Neurofibromatosis (NF) is an autosomal dominant disorder affecting 1 in 2500 individuals in the U.S. While its manifestations can be highly variable, most affected individuals will develop multiple neurofibromas (benign tumors arising from nerve sheaths) by adulthood, and frank malignancy occurs at an increased rate, either as a result of malignant degeneration of preexisting plexiform neurofibromas or as a primary event. This condition thus belongs in the category of familial cancer syndromes. Recent rapid progress in the understanding of the molecular basis of cancer has occurred with the identification, cloning, and sequencing of cellular oncogenes. The possibility that NF represents an activating mutation in such an oncogene, which is in itself insufficient to induce malignant transformation but facilitates the development of tumor formation when other events occur, leads to the experiments proposed here. We propose to use DNA-mediated gene transfer to look for activated oncogene activity in NF. It is already known that NF DNA will not transform NIH 3T3 cells, but this system detects only a subset of known oncogenes (those of the ras category). We will transfect primary rat embryo fibroblasts with NF DNA plus an activated ras gene to look for activated oncogenes of other classes and will also use NF cells as the recipients in DNA transfection experiments. For this purpose, we have recently developed efficient methods of transfecting human skin fibroblasts. In addition, we will look for oncogene rearrangements in different families with NF by Southern blotting and chromosome analysis, and for abnormalities of oncogene expression by Northern blotting of fibroblast mRNA. A broad panel of human oncogene probes will be used for these blotting experiments. An alternative mechanism for NF will also be investigated in which the NF mutation occurs in a gene that normally acts to inhibit growth; so long as the other allele of this gene is active, no phenotypic effects occur. If the other allele is inactivated by somatic mutation or chromosome loss, however, abnormal growth would ensue. This model has already been shown to apply in retinoblastoma. We propose to test this model and attempt to map the NF locus by careful cytogenetic analysis of neurofibromas and neurofibrosarcomas from patients with NF, as well as by insertional mutagenesis of NF fibroblasts using a highly efficient retrovirus vector.
