1. ¹Microvascular and Molecular Neurooncology Laboratory, New York University Medical Center, New York, New York
2. ²Department of Pathology, Division of Neuropathology, New York University Medical Center, New York, New York
3. ³Division of Surgical Pathology, New York University Medical Center, New York, New York
4. ⁴Division of Pediatric Pathology, New York University Medical Center, New York, New York
5. ⁵Department of Neurosurgery, New York University Medical Center, New York, New York
6. ⁶Department of Pharmacology, New York University Medical Center, New York, New York
7. ⁷Kaplan Cancer Center, New York University Medical Center, New York, New York
8. ⁸Regeneron Pharmaceuticals, Tarrytown, New York

Correspondence: Dr. D. Zagzag, Department of Pathology, Division of Neuropathology, New York University Medical Center, 550 First Avenue, New York, NY 10016. Fax: 212-263-8994; E-mail: zagzag@is3.nyu.edu

Received 11 January 2000.

Abstract

Vascular changes in gliomas were analyzed by implanting fluorescent-labeled glioma 261 cells in the brains of 28 mice. Seven animals were killed each week for 4 weeks. We investigated the expression of angiopoietin-2 (Ang-2) by in situ hybridization and compared it with the distribution of apoptotic cells identified by DNA strand breaks (using the terminal deoxynucleotidyl transferase-mediated biotinylated deoxyuridine triphosphate nick end labeling [TUNEL] method) and transmission electron microscopy (TEM). As early as 1 week after implantation, tumor cells accumulated around vessels, which expressed Ang-2 and were TUNEL negative. TEM showed tumor cells adjacent to the vascular cells "lifting up" the normal astrocytic feet processes away from the endothelial cells and disrupting normal pericytic cuffing. After 2 weeks the number of perivascular glioma cells had increased. No increase in the number of blood vessels was detected at this time. Vascular cells remained positive for Ang-2 and rare ones were TUNEL positive. TEM showed closely packed proliferating perivascular tumor cells. After 3 weeks, there was vascular involution with scant zones of tumor necrosis. Ang-2 was still detected in vascular cells, but now numerous vascular cells were TUNEL positive. In addition, TEM showed apoptotic vascular cells. After 4 weeks, there were extensive areas of tumor necrosis with pseudopalisading and adjacent angiogenesis. Ang-2 was detected in vascular cells at the edge of the tumors in the invaded brain and in vessels surrounded by tumor cells. At both 3 and 4 weeks, most of the TUNEL-positive tumor cells lacked morphological features characteristic of apoptosis and displayed features consistent with necrotic cell death as determined by TEM. Only rare tumor cells appeared truly apoptotic. In contrast, the TUNEL-positive endothelial cells and pericytes were round and shrunken, with condensed nuclear chromatin by TEM, suggesting that vascular cells were undergoing an apoptotic cell death. These results suggest that vascular cell apoptosis and involution preceded tumor necrosis and that angiogenesis is a later event in tumor progression in experimental gliomas. Moreover, Ang-2 is detected prior to the onset of apoptosis in vascular cells and could be linked to vascular involution.