Epithelial-to-mesenchymal transition markers are differentially expressed in epithelial cancer cell lines after everolimus treatment

Phenotypic changes on central nervous system (CNS) tumor cell lines cultured in vitro 2D and 3D models and treated with cisplatin

Despite aggressive therapy, most patients with brain tumors present disease relapse due to the cellular and molecular nature of these tumors. One of the models that best explains the heterogeneity observed in CNS tumors is the presence of cancer stem cells (CSCs). In this paper, we evaluated platinum-based response in brain tumor U-87 MG, LN-18, and KELLY cell lines cultured in monolayer (2D) and neurosphere (CSC enrichment- 3D) models. We evaluated mRNA expression of heat shock proteins (HSPA1A, HSPB1, HSPA1AL, TRAP1, and HSPD1), and DNA repair gene ERCC1. Changes in cell cycle and glycosylation profile were assessed by flow cytometry. After treatment with cisplatin, we found that the mRNA expression of HSPs markedly increased in the U-87 MG and LN-18 neurosphere cells. In KELLY monolayer cells, cisplatin induced upregulation of all genes. In KELLY neurosphere cells, only the HSPA1A, HSPB1, TRAP1, and HSPD1 genes were upregulated. The proportion of cells in the G₀/G₁ phase was significantly higher in U-87 MG neurosphere cisplatin-treated cells. A trend towards a greater proportion of cells in the S phase of U-87 MG monolayer cisplatin-treated cells was also observed. On the other hand, a significant decrease in the number of cells in the S phase and an increase in G₂/M was observed in LN-18 monolayer cisplatin-treated cells. Glycosylation analysis using lectins showed a higher surface binding for PNA in the U-87 MG treated monolayer and a lower binding for Concanavalin A in the treated neurospheres. The binding of Isolectin GS-IB4, GSII, and SBA in KELLY monolayer cisplatin-treated cells was lower whereas the proportion of cells labeled with Concanavalin A was higher. In the KELLY neurosphere cisplatin-treated cells, the binding of Concanavalin A was lower than nontreated cells. Thus, our findings strongly supported the idea that definitions of phenotypic characteristics may help to establish better therapeutic strategies for brain tumors.