MEK1 is associated with carboplatin resistance and is a prognostic biomarker in epithelial ovarian cancer

Abstract

Background: Primary systemic treatment for ovarian cancer is surgery, followed by platinum based chemotherapy. Platinum resistant cancers progress/recur in approximately 25% of cases within six months. We aimed to identify clinically useful biomarkers of platinum resistance.

Methods: A database of ovarian cancer transcriptomic datasets including treatment and response information was set up by mining the GEO and TCGA repositories. Receiver operator characteristics (ROC) analysis was performed in R for each gene and these were then ranked using their achieved area under the curve (AUC) values. The most significant candidates were selected and in vitro functionally evaluated in four epithelial ovarian cancer cell lines (SKOV-3-, CAOV-3, ES-2 and OVCAR-3), using gene silencing combined with drug treatment in viability and apoptosis assays. We collected 94 tumor samples and the strongest candidate was validated by IHC and qRT-PCR in these.

Results: All together 1,452 eligible patients were identified. Based on the ROC analysis the eight most significant genes were JRK, CNOT8, RTF1, CCT3, NFAT2CIP, MEK1, FUBP1 and CSDE1. Silencing of MEK1, CSDE1, CNOT8 and RTF1, and pharmacological inhibition of MEK1 caused significant sensitization in the cell lines. Of the eight genes, JRK (p = 3.2E-05), MEK1 (p = 0.0078), FUBP1 (p = 0.014) and CNOT8 (p = 0.00022) also correlated to progression free survival. The correlation between the best biomarker candidate MEK1 and survival was validated in two independent cohorts by qRT-PCR (n = 34, HR = 5.8, p = 0.003) and IHC (n = 59, HR = 4.3, p = 0.033).

Conclusion: We identified MEK1 as a promising prognostic biomarker candidate correlated to response to platinum based chemotherapy in ovarian cancer.

Figure 1

Overview of the study.

Figure 2

Carboplatin sensitivity and silencing of the candidate genes. Dose–response curves of each cell line against carboplatin, after 48 hours drug administration (A). Relative viability after 48 hours carboplatin administration and silencing of four genes compared to the negative control siRNA transfected groups in each of the four cell lines (mean with SEM) *p < 0.001, **p < 0.01 (B).

Figure 3

MEK1 inhibition with carboplatin treatment. Silencing of MEK1 significantly increases the ratio of the apoptotic cells, and decrease the ratio of the viable cells after 48 hours carboplatin treatment compared to the negative control siRNA transfected cells. *: p < 0.05 (A) Dose–response curves of SKOV-3 and CAOV-3 cell lines against the MEK1 inhibitor PD0325901 (B). Effects of 48 hour treatment with carboplatin and PD0325901 as single agents and in combination. SKOV-3: C1: 212 μM carboplatin, C2: 141 μM carboplatin, PD: 554 nM PD0325901 in SKOV3 cell line. CAOV-3: C1: 111 μM carboplatin, C2: 74 μM carboplatin, PD: 277 nM PD0325901 in CAOV-3 cell line (mean with SEM) *p < 0.0001 C).

Figure 4

Correlation between MEK1 expression and survival after platinum treatment in EOC patients. Expression measured by qRT-PCR: relapse-free survival of 34 patients with low and high MEK1 expressing tumors (A). Expression tested with IHC: overall survival of 59 independent patients with low and high MEK1 expression (B). Representative images of immunohistochemistry, low and high expression of MEK1 at low and high magnifications (C).