Imaging glioma initiation in vivo through a polished and reinforced thin-skull cranial window.

Glioma is the one of the most lethal forms of human cancer. The most effective glioma therapy to date-surgery followed by radiation treatment-offers patients only modest benefits, as most patients do not survive more than five years following diagnosis due to glioma relapse (1,2). The discovery of cancer stem cells in human brain tumors holds promise for having an enormous impact on the development of novel therapeutic strategies for glioma (3). Cancer stem cells are defined by their ability both to self-renew and to differentiate, and are thought to be the only cells in a tumor that have the capacity to initiate new tumors (4). Glioma relapse following radiation therapy is thought to arise from resistance of glioma stem cells (GSCs) to therapy (5-10). In vivo, GSCs are shown to reside in a perivascular niche that is important for maintaining their stem cell-like characteristics (11-14). Central to the organization of the GSC niche are vascular endothelial cells (12). Existing evidence suggests that GSCs and their interaction with the vascular endothelial cells are important for tumor development, and identify GSCs and their interaction with endothelial cells as important therapeutic targets for glioma. The presence of GSCs is determined experimentally by their capability to initiate new tumors upon orthotopic transplantation (15). This is typically achieved by injecting a specific number of GBM cells isolated from human tumors into the brains of severely immuno-deficient mice, or of mouse GBM cells into the brains of congenic host mice. Assays for tumor growth are then performed following sufficient time to allow GSCs among the injected GBM cells to give rise to new tumors-typically several weeks or months. Hence, existing assays do not allow examination of the important pathological process of tumor initiation from single GSCs in vivo. Consequently, essential insights into the specific roles of GSCs and their interaction with the vascular endothelial cells in the early stages of tumor initiation are lacking. Such insights are critical for developing novel therapeutic strategies for glioma, and will have great implications for preventing glioma relapse in patients. Here we have adapted the PoRTS cranial window procedure (16)and in vivo two-photon microscopy to allow visualization of tumor initiation from injected GBM cells in the brain of a live mouse. Our technique will pave the way for future efforts to elucidate the key signaling mechanisms between GSCs and vascular endothelial cells during glioma initiation.