Research On The Mechanism Of Hedyotis Diffusa In The Treatment Of Gastrointestinal Cancer Based On Network Pharmacology

Gastric cancer(GC)is the fifth most commonly diagnosed cancer worldwide and the third leading cause of cancer death worldwide.Colorectal cancer(CRC)is the third most commonly diagnosed cancer and the second leading cause of cancer death worldwide.Traditional Chinese medicine(TCM)has been widely applied to clinically treat tumor,with great advantages and potentials.TCM exhibits therapeutic efficacy by the synergistic effects of multi-component,multi-target and multi-pathway,so it is relatively difficult to analyze the complicated mechanism of TCM only by traditional experimental approaches.The active components,targets and therapeutic mechanism of most TCMs have not been fully elucidated,which hinders the application of TCM and TCM-based drug discovery.Hedyotis diffusa Willd.(HDW)is an annual herb of the Rubiaceae family,which possesses the functions of clearing heat and toxic materials,expelling carbuncle and disintegrating a mass,and inducing urination and removing dampness.Clinically,the herb has often been applied as a critical element in many TCM formulae for the treatment of various cancers,including GC and CRC.Although the multiple anti-cancer activities of HDW have been widely reported,the molecular mechanism of HDW on GC and CRC remain largely unclear.Network pharmacology is based on disease-gene-target-drug interaction networks,which could illustrate the action of drugs on the biological networks of body systems from a holistic perspective and has been widely used in exploring the mechanism of TCM.ObjectiveTo discover the potential prognostic biomarkers for GC and CRC by the integrated gene expression profiling analysis.To systematically investigate the mechanism of HDW on GC and CRC by the network pharmacology method.Methods1.Integrated gene expression profiling analysisNine GC microarray datasets were downloaded from the GEO database.For each of them,the differential expression analysis was performed by the limma package in R.The nine datasets were integrated with the RobustRankAggreg package to detect consistent differentially expressed genes(DEGs).The mRNA sequencing data and clinical information of GC patients were downloaded from the TCGA database,and the edgeR package was used to screen DEGs.The consistent DEGs in the nine microarray datasets were intersected with the DEGs in the TCGA dataset,and the eventually consistent DEGs between the microarray and sequencing data were reserved for further study.The protein-protein interaction(PPI)information of the DEGs was acquired from the STRING database,and the PPI network was constructed with the Cytoscape software.The MCODE plug-in in the Cytoscape software was used to screen hub clustering modules in the network.The gene ontology(GO)enrichment analysis was performed with DAVID,and the Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway enrichment analysis was performed with the clusterProfiler package.After the survival information and gene expression data in the TCGA GC dataset were merged,the univariate Cox regression analysis was performed with the survival package and the genes with P value<0.05 were further used to conduct the multivariate Cox regression analysis.The survival-related linear risk assessment model was built based on the selected gene expression profiles and regression coefficients,and the risk score for each sample was calculated.The Kaplan-Meier method with the log-rank test was used to assess the difference of overall survival between high-and low-risk groups.The time-dependent receiver operating characteristic(ROC)curve was used to assess the predictive accuracy of this prognostic model for time-dependent cancer death.Six CRC microarray datasets were downloaded from the GEO database.For each of them,the differential expression analysis was performed by the limma package.The six datasets were integrated with the RobustRankAggreg package to detect consistent DEGs.The mRNA sequencing data and clinical information of patients with colorectal adenocarcinoma were downloaded from the TCGA database,and the edgeR package was used to screen DEGs.The consistent DEGs in the six microarray datasets were intersected with the DEGs in the TCGA dataset,and the eventually consistent DEGs between the microarray and sequencing data were reserved for further study.The GSEA for the seven datasets was respectively performed by the clusterProfiler package,under the functional annotations of the KEGG database.After the survival information and gene expression data in the TCGA CRC dataset were merged,the univariate Cox regression analysis was performed with the survival package.The genes with P value<0.05 were further used to conduct the LASSO Cox regression analysis by the glmnet package.The survival-related genes identified by the LASSO Cox regression analysis were finally used to perform the multivariate Cox regression analysis.The survival-related linear risk assessment model was built based on the selected gene expression profiles and regression coefficients,and the risk score for each sample was calculated.The Kaplan-Meier method with the log-rank test was used to assess the difference of overall survival between high-and low-risk groups.The time-dependent ROC curve was used to assess the predictive accuracy of this prognostic model for time-dependent cancer death.2.Network pharmacology analysisThe herbal compounds of HDW were acquired from TCMSP and TCMID,and the targets of these chemicals were predicted by SuperPred.The GC-related genes were acquired from TTD,OMIM,PharmGKB and DigSee.The PPI information