Differentiation of Human Glioblastoma Stem Cells as a Prospective Tool to Find New Biomarkers and Molecular Targets: A Proteomic and Lipidomic Approach

Although glioblastoma stem cells (GSCs) constitute a small subpopulation of this highly heterogeneous brain cancer, they are widely recognised as key drivers of tumour recurrence after surgical resection and resistance to radiotherapy and chemotherapy. Differentiation of GSCs has, therefore, emerged as a strategy to increase glioblastoma susceptibility to therapeutic and surgical interventions. Cellular and molecular heterogeneity of glioblastoma limits the effectiveness of any treatment that relies on single molecules identified as disease-specific biomarkers. In this context, lipidomic and proteomic profiles constitute a source of information that can be translated into the identification of clinically relevant diagnostic and prognostic biomarkers, and novel therapeutic targets towards eradication of this lethal disease. Our initial studies demonstrated that BMP4 induced differentiation of the human glioblastoma stem cells G144 with approximately a 30-fold increase in expression of the astrocytic marker GFAP, which was coupled with a reduction in cell proliferation (33% decrease in EDU incorporation), invasion (61% decrease) and tumoursphere growth (53% decrease in area), and enhanced senescence-associated β-galactosidase activity (5.9-fold increase). Lipidomic and proteomic analysis (using data from 3 and 4 independent experiments, respectively) was then performed, allowing the identification of specific lipids and proteins exclusively detected or predominant in stem or differentiated G 144 cells. Importantly, our findings, which revealed a 6-fold increase in GFAP expression upon downregulation of a specific protein detected only in stem cells (SPRY4), pave the way for the employment of the RNAi-mediated knockdown strategy as a mechanistically distinct alternative to the use of BMP4, and highlight this protein as a candidate regulator associated with the G144 stem-like phenotype.