Transcriptome analysis of newly established carboplatin-resistant ovarian cancer cell model reveals genes shared by drug resistance and drug-induced EMT.

ABSTRACT

BACKGROUND: In ovarian cancer (OC) therapy, even initially responsive patients develop drug resistance. METHODS: Here, we present an OC cell model composed of variants with differing degrees of acquired resistance to carboplatin (CBP), cross-resistance to paclitaxel, and CBP-induced metastatic properties (migration and invasion). Transcriptome data were analysed by two approaches identifying differentially expressed genes and CBP sensitivity-correlating genes. The impact of selected genes and signalling pathways on drug resistance and metastatic potential, along with their clinical relevance, was examined by in vitro and in silico approaches. RESULTS: TMEM200A and PRKAR1B were recognised as potentially involved in both phenomena, also having high predictive and prognostic values for OC patients. CBP-resistant MES-OV CBP8 cells were more sensitive to PI3K/Akt/mTOR pathway inhibitors Rapamycin, Wortmannin, SB216763, and transcription inhibitor Triptolide compared with parental MES-OV cells. When combined with CBP, Rapamycin decreased the sensitivity of parental cells while Triptolide sensitised drug-resistant cells to CBP. Four PI3K/Akt/mTOR inhibitors reduced migration in both cell lines. CONCLUSIONS: A newly established research model and two distinct transcriptome analysis approaches identified novel candidate genes enrolled in CBP resistance development and/or CBP-induced EMT and implied that one-gene targeting could be a better approach than signalling pathway inhibition for influencing both phenomena.