was acquired from the STRING database.The compound-compound target network,PPI network of compound targets,PPI network of GC-related genes and PPI network of compound target-GC target were constructed with the Cytoscape software.The topological features of these interaction networks were calculated.The MCODE plug-in in the Cytoscape software was used to screen hub clustering modules in the PPI network of compound target-GC target.GO and KEGG pathway enrichment analyses for the genes in the hub modules were performed with DAVID.The herbal compounds of HDW were acquired from TCMSP,TCMID and TCM Database@Taiwan,and the targets of these chemicals were predicted by PharmMapper.The CRC-related genes were acquired from DisGeNET.The PPI information was acquired from the STRING database.The compound-compound target network,PPI network of CRC-related genes and compound-CRC target PPI network were constructed with the Cytoscape software.The topological features of these interaction networks were calculated to discover the hub targets regulated by HDW.GO and KEGG pathway enrichment analyses for the hub targets were performed with DAVID.Results1.The result of the integrated gene expression profiling analysis for GC and CRCA total of 411 DEGs were obtained after the integrated analysis of the nine GC microarray datasets,and 4623 DEGs were obtained from the TCGA GC dataset.After intersecting these DEGs,we finally obtained 268 consistent DEGs(149 down-regulated and 119 up-regulated genes).The PPI network and module analyses showed three hub clustering modules and nine hub genes(TOP2A,COL1A1,COL1A2,NDC80,COL3A1,CDKN3,CEP55,TPX2 and TIMP1).The functional enrichment analysis showed that the down-regulated genes were mainly involved in various metabolic processes,including the metabolism of xenobiotics,cofactors,vitamins,amino acids and carbohydrates.For the up-regulated genes,they were significantly enriched in cancer-related pathways,such as ECM-receptor interaction,PI3K-Akt signaling pathway and Toll-like receptor signaling pathway.The survival analysis showed a nine-gene prognostic signature,in which COL8A1,SMPD3 and PLEKHS1 were protective prognostic genes,while CST2,AADAC,SERPINE1,ASPN,ITGBL1 and MAP7D2 were risky prognostic genes.A total of 990 DEGs were obtained after the integrated analysis of the six CRC microarray datasets,and 4131 DEGs were obtained from the TCGA CRC dataset.After intersecting these DEGs,we finally obtained 885 consistent DEGs(458 down-regulated and 427 up-regulated genes).The GSEA showed that 32 pathways(nine activated and 23 suppressed pathways)were found in three or more than three datasets.The survival analysis showed a seven-gene prognostic signature,in which AXIN2,CXCL1,ITLN1,CPT2 and CLDN23 were protective prognostic genes,while TIMP1 and LZTS3 were risky prognostic genes.2.The result of the network pharmacology analysis for the mechanism of HDW on GC and CRCIn the network pharmacology analysis for the mechanism of HDW on GC,carbonic anhydrase isoenzymes,p53 and PIK3CA might be the key targets of HDW.HDW showed a close relation with the proteins related to cell cycle,apoptosis and angiogenesis,such as CDK2,p27Kip1,cyclin D1,cyclin B1,cyclin A2,p53,AKT1,BCL2,MAPKl,VEGFA and PIK3CA.In the functional enrichment analysis,the hub proteins regulated by HDW were significantly enriched in nucleotide excision repair,apoptosis,cell cycle,PI3K-Akt-mTOR signaling pathway,VEGF signaling pathway and Ras signaling pathway.Compared with the result of the KEGG pathway enrichment analysis in the integrated gene expression profiling analysis for GC,we found five consistent pathways,namely,cell cycle,PI3K-Akt signaling pathway,FoxO signaling pathway,focal adhesion and TNF signaling pathway.In the network pharmacology analysis for the mechanism of HDW on CRC,we showed ten potential hub targets of HDW on CRC,namely,HRAS,PIK3CA,KRAS,p53,APC,BRAF,GSK3B,CDK2,AKT1 and RAF1.The GO enrichment analysis showed that the hub targets regulated by HDW were significantly enriched in regulation of peptidyl-serine phosphorylation,ERBB2 signaling pathway and Ras protein signal transduction.The KEGG pathway enrichment analysis showed that the hub targets regulated by HDW were significantly enriched in colorectal cancer,pathways in cancer,PI3K-Akt signaling pathway and MAPK signaling pathway.Compared with the result of the GSEA in the integrated gene expression profiling analysis for CRC,we found four consistent pathways,namely,MAPK signaling pathway,regulation of actin cytoskeleton,Rap1 signaling pathway and insulin signaling pathway.ConclusionThe present study comprehensively applied the network pharmacology and the integrated gene expression profiling analysis to systematically disclose the active compounds,targets and pathways of HDW on GC and CRC.This study preliminary suggests th at the mechanism of HDW on GC might be mainly attributed to its synergistic regulation on the signaling pathways associated with apoptosis,cell cycle,cell differentiation,cell proliferation,cell migration,cell invasion and angiogenesis.This study also preliminary suggests that the mechanism of HDW on CRC might be mainly attributed to its synergistic regulation on the commonly mutated genes in patients with CRC.The study would be beneficial for providing clues to understand and evaluate the synergetic effects of TCM in controlling complex diseases and for facilitating the application of network pharmacology in exploring the pharmacological mechanism of anticancer TCMs.However,the mechanism of HDW on GC and CRC still needs to be further investigated with more biological experiments since this study was conducted based on data analysis.