues, kidneys, heart, lungs and livers of mice were harvested and fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin. The paraffin-embedded tissue samples were sectioned into 5 μm slices and then stained with H&E. Finally, the sections were placed under an upright fluorescence microscope to observe the pathological changes in the transplanted tumor tissues of each group. The images were then captured and analyzed.

Statistical Analysis

All experiments were performed in triplicate as independent biological replicates. Data are expressed as mean ± SD. GraphPad Prism software was used for statistical analysis and graphing. For comparisons involving more than two groups, one-way analysis of variance (ANOVA) was used, followed by Tukey post-hoc test for multiple pairwise comparisons. For comparisons between two groups, a two-tailed, unpaired Student t test was used. A p-value < 0.05 was considered statistically significant.

Results

Hispidulin Has the Potential to Affect the Occurrence and Progression of Hepatocellular Carcinoma Through CDK1, CCNB1, SRC, and AURKB

The workflow of this research is shown in Figure 1. Network pharmacology analysis identified 111 targets for hispidulin and 8,237 for hepatocellular carcinoma, with 98 targets overlapping (Figure 2A). Hispidulin has high bioavailability, does not cross the blood-brain barrier, and is free from hepatotoxicity, carcinogenicity, immunotoxicity, mutagenicity, and cytotoxicity (Figure 2C). The “drug-target-disease” regulation network (Figure 2B) has 103 nodes and 196 edges. The PPI network (Figure 2D) which contains 69 nodes and 139 edges. The “MCC” algorithm in the “CytoHubba” plug-in was used to identify the core genes as CDK1, SRC, CCNB1, CCNB2, CDK6, and AURKB. (Figure 2E).

Figure 1.

The workflow of this research.

Figure 2.

CDK1, CCNB1, SRC and AURKB may be the core targets of hispidulin intervention in the occurrence and progression of hepatocellular carcinoma. (A) Drug-disease target intersection Venn diagram. (B) “Drug-target-disease” regulation network. (C) Chemical structure, ADME characteristics and toxicological parameters of hispidulin. (D) Intersection target protein-protein interaction network. (E) Hub targets (top 6). (F) Validation of the mRNA expression levels of hub targets. (G) Validation of the protein expression levels of hub targets. (H) Overall Survival of the 6 hub genes overexpressed in LIHC patients. (*p<0.05, **:p<0.01).

Bioinformatics analysis revealed that mRNA levels of CDK1, SRC, CCNB1, CCNB2, and AURKB were significantly higher in HCC tissues than in normal liver tissues, whereas CDK6 did not meet the significance threshold (Supplementary Table S1 and Figure 2F). HPA database analysis showed that the protein expression levels of CDK1, CCNB1, SRC and AURKB in liver cancer tissues were higher than those in normal liver tissues (Figure 2G). Using the GEPIA database based on TCGA data, the expression of core genes and the survival prognosis of patients with liver cancer were analyzed. The results showed that the high expression of CDK1, SRC, AURKB, and CCNB1 was associated with a poor prognosis in patients with liver cancer (Figure 2H) (Logrank p < 0.05).

Hispidulin Inhibits the Growth of Hepatocellular Carcinoma

Hepa1-6 and Huh7 cells were treated with different concentrations of hispidulin, and cell viability was assessed using the CCK-8 assay. The experimental results showed that hispidulin could inhibit the proliferation of Hepa1-6 cells and Huh7 cells at 24 h (Figure 3A), 48 h (Figure 3B) and 72 h (Figure 3C). Moreover, the results of the colony formation assay showed that hispidulin could inhibit the proliferation of Hepa1-6 and Huh7 cells (Figure 3D and E).

Figure 3.

Hispidulin inhibits the growth of hepatocellular carcinoma. (A) Hispidulin inhibits the growth of hepatocellular carcinoma cells in vitro for 24 hours (n=3). (B) Hispidulin inhibits the growth of hepatocellular carcinoma cells in vitro for 48 hours (n=3). (C) Hispidulin inhibits the growth of hepatocellular carcinoma cells in vitro for 72 hours (n=3). (D) Hispidulin inhibits colony formation of Hepa1-6 cells (n=3). (E) Hispidulin inhibits colony formation of Huh7 cells (n=3). Error bars indicate standard deviation (SD). The data presented here are based on three independent experiments. (*p<0.05, **p<0.01).

