Method for analyzing and identifying resveratrol-induced hepatocellular carcinoma apoptosis and autophagy key targets based on network pharmacology and clinical data

Abstract

The invention discloses a method for analyzing and identifying resveratrol-induced hepatocellular carcinoma apoptosis and autophagy key targets based on network pharmacology and clinical data, which comprises the following steps of: respectively acquiring resveratrol, hepatocellular carcinoma, apoptosis and autophagy targets, and acquiring intersection targets through Wehn analysis; screening core targets through a protein interaction network topology algorithm; through gene expression analysis, the correlation between the core target spot and the occurrence and development of the hepatocellular carcinoma is verified; through survival analysis, the correlation between the gene expression of the core target spot and the prognosis of the hepatocellular carcinoma patient is verified. Through network pharmacology and clinical data analysis, the resveratrol can regulate apoptosis and autophagy of hepatocellular carcinoma through six core targets with remarkable clinical correlation, such as TP53 and the like.

Description

Technical Field

[0001] This study belongs to the field of pharmacology and involves a screening method based on network pharmacology and clinical data analysis to systematically reveal the key molecular targets of resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma. Background Art

[0002] Liver cancer presents a significant challenge in global cancer prevention and control, consistently ranking among the top three causes of cancer mortality worldwide. Primary liver cancer primarily includes hepatocellular carcinoma (HCC) and intrahepatic bile duct carcinoma, with HCC accounting for approximately 80% of liver cancer cases worldwide. HCC is afflicted by numerous factors, is difficult to diagnose early, and has a high recurrence rate, necessitating the development of safe and effective treatment strategies.

[0003] Resveratrol (3,4′,5-trihydroxy-1,2-diphenylethylene, Resveratrol) is a natural polyphenol compound widely distributed in nature, found in over 100 plant species from 34 families, including grapes, peanuts, and blueberries. It has high value in food and pharmaceutical applications. Resveratrol exhibits multiple biological activities, including antioxidant, anti-inflammatory, metabolic regulation, neuroprotective, liver protective, and anti-tumor activities. Furthermore, resveratrol synergistically exerts its therapeutic effects on liver disease through its anti-inflammatory, antioxidant, and metabolic regulation activities, demonstrating unique advantages in the prevention and treatment of hepatocellular carcinoma.

[0004] Apoptosis and autophagy, two major forms of programmed cell death, regulate cell fate through different mechanisms. Activation of apoptosis can directly eliminate cancerous cells and exert anti-tumor effects. Under normal physiological conditions, autophagy protects cells by degrading damaged organelles and proteins; however, under conditions such as hypoxia, nutrient deficiency, or long-term stress, activation of autophagy may also induce cell death. Apoptosis and autophagy can be synergistically regulated through protein nodes such as p53, Bcl-2, and Beclin-1. Elucidating and targeting these interaction nodes will provide new insights into cancer treatments based on autophagy and apoptosis mechanisms.

[0005] Network pharmacology is a comprehensive research approach based on systems biology. It integrates multidimensional data on drugs, targets, diseases, and pathways to construct complex interaction networks to reveal the multi-target mechanisms of drug action and their systemic regulatory mechanisms. Network pharmacology emphasizes the synergistic effects of drugs on disease through multiple targets and pathways, revealing the global regulatory networks of drugs in complex diseases. This approach has demonstrated significant advantages in the analysis of the mechanisms of traditional Chinese medicine compounds and the study of natural products, and has become a vital tool in modern drug development. Although network pharmacology can predict potential drug targets and pathways, the clinical relevance of its results requires further validation. In recent years, with the rapid development of cancer genomics databases such as TCGA (The Cancer Genome Atlas) and UALCAN, clinical data analysis has become an important method for validating network pharmacology predictions. By analyzing the expression patterns of target genes in tumor tissues, their dynamic changes across different clinical stages, and their association with patient survival, targets with therapeutic potential can be identified, providing clinically relevant evidence for drug mechanism studies. Summary of the Invention

[0006] The present invention aims to provide a method based on network pharmacology and clinical data analysis to identify key targets for resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma. Using network pharmacology, the core targets were screened and their clinical relevance was verified through clinical data analysis, providing a theoretical basis for the treatment of hepatocellular carcinoma with resveratrol through apoptosis and autophagy pathways.

