Cancer development in humans and animals is a multistep process involving at least two classes of genes, proto-oncogenes and tumor suppressor genes. We have shown that neoplastic transformation of Syrian hamster embryo cells (SHE) in culture is a multistep process involving both activation of proto-oncogenes and inactivation of a tumor suppressor gene. The loss or inactivation of tumor suppressor genes is an essential step in the multistep neoplastic transformation of SHE cells. Non-tumorigenic variants have been isolated that have lost (sup-) or retained (sup+) the ability to suppress tumorigenicity of tumor cells in cell hybrids. Fusions of sup+ or sup- variants with different tumor cells show different patterns of suppression indicating that a family of tumor suppressor genes exists in these fibroblast cells. Currently, several strategies to clone tumor suppressor genes are in progress. cDNA libraries of sup+ hamster cells have been screened with RNA from sup+ or sup- cells and differentially expressed cDNAs have been cloned. Two-dimensional gel analyses of proteins showed that a reduction in the expression of tropomyosin I correlates with the loss of the tumor suppressor function. A cellular phenotype associated with the loss of tumor suppressor gene function has also been found. Sup- cells suspended in agar respond reversibly to transforming and normal growth factors by forming colonies in agar whereas sup+ cells fail to grow., Tumor suppressor genes can be mapped to specific chromosomes by introduction of normal chromosomes into tumor cells by microcell fusion. We have shown that normal human chromosome 11 suppresses cervical carcinoma cells, lung adenocarcinoma cells, rhabdomyosarcoma cells, and Wilms' tumor cells, whereas chromosome 3 suppresses renal carcinoma and lung adenocarcinoma cells. An uterine endometrial cancer cell was suppressed by three different chromosomes (Nos. 1, 6, and 9). In addition to the tumor suppressor genes described above that are expressed in some immortal cell lines, tumorigenicity also may be limited by cellular senescence. Our results indicate that a gene(s), possibly involved in the senescence phenotype, can be mapped to human chromosome 1.