Hispidulin Inhibits Migration and Induces Apoptosis of Hepatocellular Carcinoma

The wound‑healing assay showed that the migration rate of Hepa1-6 and Huh7 cells decreased with increasing hispidulin concentration (Figure 4A and B). It was proven that the migration ability of Hepa1-6 and Huh7 cells was inhibited. The expression of pro-apoptotic protein Bax and anti-apoptotic protein Bcl-2 was evaluated by Western blotting. The results suggested that the expression of the apoptosis-related protein Bcl-2 decreased and that of Bax increased after treatment with hispidulin. Flow cytometry showed that the apoptosis rate of Hepa1-6 and Huh7 cells increased with increasing hispidulin concentration (Figure 5A and B).

Figure 4.

Hispidulin inhibits hepatocellular carcinoma migration. (A) Hispidulin inhibits the migration of Hepa1-6 cells. (B) Hispidulin inhibits the migration of Huh7 cells. The data presented here are based on three independent experiments. (*p<0.05, **p<0.01).

Figure 5.

Hispidulin induces hepatocellular carcinoma apoptosis. (A) Hispidulin induced apoptosis of Hepa1-6 cells (n=3). (B) Hispidulin induced apoptosis of Huh7 cells. Error bars indicate standard deviation (SD). The data presented here are based on three independent experiments. (*p<0.05, **p<0.01).

Hispidulin Reduces CDK1, CCNB1, SRC, and AURKB Protein Expression and Decreases PI3K/AKT Signaling Pathway

The study shows that a molecular docking binding energy below −5 kcal/mol indicates spontaneous binding, while below −7 kcal/mol suggests strong activity in the binding configuration. The docking energies of hispidulin and CDK1, CCNB1, SRC, AURKB were all less than −7 kcal/mol (Table 1). It shows that the drug and the target can be well combined, and the conformational activity is strong (Figure 6A–D). The results of KEGG enrichment analysis showed that the most enriched pathway for drug action on disease targets was the PI3K/AKT signaling pathway (the red highlight box section in Figure 6E). In vitro study, the expression levels of CDK1, CCNB1, SRC, and AURKB proteins decreased following treatment with hispidulin. The expression of PI3K-AKT pathway protein decreased, and the pathway signal was weakened (Figure 7A–G).

Table 1.

Molecular Docking Parameters

Target	PDBID	Grid Center			NPT			Spacing	Binding Energie
CDK1	6GU2	31.499	22.784	20.458	40	40	40	1	−8.3(kcal/mol)
CCNB1	5IQF	15.42	−59.26	19.029	40	40	40	1	−9.6(kcal/mol)
SRC	2SRC	20.671	33.876	67.502	40	40	40	1	−8.9(kcal/mol)
AURKB	4AF3	15.553	−16.602	−3.528	40	40	40	1	−8.7(kcal/mol)

Figure 6.

Hispidulin forms a stable structure with core targets CDK1, CCNB1, SRC, and AURKB in virtual docking, and the most enriched pathway of core genes is PI3K-AKT signaling pathway. (A) Hispidulin-CDK1. (B) Hispidulin-CCNB1. (C) Hispidulin-SRC. (D) Hispidulin-AURKB. (E) KEGG enrichment analysis.

Figure 7.

Hispidulin reduces CDK1, CCNB1, SRC, and AURKB protein expression and decreases PI3K/AKT signaling pathway. (A) Western blotting results. (B) The relative expression of PI3K. (C) The relative expression of P-AKT/AKT. (D) The relative expression of SRC. (E) The relative expression of CCNB1. (F) The relative expression of CDK1. (G) The relative expression of AURKB. Full-length membrane images see Supplementary Information. The data presented here are based on three independent experiments. (*p<0.05, **p<0.01).

