Screening Of Key Genes As Potential Biomarkers For Hepatoblastoma And Related Biological Functions Research

Objective: Hepatoblastoma(HB)serves as the most frequent hepatic tumor in children,and the prognosis for patients with advanced-stage or chemotherapy-refractory HB is still dismal.The purpose of the current study was to detect genes participating in HB growth using bioinformatics analysis and subsequent experimental verification.Methods:Based on GEO database(https://www.ncbi.nlm.nih.gov/geo/),we downloaded raw data of two microarray datasets(GSE131329 and GSE75271).After performing data merging and batch effect correction,we utilized the “limma” R package to detect differentially expressed genes(DEGs)between HB and normal liver tissues.Then,STRING online database(https://string-db.org)was utilized to construct Protein-protein interaction(PPI)network,followed by a better visualization through Cytoscape software.Subsequently,MCODE pluggin in the Cytoscape software was applied for the exploration of closely related sub-modules,followed by detection of critical genes in PPI network by Cyto Hubba pluggin.The “WGCNA” R package was utilized to perform weighted gene co-expression network analysis(WGCNA)to identify HB-associated key modules and critical genes.Genes from both PPI network and WGCNA analyses were intersected to get the final key genes.Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)analyses were utilized to obtain the most critical biological functions as well as related pathways involved in the growth of HB.According to Oncopression database,receiver operating characteristic(ROC)curve analysis and literature search results,the key genes were further verified.Lastly,we verify the biological functions of the key genes in HB cell lines in vitro using si RNA knockdown cellular models together with colony formation assays,CCK-8 assays,scratch assays,and Transwell invasion assays.Results: A total of 856 DEGs were detected between HB and normal liver tissues.Through GO and KEGG analyses,cell cycle phase transition and PI3K/AKT signaling were found to be in association with HB growth.Through 4 methods in Cyto Hubba plugin,CDK1,AURKA,CDC20,AURKB,CCNA2,and PLK1 were found to be the critical genes during PPI analysis.Based on the three highest MCODE scores,the most important three modules were identified.Module 1,with the highest MCODE score,was analyzed to identify cell cycle,DNA replication,and oocyte meiosis as the enriched biological functions.The blue module was obtained from WGCNA analysis and also applied for GO and KEGG analyses,resulting in similar enriched biological funcitions and pathways to those in the module 1.Five genes,including AURKA,CDK1,AURKB,CCNA2,and CDC20,were identified as critical genes using PPI and WGCNA analytic methods.Using ROC curve analytic method,literature search,and analysis of Oncopression database,CCNA2,CDK1,and CDC20 were revealed as the final genes to perform experimental verification.Knockdown experiement indicated that either gene of these 3 key genes knockdown resulted in the suppression of biological funcitions in HB cell lines.Conclusions: CCNA2,CDK1,and CDC20 were identified as noval promising biomarkers and therapeutic targets for HB diagnosis and therapy in the future.