[0007] To achieve the above objectives, the technical solution of the present invention provides a method for predicting key targets of resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma based on network pharmacology and clinical data analysis, which specifically includes the following steps:

[0008] S1: Obtain information on resveratrol therapeutic targets

[0009] Search for "Resveratrol" in the Pubchem database (Public Chemical Database) to obtain the SMILES sequence number and molecular structure SDF file of resveratrol. Import the SMILES sequence number into the Swiss Target Prediction database, select the species as "Homo sapiens", perform target prediction, and screen out targets with probability>0. Import the molecular structure SDF file into the PharmMapper database, select the target type as "Human Protein Targets Only", perform target prediction, and remove non-Homo sapiens genes and incorrect targets. Search for "Resveratrol" in the ChEMBL database to obtain target data, perform standard naming through the UniProt database, and remove non-Homo sapiens genes and duplicates. Use the DrugBank database and TCMSP database to search for resveratrol targets, and perform standard naming through the UniProt database. Integrate all target data and remove duplicates to obtain resveratrol-related targets.

[0010] S2: Obtaining hepatocellular carcinoma target information

[0011] Using the keyword "Hepatocellular Carcinoma," we searched the GeneCards, OMIM, and DisGeNET databases to identify HCC-related targets. We selected targets with a relevance score in the top 5% for the GeneCards database and a score > 0.1 for the DisGeNET database. We integrated all target data, removed duplicates, and identified HCC-related targets.

[0012] S3: Obtaining apoptosis and autophagy target information

[0013] Apoptosis-related targets were obtained from the GeneCards database, and autophagy-related targets were obtained from the GeneCards database and the HADb database. Targets with the top 5% relevance scores were selected from the GeneCards database.

[0014] S4: Obtain intersection targets

[0015] The Wayne online analysis tool of the Micro-Informatics Online Platform (https: / / www.bioinformatics.com.cn / ) was used to create a Venn diagram of resveratrol, hepatocellular carcinoma, apoptosis, and autophagy targets to obtain intersection targets.

[0016] S5: Constructing protein interaction networks

[0017] The intersection targets were imported into the STRING database, the species was selected as "Homo sapiens", and the minimum interaction score was set to "medium confidence>0.4" for protein-protein interaction (PPI) analysis. The PPI network was constructed using Cytoscape_v3.10.0 software, where nodes represent intersection targets and edges represent the interactions between any two target proteins.

[0018] S6: Screening core targets in the PPI network

[0019] Core targets were screened using the Cytoscape plug-in cytoHubba. Five algorithms, namely "Degree," "DMNC," "Closeness," "Betweenness," and "Stress," were used to analyze the PPI network. The Degree algorithm assesses node importance by counting the number of direct connections. A greater number of connections indicates a more critical node in the network. The DMNC algorithm scores nodes based on the density of connections within their neighborhood. A higher DMNC value reflects the density of connections within the local network, suggesting that the node may play a core regulatory role in a specific functional module. The Closeness algorithm measures centrality by calculating the inverse of the sum of the shortest paths from a node to all other nodes. A higher Closeness value indicates that the node is more efficient in information transmission. The Betweenness algorithm measures the frequency with which a node serves as the shortest path in the network. A higher Betweenness value indicates that the node is more likely to play an irreplaceable role as a bridge connecting different functional modules. The Stress algorithm measures the node's transit load by calculating the total number of shortest paths within the node. A higher Stress value indicates that the node carries a heavier burden of information transmission within the network. The five algorithms comprehensively evaluated the different centrality characteristics of nodes in the PPI network, screened targets with scores above the mean, and used the Wayne online analysis platform to perform intersection analysis to finally obtain the core targets.

[0020] S7: Gene expression analysis of core targets

[0021] Using the Gene_DE module of the TIMER2 database, the expression differences of core genes in tumor tissues and adjacent normal tissues in the TCGA database were studied. The gene expression level was represented by log2 TPM (Transcripts Per Million). The significance of the expression difference was calculated by the Wilcoxon test. In addition, the UALCAN database analysis tool was used to mine the TCGA database information to study the expression of core genes in hepatocellular carcinoma tissues with different tumor grades. In UALCAN, the gene expression level was represented by TPM. The significance of the difference in expression levels was evaluated by the t-test (Student's t-test).