Hispidulin Inhibits the Growth of Hepatocellular Carcinoma

In animal studies, hispidulin has been shown to inhibit tumor growth. After treatment with hispidulin, the tumor volume and weight of mice in the experimental group were significantly smaller than those in the control group (Figure 8A–C). HE staining of tumor tissue sections further revealed that, compared to the control group, the tumor condition improved following the intervention with hispidulin (Figure 8D).

Figure 8.

Hispidulin inhibits the growth of hepatocellular carcinoma. (A) Tumor changes in mice after treatment. (B) Time variation trend of tumor volume in mice. (C) Tumor weight in mice. (D) Tumor histopathology in mice. The data presented here are based on three independent experiments. (*p<0.05, **p<0.01).

Hispidulin Exhibits No Significant Toxicity to Normal Cells and Tissues

Hematoxylin and eosin (HE) staining of the liver, heart, kidneys, and lungs revealed no apparent organ damage following hispidulin treatment (Figure 9A). Serum biochemical analysis revealed that alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were lower in the hispidulin-treated group than in the control group (Figure 9B and C), indicating that hispidulin may improve liver function in tumor-bearing mice. Additionally, monitoring the weight changes of the mice revealed no significant difference in body weight between the control group and the hispidulin treatment group throughout the growth period (Figure 9D). Lastly, while the red blood cell count (RBC), hemoglobin (HGB), and platelet count (PLT) in the mice’s blood routine showed slight decreases compared to the control group after hispidulin treatment, these changes were not statistically significant (Figure 9E–H).

Figure 9.

Hispidulin exhibits no apparent toxicity to normal tissues and may improve liver function. (A) H&E staining of the major organs from mice treated with different therapies. (B) ALT. (C) AST. (D) Weight change of mice. (E) WBC. (F) RBC. (G) HGB. (H) PLT. The data presented here are based on three independent experiments. (*:p<0.05, **:p<0.01).

Discussion

Hispidulin is a monomethoxy and trihydroxy flavonoid with physicochemical properties consistent with Lipinski’s rule of five (molecular weight: 300.26, rotatable bonds: 2, H-bond acceptors: 6, H-bond donors: 3, MLOGP: 0.22). Toxicity predictions indicated no hepatotoxicity, mutagenicity, carcinogenicity, cytotoxicity, or immunotoxicity, supporting its suitability as a drug-like candidate.

Bioinformatics analysis of hispidulin targets in HCC identified CDK1, SRC, CCNB1, and AURKB as core genes significantly upregulated at the mRNA level in HCC tissues compared with normal liver. Higher expression of these genes was associated with poorer overall survival, and their protein levels were elevated in HCC tissues relative to normal liver. SRC is a proto-oncogenic tyrosine kinase that drives cancer cell proliferation, migration, and invasion.^28–30 In HCC, SRC signaling is further amplified through membrane anchoring mechanisms³¹ and contributes to immune microenvironment remodeling.^32,33 CDK1, a central cell cycle checkpoint kinase,³⁴ controls the S-G2-M phase transition.³⁵ CDK1 inhibition arrests HCC cells at the G2-M boundary,³⁶ and elevated CDK1 expression has been linked to immune evasion in several cancer types.^37,38 CCNB1, the regulatory subunit of the CDK1-cyclin B complex,³⁹ promotes HCC cell proliferation, migration, metastatic progression, and sorafenib resistance.^40–42 Its expression also modulates immune activity within the tumor microenvironment.⁴² AURKB, an essential mitotic regulator, maintains chromosomal stability during cell division. AURKB overexpression is frequently observed in HCC and correlates with poor prognosis and altered immune status.^43,44 Together, these bioinformatics findings, supported by prior experimental evidence, establish CDK1, SRC, CCNB1, and AURKB as functionally relevant targets of hispidulin in HCC. Notably, although hispidulin has been reported to inhibit PI3K/AKT signaling in HepG2 cells¹⁸ and other cancer types,¹⁶ the specific upstream targets through which this pathway is modulated in HCC have not been systematically identified. Our study fills this gap by combining network pharmacology, bioinformatics, molecular docking, and experimental validation in dual HCC cell lines and an in vivo model.

