Under normal conditions, little or no SCF expression is detectable in normal cerebrum; however we found it to be expressed at high levels both in glioma cells lines and in gliomas when compared to non-tumor brain. Additionally, there was a statistically significant higher level of SCF expression in high-grade gliomas compared to low-grade gliomas. Since high-grade gliomas are characterized by a much greater amount of tumor-associated angiogenesis compared to low-grade gliomas, the positive correlation of SCF expression with increasing glioma grade is consistent with a potential role for SCF in glioma-associated angiogenesis. We have demonstrated that the SCF receptor, c-Kit, is expressed on the surface of all endothelial cells (ECs) examined and that exposure of BMVEC-b, HUVEC and HMVEC-d in basal medium to SCF resulted in thymidine incorporation and cellular proliferation in all 3 EC lines in a dose-dependent manner even at low concentrations in the absence of other cytokines such as VEGF. SCF also induced EC migration and differentiation in an in vitro wound healing assay and capillary tube formation assay. These data demonstrated the ability of SCF to induce proliferation, migration and differentiation of BMVEC-b in vitro. We next subcutaneously implanted Matrigel impregnated with SCF, b-FGF (positive control) or vehicle alone into the adult SCID mice. The data obtained demonstrated that SCF can promote angiogenesis in vivo. By a similar technology we also demonstrated that suppression of SCF in glioma cells results in significant inhibition of glioma-induced angiogenesis in vivo. We next evaluated whether suppression of SCF would effect the survival of animals with intracranial gliomas. U373/as-SCF or U373/vector cells were stereotactically implanted to the cerebral subcortex of adult athymic nu/nu mice. Log-rank analysis of the Kaplan-Meier survival curves demonstrated a significant survival advantage for the U373/as-SCF bearing mice compared to the U373/vector control bearing animals (p less than 0.05), despite the fact that the growth rate of both cells types in vitro was identical. To confirm these results in actual tumor samples, immunohistochemical analysis of multiple surgical specimens from patients with glioblastoma revealed profound expression of SCF in cerebral cortex infiltrated by glioma cells secondary to both tumor-and neuronal-associated SCF expression. In summary, SCF expression appears to reside most prominently in the invasive front of the infiltrating glioma, suggesting its roles in the tumor progression. Given our data demonstrating the importance of SCF in tumor and host cell-induced angiogenesis, we hypothesize that a previously unrecognized, but major anti-tumor mechanism of Gleevec may be as an anti-angiogenic agent through its ability to potently inhibit c-kit signaling. We have therefore embarked on a series of in vivo experiments to look at the effects of Gleevec on glioma-mediated angiogenesis in our orthotopic glioma models. Thus, we will embark on a series of preclinical studies evaluating the combination of Gleevec with specific VEGF inhibitors (LY317615, Avastin, etc.). The poor penetration of Gleevec through an intact blood-brain barrier, however, will also force us to screen other tyrosine kinase inhibitors that have activity against c-kit but may have more favorable pharmacokinetics.