[0022] S8: Survival analysis of core targets

[0023] Using the Pan-Cancer analysis tool of the "RNA-seq" module in the Kaplan Meier Plotter database, based on the gene expression data of hepatocellular carcinoma patients in the TCGA database, the correlation between gene expression levels and the overall survival (OS) of patients was analyzed. The best cutoff point was automatically selected through the "Auto select best cutoff" option to distinguish the high-expression group and the low-expression group. The correlation between the expression of core genes and the survival of hepatocellular carcinoma patients was evaluated by Kaplan-Meier survival analysis. Cox Proportional Hazards Regression Analysis was used to evaluate the impact of gene expression on patient survival. Cox proportional hazards regression analysis measures the impact of the gene on the survival risk by calculating the Hazard Ratio (HR) and its 95% Confidence Interval (CI) for each core gene. An HR greater than 1 indicates an increased survival risk in the high-expression group, and an HR less than 1 indicates a decreased survival risk in the high-expression group. The significance of the difference in survival curves between the high-expression group and the low-expression group was determined by the log-rank test. Description of the drawings

[0024] Figure 1 Venn diagram of the intersection targets of hepatocellular carcinoma, resveratrol, autophagy, and apoptosis

[0025] Figure 2 Protein interaction network diagram of the intersection targets

[0026] Figure 3 Diagram of the screening results of core targets

[0027] Figure 4Boxplot of gene expression of core targets in hepatocellular carcinoma tissue and normal liver tissue

[0028] Figure 5 Boxplots of gene expression of core targets in hepatocellular carcinoma tissues of different tumor grades

[0029] Figure 6 Kaplan-Meier survival curve of core target gene expression DETAILED DESCRIPTION

[0030] In order to enable researchers in this field to better understand the technical solutions of the present invention, the following preferred embodiments are used in conjunction with the accompanying drawings to describe the technical solutions provided by the present invention in detail. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0031] Example

[0032] 1. Obtain information on resveratrol therapeutic targets

[0033] Search for "Resveratrol" in the Pubchem database to obtain the SMILES sequence number and molecular structure SDF file of resveratrol. Import the SMILES sequence number into the Swiss Target Prediction database, select the species as "Homosapiens", perform target prediction, and screen out targets with probability>0. Import the molecular structure SDF file into the PharmMapper database, select the target type as "Human Protein Targets Only", perform target prediction, and remove non-Homo sapiens genes and incorrect targets. Search for "Resveratrol" in the ChEMBL database to obtain target data, perform standard naming using the UniProt database, and remove non-Homo sapiens genes and duplicates. Use the DrugBank database and TCMSP database to search for resveratrol targets, and perform standard naming using the UniProt database. Integrate all target data and remove duplicates to obtain resveratrol-related targets.

[0034] 2. Obtaining hepatocellular carcinoma target information

[0035] Using the keyword "Hepatocellular Carcinoma," we searched the GeneCards, OMIM, and DisGeNET databases to identify HCC-related targets. We selected targets with a relevance score in the top 5% for the GeneCards database and a score > 0.1 for the DisGeNET database. We integrated all target data, removed duplicates, and identified HCC-related targets.

[0036] 3. Obtaining apoptosis and autophagy target information

[0037] Apoptosis-related targets were obtained from the GeneCards database, and autophagy-related targets were obtained from the GeneCards database and the HADb database. Targets with the top 5% relevance scores were selected from the GeneCards database.

[0038] 4. Obtain intersection targets

[0039] The Wayne analysis tool of the Weishengxin online platform was used to create a Venn diagram of resveratrol, hepatocellular carcinoma, apoptosis and autophagy targets to obtain intersection targets.

[0040] 5. Constructing protein interaction networks

[0041] The intersection targets were imported into the STRING database, the species was selected as "Homo sapiens", the minimum interaction score was set to "medium confidence>0.4", and PPI analysis was performed. The PPI network was constructed using Cytoscape_v3.10.0 software, where nodes represent intersection targets and edges represent the interactions between any two target proteins.

[0042] 6. Screening core targets in the PPI network

[0043] Core targets were screened using the Cytoscape plug-in cytoHubba. Five algorithms, namely "Degree," "DMNC," "Closeness," "Betweenness," and "Stress," were used to analyze the PPI network. The Degree algorithm assesses node importance by counting the number of direct connections. A greater number of connections indicates a more critical node in the network. The DMNC algorithm scores nodes based on the density of connections within their neighborhood. A higher DMNC value reflects the density of connections within the local network, suggesting that the node may play a core regulatory role in a specific functional module. The Closeness algorithm measures centrality by calculating the inverse of the sum of the shortest paths from a node to all other nodes. A higher Closeness value indicates that the node can more efficiently transmit information and has the potential to rapidly influence the entire network. The Betweenness algorithm measures the frequency with which a node serves as the shortest path in the network, specifically, how many pairs of nodes must pass through it to connect via the shortest path. A higher Betweenness value indicates that the node is more likely to play an irreplaceable role as a bridge in the interaction between different functional modules. The Stress algorithm measures the node's transit load by calculating the total number of shortest paths it is within. A higher Stress value indicates that the node carries a heavier burden of information transmission within the network. The five algorithms comprehensively evaluated the different centrality characteristics of nodes in the PPI network, screened targets with scores above the mean, and used the Wayne online analysis platform to perform intersection analysis to finally obtain the core targets.

