| Glioblastoma Multiforme (GBM) is a deadly brain malignancy. Despite surgical resection and aggressive concomitant chemoradiotherapy, median survival is only 14.6 months. A major impeding factor in designing effective therapies against glioblastoma (GBM) is its extensive intratumoral molecular and microenvironmental heterogeneity. Glioblastoma stem cells (GSCs) reside at the apex of a cellular hierarchy in GBM. Due to their highly tumorigenic nature and their resistance to existing therapies, GSCs represent critical therapeutic targets. How GSCs adapt to variable microenvironmental conditions within tumors is not known. I used genetically modified primary human GBM samples to identify heterogeneous GSC populations within any given GBM. These distinct populations, defined by their varying dependence on Notch signaling and expression levels of the CD133 glycoprotein at the cell surface, differ in their differentiation capacities, angiogenic profiles, metabolic preferences and growth-promoting signaling mechanisms, thereby enabling adaptive tumor growth in a range of tumor microenvironments. Furthermore, the characterization of these GSC subpopulations led to the identification of a novel signaling mechanism required for tumor growth in hypoxic conditions. This mechanism is mediated by GPR133, an orphan adhesion-type G-protein coupled receptor, which is transcriptionally activated by Hif1? and elevates intracellular cyclic AMP. Finally, I extended my studies to implement a viral gene delivery approach to selectively transduce GSCs expressing CD133 on the cell surface, as a paradigm for specific GSC targeting in the laboratory and potentially the clinic. In summary, my results demonstrate that the tumor microenvironment shapes the metabolic and signaling heterogeneity in the GSC population to ensure tumor growth. I propose that future therapies aiming to target the GSC population take into account its heterogeneity at the cellular and molecular level. My results suggest that, in order to achieve successful therapies, combinatorial approaches targeting heterogeneous GSCs populations are required. |
