Individuals a ected with neuro®bromatosis 1 (NF1) harbor increased numbers of GFAP-immunoreactive cerebral astrocytes and develop astrocytomas that can lead to blindness and death. Mice heterozygous for a targeted Nf1 mutation (Nf1+/7) were employed as a model for the human disease to evaluate the hypothesis that reduced NF1 protein (neuro®bromin) expression may confer a growth advantage for astrocytes, such that inactivation of only one NF1 allele is su cient for abnormal astrocyte proliferation. Here, we report that Nf1+/7 mice have increased numbers of cerebral astrocytes and increased astrocyte proliferation compared to wild-type littermates. Intriguingly, primary Nf1+/7 astrocyte cultures failed to demonstrate a cell-autonomous growth advantage unless they were cocultured with C17 neuronal cells. This C17 neuronal cellinduced Nf1+/7 increase in proliferation was blocked by MEK inhibition (PD98059), suggesting a p21-rasdependent e ect. Furthermore, mice heterozygous for a targeted mutation in another GAP molecule, p120-GAP, demonstrated no increases in cerebral astrocyte number. These ®ndings suggest that reduced NF1 expression results in a cell context-dependent increase in astrocyte proliferation that may be su cient for the development of astrocytic growth abnormalities in patients with NF1.

Neurofibromatosis 1 (NF1) is a common autosomal dominant cancer predisposition syndrome, in which 15% to 20% of affected individuals develop astrocytomas. Neurofibromin, the protein product of the NF1 gene, functions as a tumor suppressor, largely by inhibiting Ras activity. While loss of neurofibromin has been implicated in the molecular pathogenesis of other NF1-associated tumors, there is no formal evidence demonstrating loss of neurofibromin function in NF1-associated astrocytomas. In this report, we describe an NF1 patient from whom both astrocytoma tumor tissue as well as corresponding non-neoplastic white matter were available for analysis. Loss of neurofibromin expression was observed in the tumor and was associated with elevated levels of Ras-GTP. However, elevated Ras-GTP levels were not the result of oncogenic Ras mutations, altered p120-GAP function, growth factor receptor activation, or abnormal p53, Rb, or p16 expression. Furthermore, increased Raf-MAPK and PI3-K/Akt activity was detected in the NF1 astrocytoma compared with the corresponding normal white matter. These results support a role for neurofibromin as the critical GAP in the molecular pathogenesis of NF1 astrocytomas.

No abstract