[0044] 7. Gene expression analysis of core targets

[0045] The Gene_DE module of the TIMER2 database was used to investigate differential expression of core genes in the TCGA database between tumor tissues and adjacent normal tissues. Gene expression levels were expressed as log2 TPM. The significance of expression differences was calculated using the Wilcoxon test. Furthermore, the UALCAN database analysis tool was used to mine TCGA database information and investigate the expression of core genes in HCC tissues of different tumor grades. In UALCAN, gene expression levels are expressed as TPM. The significance of expression level differences was assessed using the t-test.

[0046] 8. Survival analysis of core targets

[0047] Using the Pan-Cancer analysis tool in the "RNA-seq" module of the Kaplan Meier Plotter database, based on the gene expression data of hepatocellular carcinoma patients in the TCGA database, the correlation between gene expression levels and the overall survival of patients was analyzed. Through the "Auto select best cutoff" option, the best cutoff point was automatically selected to distinguish the high-expression group and the low-expression group. The correlation between the expression of core genes and the survival of hepatocellular carcinoma patients was evaluated by Kaplan-Meier survival analysis. Cox proportional hazards regression analysis was used to evaluate the impact of gene expression on patient survival. Cox proportional hazards regression analysis measures the impact of a gene on the survival risk by calculating the hazard ratio and its 95% confidence interval for each core gene. An HR greater than 1 indicates an increased survival risk in the high-expression group, and an HR less than 1 indicates a decreased survival risk in the high-expression group. The significance of the difference in survival curves between the high-expression group and the low-expression group was determined by the log-rank test.

[0048] Results

[0049] 1. Results of target acquisition and screening of intersection targets

[0050] To obtain the targets of resveratrol regulating apoptosis and autophagy in hepatocellular carcinoma, 554 potential therapeutic targets of resveratrol were screened and determined from the SwissTargetPrediction, PharmMapper, ChEMBL, DrugBank, and TCMSP databases; 1341 hepatocellular carcinoma-related targets were screened and determined from the GeneCards, OMIM, and DisGeNET databases; 1093 apoptosis-related targets were screened and determined from the GeneCards database; 580 autophagy-related targets were screened and determined from the GeneCards and HADb databases. Venn analysis was performed on the four types of targets to obtain 55 intersection targets, and the results are as Figure 1 shown.

[0051] 2. Results of protein-protein interaction network construction and core target screening

[0052] The intersection targets were imported into the STRING database for PPI network analysis, and the PPI network was constructed using Cytoscape 3.10.0 software. The results are as Figure 2 shown. The network contains 55 nodes and 2220 edges. The nodes were visualized according to the Degree value from large to small: the node color gradually changes from dark red to yellow, and the position is arranged from the inner circle to the outer circle. Five topological parameters, namely "Degree", "DMNC", "Closeness", "Betweenness", and "Stress", were used for analysis to screen the core targets in the PPI network. The results are as Figure 3As shown in the figure. By Venn analysis, the intersection targets with scores higher than the mean were screened, and 16 core targets were obtained, including TP53, CDKN1A, MYC, KRAS, ERBB2, BCL2, BCL2L1, CASP3, CASP8, CASP9, MTOR, HIF1A, MAPK3, MAPK8, CTNNB1, and HSP90AA1.

[0053] 3. Results of gene expression analysis of core targets

[0054] To identify core targets that are significantly correlated with the occurrence and development of hepatocellular carcinoma, the expression data of 16 core target genes in normal liver tissue and hepatocellular carcinoma tissue were obtained from the TCGA database for analysis. The results are as Figure 4 shown. There were no significant differences in the expression levels of BCL2, HIF1A, and MAPK8 between hepatocellular carcinoma tissue and normal liver tissue; the expression levels of TP53, KRAS, ERBB2, BCL2L1, CASP3, CASP8, CASP9, MTOR, MAPK3, CTNNB1, and HSP90AA1 were significantly higher in hepatocellular carcinoma tissue than in normal liver tissue (P < 0.001, and the significance of the expression difference of KRAS and ERBB2 was P < 0.05); the expression level of MYC was significantly lower in hepatocellular carcinoma tissue than in normal liver tissue (P < 0.001). These target genes with significant differential expression may be closely related to the occurrence and development of hepatocellular carcinoma. Further analysis of the expression of core target genes in hepatocellular carcinoma tissue with different tumor grades showed the results as Figure 5 shown. Among hepatocellular carcinoma tissues with different tumor grades, 7 target genes, such as TP53, MYC, ERBB2, CASP3, CASP8, MAPK3, and HSP90AA1, showed significant differential expression (P < 0.05). These target genes may be closely related to the progression of hepatocellular carcinoma

