Cancer genes come of age.

Abstract

In 1866, Paul Broca sketched the pedigree of his wife’s family and discerned a hereditary diathesis to cancer. The insight seems to have attracted little attention in Broca’s time (indeed, this child of Broca’s brain went unmentioned by Carl Sagan in our time). During the century that followed, however, biologists began to seek genetic explanations for tumorigenesis. Now the quest has reached fruition: the long-imagined “cancer genes” have been brought into view. They were unearthed first by the predilection of retroviruses to transduce cellular genes-it was the search for retroviral oncogenes in the DNA of normal cells that incarnated cancer genes (Bishop, Ann. Rev. Biothem., in press). Two further strategies have since added fuel to the fire: integration of retroviral DNA can finger cellular genes whose activation apparently initiates tumorigenesis (Varmus, Cancer Surv. 7, 30931 9, 1982); and DNA taken from a variety of tumors can convert rodent cells to neoplastic growth, as if the DNA contained an active oncogene-of the sort we were once accustomed to finding only in viruses (Weinberg, Adv. Cancer Res. 36, 149-163, 1982; Cooper, Science 2 7 7, 801-806, 1982). Oncogenes of retroviruses have shown us that dominant genetic loci can convert cells to neoplastic growth. The experimental simplicity offered by viral oncogenes may be misleading, since natural tumorigenesis is thought to arise from several discrete events within the emerging tumor cell. It is nevertheless from tumor viruses that we have gained our most promising glimpse of biochemical mechanisms that may lead to neoplastic growth. Eighteen different oncogenes have now been found in retroviruses, causing different tumors by a variety of mechanisms. The proteins encoded by these genes display a provocative diversity (Bishop, op. cit.): some are tyrosinespecific protein kinases located on the plasma membrane; some are nuclear proteins that may bind to DNA within chromatin; some are glycoproteins that linger inexplicably on intracellular membranes throughout their life span; some possess structural features and biochemical properties evocative of electrogenic nucleoside triphosphatases (Gay and Walker, Nature 307, 262-264, 1983); and some remain wholly uncharacterized, offering promise of even further diversity. What does this diversity connote? That there is more than one way to create a cancer cell, of course. But a greater truth may lie beyond. It seems likely that the growth and division of cells are regulated by an interdigitating network spanning from the surface of the cell to the heart of the nucleus. If that Minireviews