Implication des vésicules extracellulaires des cellules initiatrices tumorales dans l’augmentation de la perméabilité vasculaire du glioblastome

摘要

Brain microvessels are characterized by specific structure and organization within the neurovascular unit. Through highly selective endothelial junctions, the blood-brain barrier (BBB) controls exchanges of cells, fluids, plasmatic proteins and metabolites between blood and the cerebral compartment. VE-cadherin, a transmembrane protein of endothelial junctions, is of most importance in the vascular integrity. Indeed, its destabilization leads to BBB weakening and also breaking in some pathology. Glioblastoma is a highly aggressive brain tumour characterized by a high vascularization rate and abnormal vascular permeability. These properties promote in turn perivascular œdema, harmful for the patient. Since the last decade, a growing number of studies link glioblastoma stem-like cell (GSC) population to the initiation, recurrence and aggressiveness of such cancer. Interestingly, GSCs are located within the vascular niche, a specific microenvironment where they survive, communicate and exchange factors with the microvascular endothelium. On the base of a coculture model between patient-derived GSCs and brain microvascular endothelial cells which recapitulate BBB properties, our laboratory has focused on Semaphorin 3A (Sema3A). Sema3A is a GSC secreted protein and acts through its coreceptor Neuropilin-1 (Nrp-1) which in turn destabilizes VE-cadherin and promotes vascular permeability. During my thesis, we have identified and characterized Sema3A at the membrane of GSC secreted extracellular vesicles (EVs). A growing number of studies highlight EVs as important actors of tumour biology, in this way we have demonstrated that GSC-derived EVs can be uptake by endothelial cells and modulate their intrinsic properties. Through original in vivo models in combination with genetic (RNA interference) and pharmacologic strategies (humanised blocking antibodies), we have demonstrated that EV-carried Sema3A acts specifically through endothelial cells Nrp-1 to promote permeability. Furthermore, in orthotopic GSC xenograft we have identified a significant increase in the Sema3A EV-fraction collected from peripheral blood. Interestingly, similar results were obtained from newly diagnosed glioblastoma blood samples. Moreover, Sema3A from this fraction is a potent propermeability factor that can act at distance through Nrp-1 both in vitro and in vivo. Altogether, our results suggest that EV-carried Sema3A orchestrates loss of barrier integrity in glioblastoma and may be of interest for prognostic purposes.