Tumor suppressor genes: a new era for molecular genetic studies of cancer.

Previous emphasis in cancer research has been placed on genes in which activating mutations are found in experimental systems and sometimes in human tumors, and many of these genes are the cellular homologs of retroviral oncogenes. Studies of genes whose functions are necessary for maintenance of the normal cellular state, but for which loss-of-function mutations lead to tumor development, are limited. The latter genes have been variously termed 'tumor suppressor genes', 'recessive oncogenes', and 'anti-oncogenes', and each term defines a specific aspect of their properties and may not always be applicable. The retinoblastoma (RB) gene is the first such gene to be identified, and was isolated based on its chromosome localization and on the recessive nature of the tumor phenotype. That is, both wild type RB alleles must be inactivated in a single cell for neoplastic transformation to occur, and deletions at the chromosomal locus now known to contain RB are often found in retinoblastoma cells. Candidate genes for Wilms' tumor and neurofibromatosis type I have also been identified recently, and loss of function of these genes seems to be indicated for these diseases. Allelic loss of chromosome 17p13 is frequently observed in many tumor types. The p53 gene was mapped to this chromosomal region and has been shown to be a tumor suppressor gene, and germ-line mutations of p53 recently were found to be correlated with Li-Fraumeni syndrome, a syndrome characterized by multiple neoplasms. Rapid progress in studies of tumor suppressor genes points to diverse mechanisms for their functioning in the negative regulation of cell growth. A scenario depicting cell growth control by positive and negative regulators, based on new and emerging findings, is the main focus of this review.