KEGG enrichment analysis revealed that the PI3K/AKT pathway contained the largest number of enriched target genes among all identified pathways. This pathway regulates cell growth, proliferation, survival, and metabolism, and its aberrant activation is a hallmark of many cancers, including HCC.^45,46 In this study, the relationship between the core targets and the PI3K/AKT pathway followed a clear upstream-downstream pattern: CDK1 activates PI3K/AKT signaling to promote tumor progression,^47,48 CCNB1 modulates PI3K/AKT phosphorylation,^49,50 SRC regulates cell cycle proteins through PI3K/AKT-dependent mechanisms,⁵¹ and AURKB activates PI3K/AKT signaling to drive tumor development.^52,53 Our molecular docking results confirmed that hispidulin binds stably to all four target proteins with favorable binding energies (Table 1). Thus, the present findings support a model in which hispidulin simultaneously downregulates these four core targets, collectively attenuating PI3K/AKT signaling and ultimately suppressing HCC proliferation, migration, and tumor growth. It should, however, be noted that other KEGG pathways enriched in this analysis may also contribute to hispidulin’s anti-HCC effects, and their individual roles remain to be investigated in future studies. From a translational perspective, given that PI3K/AKT pathway activation is implicated in resistance to sorafenib and regorafenib,^3,54,55 the ability of hispidulin to attenuate this signaling axis suggests potential utility as an adjunct to existing targeted therapies.

In vitro and in vivo experiments confirmed that hispidulin significantly suppressed HCC cell growth, reduced migration, and induced apoptosis. Hispidulin treatment downregulated CDK1, SRC, CCNB1, and AURKB protein expression and attenuated PI3K/AKT signaling in a concentration-dependent manner. In the xenograft model, hispidulin inhibited tumor growth without apparent organ toxicity and was associated with improved liver function markers (ALT, AST). It is important to acknowledge several limitations of this study. First, the highest in vitro concentration used (200 uM) exceeds typical physiological drug exposures. While the concentration-dependent effects, predominance of apoptosis over necrosis, and target-specific protein downregulation support the specificity of the observed responses, future validation using lower concentrations or physiologically relevant models (eg, 3D organoids, PDX models) would strengthen the translational relevance of these findings. Second, only the core targets and the most significantly enriched pathway (PI3K/AKT) were experimentally validated, and the contributions of other predicted targets and pathways remain to be elucidated. Third, the pharmacokinetic profile and long-term tissue distribution of hispidulin were not characterized, and these should be addressed in future work before clinical translation.

In summary, this study demonstrates that hispidulin exerts anti-HCC effects through a multi-target mechanism involving CDK1, SRC, CCNB1, and AURKB, with the PI3K/AKT pathway serving as a central signaling node among multiple enriched pathways. These findings provide a mechanistic basis for further development of hispidulin as a multi-target therapeutic candidate for HCC.

Conclusion

In conclusion, network pharmacology and bioinformatics analyses revealed that the ADME and toxicological profiles of hispidulin meet the criteria for drug-likeness. Hispidulin downregulates the expression of core target genes CDK1, CCNB1, SRC, and AURKB, and attenuates PI3K/AKT signaling among multiple enriched pathways, thereby interfering with the progression of hepatocellular carcinoma. These findings provide mechanistic evidence supporting hispidulin as a multi-target therapeutic candidate for HCC. As a naturally occurring flavonoid widely distributed in plants, hispidulin holds potential for further development into a targeted therapeutic agent for hepatocellular carcinoma.

Funding Statement

This work was supported by the Key Scientific Research Project of Jiangsu Health Commission ZD2021061.

Data Sharing Statement

The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding authors. E-mail: yanganyin1996@163.com.

Ethics Approval and Consent to Participate

This study followed the requirements of the National Standard of the People^’s Republic of China—Laboratory Animal—Guideline for Ethical Review of Animal Welfare (GB/T 35892-2018) to ensure the welfare of the laboratory animals, and was approved by the Suzhou Weiyuan Biotechnology Co., Ltd. Park Branch Animal Ethics Committee (Number: IA20251109).

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

This paper has been uploaded to ResearchSquare as a preprint: https://www.researchsquare.com/article/rs-7961105/v1. All authors declared no competing financial or other interests.