Four investigators, each of whom brings special expertise to the Consortium, comprise the core of this proposal. The PI, Dr. Terry Van Dyke, has utilized mouse manipulation strategies to study cancer for 16 years. In the course of analyzing the tumor suppressors p53 and pRb [See Fig.1. Inactivation of pRb Tumor,  Fig.2. Multiple Tumor Models], her lab has established several tumor models using transgenic and knock-out strategies [Mol Cell Biol 19:3095-3102 (1999); Genes Dev 13:1246-1250 (1999); Mol Cell 2:283-292 (1998); Mol Cell Biol 18:3495-3501 (1998); Genes Dev 10:826-835 (1996); Cell 78:703-711 (1994); Mol Cell Biol 13:3255-3265 (1993)].

Dr. Eric Holland (Memorial Sloan Kettering Cancer Center) is a practicing neurosurgeon. While a research associate in the laboratory of Dr. Harold Varmus, he developed a novel approach for short-term somatic analysis of cancer genes in the mouse [Oncogene 18:5253-5260 (1999)]. The system uses RCAS viral vectors for glia-specific gene transfer in vivo that permits investigation of effects of mutations, individually and combined, on gliomagenesis in mice. [See Fig.3. Gene Transfer with the RCAS/tv-a System]

Using the RCAS/tv-a system for cell-type-specific gene transfer, the signaling pathways leading to glioma formation have been identified. PDGF gene transfer with an RCAS vector, to both astrocytes and glial precursors, results in the formation of gliomas in these mice. The polyoma virus middle T antigen, which activates similar signaling pathways and PDGF, is capable of inducing both oligodendrogliomas and astrocytomas from astrocytes. In addition, the combination of Ras and Akt induces glioblastomas when transferred to glial progenitors. This glioblastoma formation is dependant on the combination of Ras and Akt signaling and requires an undifferentiated cell of origin since the same combination does not induce gliomas from astrocytes [Nat Rev Genet 2:120-129 (2001); Toxicol Pathol 28:171-177 (2000); Nat Genet 25:55-57 (2000); Proc Natl Acad Sci USA 97:6242-6244 (2000); Am J Pathol 157:1031-1037 (2000); Proc Nat Acad Sci (USA) 95:1218-1223 (1998); Genes Dev 12:3144-3149 (1998); Genes Dev 12:3675-3685 (1998)].

Dr. David Louis (Massachusetts General Hospital) is a board-certified neuropathologist interested in the genetic lesions of human glioma. [See Fig.4. Human Glioma Lesions] [J Neuropathol Exp Neurol 57:122-130 (1998); Oncogene 16:2259-2264 (1998); J Neuropathol Exp Neurol 56:1098-1104 (1997); Am J Pathol 151:1649-1654 (1997)].
Most recently he mapped a new glioma-specific tumor suppressor gene on human chromosome 19. [See Fig.5. Chromosome 19] [Neurogenetics 1:31-36, (1997)].

Dr. R. Jude Samulski (U-NC - Chapel Hill), a leader in gene therapy, has developed viral vectors that show great potential for somatic gene delivery in vivo [Nat Med 5:587-588 (1999); Nat Biotechnol 17:181-186 (1999); Gene Ther 5:1604-1611 (1998)].

The major experimental goal of this proposal is to develop mouse models that accurately reflect stages and classes of human glioma. Many strategies developed and utilized here can also be applied more broadly to other CNS cancer models. Dr. Van Dyke's laboratory has already developed a choroid plexus tumor model [See Fig.6. Tumorigenesis Mechanisms in a Brain Tumor Model, See Fig.7. Predictable Tumor Progression ] [Oncogene 7:1167-1175 (1992); Mol Cell Biol 11:5968-5976 (1991)] that is histologically similar to the human disease.

Project 1 describes the choroid plexus model and the approaches used for its derivation. In Project 2, Dr. Holland will define the cellular responses to genetic changes observed in human glioma by somatic gene introduction into the mouse brain. In Project 3, the Van Dyke laboratory will utilize germline modification strategies to generate mouse strains with similar alterations. Close interaction between the Van Dyke and Holland laboratories will facilitate rapid identification of genetic changes to be engineered into the mouse germline. Preliminary results from these labs indicate significant progress toward producing glioma-like pathologies in the mouse. Project 4 will establish a glial-specific targeted inactivation of the pTEN gene, a common inactivation in human glioma. This mouse will be used in model derivation of Projects 2 and 3. In Projects 1-4, Dr. Louis will assess the extent to which pathologies induced in the mouse resemble those observed in humans. Histological and genetic studies in human and mouse samples will be pursued. In Project 5, the Louis laboratory will complete the isolation of the glioma-specific tumor suppressor gene, and the Van Dyke laboratory will use gene-targeting strategies to assess its tumor suppressor function. Finally, we will examine a limited number of therapeutic strategies in existing and developing mouse models, including epidermal growth factor receptor-targeted strategies by Dr. Holland and anti-angiogenesis gene therapy by Drs. Samulski and Van Dyke.