The neuroepithelial origin of ovarian carcinomas explained through anepithelial-mesenchymal-ectodermal transition enhanced by cisplatin

Acquired resistance to platinum-derived cytostatics poses major challenges in ovarian carcinoma therapy. In this work, we show a shift in the epithelial-mesenchymal transition (EMT) process towards an “ectodermal” conversion of ovarian carcinoma cells in response to cisplatin treatment, a progression we have termed epithelial-mesenchymal-ectodermal transition (EMET). EMET appears to occur via the classical EMT as judged by a) the downregulation of several epithelial markers and b) upregulation of Vimentin, accompanied by various embryonal transcription factors and, importantly, a plethora of neuronal markers, consistent with ectodermal differentiation. Moreover, we isolated cells from ovarian carcinoma cultures exhibiting a dual neural/stemness signature and multidrug resistance (MDR) phenotype. We also found that the epithelial cells differentiate from these neural/stem populations, indicating that the cell of origin in this tumor must in fact be a neural cell type with stemness features. Notably, some transcription factors like PAX6 and PAX9 were not localized in the nucleoplasm of these cells, hinting at altered nuclear permeability. In addition, the neuronal morphology was rapidly established when commercially available and primary ovarian carcinoma cells were cultured in the form of organoids. Importantly, we also identified a cell type in regular ovarian tissues, which possess similar neural/stemness features as observed in 2D or 3D cultures. The signature of this cell type is amplified in ovarian carcinoma tumors, suggesting a neuroepithelial origin of this tumor type. In conclusion, we propose that ovarian carcinomas harbor a small population of cells with an intrinsic neuronal/stemness/MDR phenotype, serving as the cradle from which ovarian carcinoma evolves. Overall design: SKOV3WT cells were incubated in DMEM medium until the cells reached about 70 % of confluence. At this point medium was replaced with DMEM containing 1 µg/ml cisplatin and incubated for 72 h. After this period, cells were renamed as SKOV3 cp. This experiment was repeated several time to get at least 3 biological replicates. Primary ovarian cancer cell line OC236 revealed a subpopulation of cells with a nested morphology, which are EpCAM negative but positive for a plethora of neuronal markers. In order to isolate such cells, we employed MicroBeads directly conjugated to EpCAM antibody (Cat. # 130–061-101,Miltenyi Biotec, Bergisch Gladbach, Germany). Briefly, cells were counted and resuspended in a separation buffer (PBS, pH 7.2, 0.5%BSA, and 2 mM EDTA), and incubated with the appropriate amount of FcR Blocking Reagent (Cat. # 130–059-901 Miltenyi Biotec) andCD326 (EpCAM) MicroBeads for 30 min at 4°C and then separated using MS MACS® Columns and a MiniMACS™ Separator. The eluted fraction was subjected to a second round of separation and positive and negative fractions were collected. The purity of these fractions was evaluated by flow cytometry. This isolation was done three times.