Strategic co-targeting of FOXM1 and EGFR-ERBB2 signaling in ovarian cancer

Abstract

Objectives: Epithelial ovarian cancer (EOC) is a challenging disease due to chemoresistance, and short-term recurrence. Chemoresistance to anti-cancer drugs substantially reduces survival in EOC. In this study, we showed that chemoresistance to lapatinib (a dual EGFR-ERBB2 kinase inhibitor) and thiostrepton (FOXM1 inhibitor), induced the epithelial-mesenchymal transition and stemness in OC cells. Aberrant expression of FOXM1 and EGFR-ERBB2 proteins are common in many solid cancers including serous EOC. However, the precise mechanism of FOXM1, EGFR-ERBB2 in the chemoresistance of OC remains unclear. Methods: We blocked FOXM1 and EGFR-ERBB2 signaling by either genetically or pharmacologically, then evaluated the effects of thiostrepton and lapatinib on OC proliferation, cell cycle changes, and migratory effects by trypan blue exclusion, 3-D spheroid formation ability, growth assays, confocal microscopy, cell cycle, apoptosis analysis, and in vivo tumor growth assay. Results: In our study, we identified the mechanism of how tumor cells relieve their dependency on EGFR-ERBB2 signaling and adopt alternative and sustainable signaling mechanisms like FOXM1 for resistance to ERBB-therapy and metastasis. To characterize the mechanism of lapatinib and thiostrepton resistance in OC cells, we developed lapatinib and thiostrepton resistant OC cells. We observed that long-term treatment of lapatinib decreased, the levels of pEGFR (Y1068) and pERBB2 (Y1248) proteins, whereas the levels of both phosphorylated and total protein of CDK1 and pPLK1 were upregulated. Strikingly, we found that FOXM1 level is upregulated in the cells are resistant to lapatinib. Next, we determined the nuclear level of FOXM1 which is important for the transcriptional activation of oncogenes. Importantly, we found an increased localization of FOXM1 in the nucleus in lapatinib resistant cells. In contrast, we found that the thiostrepton resistant OC cells express high levels of EGFR and ERBB2. In conjunction, we found that combined FOXM1 and EGFR-ERBB2 inhibition by the sub-lethal concentration of thiostrepton and lapatinib treatment reduced EGFR-ERBB2 and FOXM1 expression and induced apoptosis and inhibition of the colony-forming ability of tumor cells more than a single agent treatment. In addition, our Western blots, immunofluorescence, and immunohistochemistry assays in combination with in vivo studies demonstrated that the combined pharmacological inhibition of FOXM1 by thiostrepton reduced nuclear expression of the FOXM1 in lapatinib resistant cells while EGFR-ERBB2 inhibitor lapatinib downregulates the EGFR-ERBB2 in thiostrepton resistant cells. Taken together, our data demonstrated that the lapatinib resistant mechanism operates through FOXM1 activation likely through CDK1 or PLK1. To inhibit the levels of FOXM1, we used thiostrepton, which is capable of inhibiting the levels of FOXM1 in OC cells, and the cells which are resistant to thiostrepton express high levels of p-EGFR, p-ERBB2, and p-ERK2. In support to our results, we found that thiostrepton resistant cells express high levels of pEGFR and pERBB2, as well as high levels of EGF mRNA. Conclusions: Taken together, our data demonstrated that FOXM1 and EGFR-ERBB2 expression and activation provide a compensatory mechanism against EGFR/ERBB2 or FOXM1 therapy. Thus, inhibiting FOXM1 and EGFR/ERBB2 levels would provide an opportunity to sensitize OC cells resistant to FOXM1 or EGFR/ERBB2 therapy in OC patients.