Identification of Novel Drug Candidate for Epithelial Ovarian Cancer via In Silico Investigation and In Vitro Validation

Abstract

Epithelial ovarian cancer (EOC) has a poor prognosis and high mortality rate; patients are easy to relapse with standard therapies. So, there is an urgent need to develop novel drugs. In this study, differentially expressed genes (DEGs) of EOC were identified in The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Enrichment and protein–protein interaction (PPI) analyses were performed. The drug candidate which has the possibility to treat EOC was predicted by Connectivity Map (CMAP) databases. Moreover, molecular docking was selected to calculate the binding affinity between drug candidate and hub genes. The cytotoxicity of drug candidates was assessed by MTT and colony formation analysis, the proteins coded by hub genes were detected by Western blots, and apoptosis analysis was evaluated by flow cytometry. Finally, 296 overlapping DEGs (|log 2 fold change|>1; q-value <0.05), which were principally involved in the cell cycle (p < 0.05), and cyclin-dependent kinase 1 (CDK1) were screened as the significant hub gene from the PPI network. Furthermore, the 21 drugs were extracted from CMAPs; among them, piperlongumine (PL) showed a lower CMAP score (-0.80, -62.92) and was regarded as the drug candidate. Furthermore, molecular docking results between PL and CDK1 with a docking score of –8.121 kcal/mol were close to the known CDK1 inhibitor (–8.24 kcal/mol). Additionally, in vitro experiments showed that PL inhibited proliferation and induced apoptosis via targeting CDK1 in EOC SKOV3 cells. Our results reveal that PL may be a novel drug candidate for EOC by inhibiting cell cycle.