[0055] 4. Results of survival analysis of core targets

[0056] To identify core targets that are significantly correlated with the prognosis of hepatocellular carcinoma patients, the survival analysis tool of the Kaplan Meier Plotter database was used to analyze the association between the expression level of target genes and the overall survival of patients. The results are as Figure 6As shown, the expression levels of MYC, BCL2, BCL2L1, and MTOR had no significant effect on the patient survival period. The overall survival of patients with high expression of TP53, CDKN1A, and ERBB2 was significantly longer than that of the low-expression group (P<0.05), suggesting that the high expression of these genes was beneficial to patient survival. On the contrary, the overall survival of patients with high expression of KRAS, CASP3, CASP8, CASP9, HIF1A, MAPK, MAPK8, CTNNB1, and HSP90AA1 was significantly shorter than that of the low-expression group (P<0.05), indicating that the high expression of these genes was not conducive to the prognosis of patients.

[0057] Conclusion

[0058] Through network pharmacology screening, 16 core targets of resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma were obtained in the technology of the present invention. The clinical data analysis results of the core targets showed that there were significant differential expressions of TP53, ERBB2, CASP3, CASP8, MAPK3, and HSP90AA1 between hepatocellular carcinoma tissues and normal liver tissues, as well as between hepatocellular carcinoma tissues with different tumor grades (P<0.05), and their expression levels were also significantly correlated with the overall survival of hepatocellular carcinoma patients (P<0.05). This result suggests that the 6 targets such as TP53 are closely related to the occurrence, progression, and prognosis of hepatocellular carcinoma, and have prominent clinical value.

[0059] As can be seen from the above embodiments, the present invention provides a method for identifying resveratrol-induced apoptosis and autophagy targets in hepatocellular carcinoma based on network pharmacology and clinical data analysis, including the following steps: obtaining resveratrol, hepatocellular carcinoma, apoptosis, and autophagy targets; obtaining intersection targets; constructing a protein-protein interaction network using the intersection targets and screening out core targets; performing clinical data analysis on the core targets to verify their clinical relevance. The present invention applies network pharmacology and clinical data analysis methods to preliminarily reveal the key target of resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma, providing guidance for the future treatment of hepatocellular carcinoma by resveratrol through apoptosis and autophagy pathways.

[0060] The above description is only a preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for identifying key targets of resveratrol-induced apoptosis and autophagy in hepatocellular carcinoma based on network pharmacology and clinical data analysis, comprising the following steps:

1. Obtain resveratrol treatment target information Search with "Resveratrol" as the keyword, and obtain resveratrol treatment targets from SwissTargetPrediction, PharmMapper, ChEMBL (Chemical European Molecular Biology Laboratory), DrugBank, and TCMSP (Traditional Chinese Medicine Systems Pharmacology) databases.

2. Obtain hepatocellular carcinoma target information Search with "Hepatocellular Carcinoma" as the keyword, and obtain hepatocellular carcinoma targets from GeneCards, OMIM (Online Mendelian Inheritance in Man), and DisGeNET (Disease Gene Network) databases.

3. Obtain apoptosis and autophagy target information Search with "Apoptosis" as the keyword to obtain apoptosis targets from the GeneCards database; search with "Autophagy" as the keyword to obtain autophagy targets from the GeneCards and HADb (Human Autophagy Database) databases.

4. Obtain intersection targets Perform Venn intersection analysis on resveratrol treatment targets, hepatocellular carcinoma targets, apoptosis and autophagy targets to obtain the intersection targets of the four.

5. Construct a protein-protein interaction network Import the intersection targets into the STRING (Search Tool for the Retrieval of Interacting Genes / Proteins) database for protein-protein interaction (PPI) network analysis. Use Cytoscape_v3.10.0 software to construct a PPI network.

6. Screen core targets of the PPI network Use the CytoNCA plugin of Cytoscape_v3.10.0 software to analyze the topological parameters of the PPI network, and then use the CytoHubba plugin to screen the core targets of the PPI network.

7. Clinical data analysis and verification The TIMER2 (Tumor Immune Estimation Resource 2) and UALCAN (UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses) databases were used to analyze the expression of the core target genes of the PPI network to verify their correlation with the occurrence and development of hepatocellular carcinoma; the Kaplan-Meier Plotter database was used for the survival analysis of the core targets to verify the correlation between the gene expression of the core targets and the prognosis of hepatocellular carcinoma patients.

Images