O4.03 * GLIOBLASTOMA STIMULATED ANGIOGENESIS IS MEDIATED BY MMP-3 ACTIVATED MMP-1 VIA BRAIN ENDOTHELIAL CELL ASSOCIATED PROTEASE ACTIVATED RECEPTOR 1 (PAR1) AND IS ENHANCED UNDER HYPOXIC CONDITIONS

Abstract

Glioblastoma Multiforme is a highly aggressive brain tumour known for its heterogeneity and its vascularity. Published reports using in vivo and in vitro glioma models reveal that Matrix metalloproteinase-1 (MMP-1) over-expression induces angiogenesis. In other studies the involvement of the Protease Activated Receptor 1 (PAR1) in MMP-1's effect on tumour vessel formation has been suggested. MMP-1 expression is localised to glioma and endothelial cells with PAR1 expression mainly localised to endothelial cells in human glioma biopsies. MMP-3, also expressed in gliomas, is an activator of pro-MMP-1. Herein, we focused on the effects of MMP-3 activated MMP-1 and conditioned media from glioma cell lines cultured under hypoxic conditions to begin to examine mechanisms of MMP-1 induction of angiogenesis in an in vitro model of brain endothelial cell tube formation. Human brain endothelial cells, activated MMP-1, inhibitors (PAR1 and MMP-1) were used to investigate the possibility that MMP-1 acts through PAR1 to induce angiogenesis. This was accessed by measuring brain endothelial tube length, number of tube projections and the number of branch points. Conditioned media from glioma cell lines (SNB-19 and UP-007) cultured under either normoxic or hypoxic conditions were also tested in the absence/presence of inhibitors. PAR1 activation was examined by endothelial intracellular calcium release using live cell imaging following addition of MMP-1. Exogenous addition of MMP-3 activated MMP-1 to brain endothelial cells lead to a dose dependant increase in number of branch points and tube length (P ≤ 0.05). MMP-1 inhibition prevented this increase and inclusion of a PAR1 inhibitor resulted in preventing MMP-1's effect, suggesting MMP-1 may function through PAR1 to drive angiogenesis. Endothelial cells treated with conditioned media obtained from glioma cell lines resulted in a significant increase in all measures of tube formation, with the most significant effect being on tube number using media from glioma cells grown under hypoxia (P ≤ 0.05). MMP-1 and PAR1 inhibition blocked these effects. Although it is well documented that MMP-1 is over-expressed in high grade gliomas, it has been unclear what role MMP-1 plays as its ‘normal’ ECM substrates are not typically found in gliomas or normal brain. Results from this study extend earlier data showing in vivo induction of angiogenesis and highlight a potential signaling pathway via PAR1 activation in glioma pathology.