Oncogenes of DNA Tumor Viruses1

Experiments carried out over the past 10-12 years have created a field or approach which may properly be termed the molecular basis of cancer. One of its major accomplishments has been the identification and understanding of some of the functions of a group of cancer-causing genes, the oncogenes. The major path to the oncogenes came from the study of cancercausing viruses. The oncogenes have been recognized and stud ied by two separate but related groups of virologists focusing upon either the DNA (1) or RNA (2) tumor viruses (they even have separate meetings now that these fields have grown so large). From their studies it has become clear that the oncogenes of each virus type have very different origins. Indeed it is a curious fact that virtually every group of the DNA viruses studied (papova-papilloma, adenoviruses, herpesviruses, hepadnaviruses, poxviruses) has a representative virus which has been shown to produce, or be closely associated with, tumors in animals. Many DNA viruses have been shown to be tumor viruses under the appropriate and often specialized circum stances of an experiment (1). In contrast, among the thousands of RNA viruses, only the retroviruses contain examples of viral agents that cause tumors. While the viral oncogenes that reside in the genome of the DNA tumor viruses bear little or no homology to cellular genes, the oncogenes of retroviruses are clearly derived from and are closely related to their cellular counterparts, the protooncogenes (3). The evidence demonstra ting that viral (DNA tumor viruses) and cellular (RNA tumor viruses) derived oncogenes participate in the development of a cancer cell is quite convincing and accepted by most critical scientists in the field, but not without some reservations (4, 5). Research, pursuing the isolation and functions of the cellular oncogenes, has proceeded rapidly and been very productive with frequent reviews of this field (6-8). Less visible but equally valuable is the information accumulating about the viral onco genes of DNA tumor viruses. The justifications for focusing on this topic are, in the end, the same reasons why the cellular oncogenes are important in the understanding of cancer: (a) some of the DNA tumor viruses (human papilloma viruses 16, 18, 33; hepatitis B virus; Epstein-Barr virus) are closely and continuously associated with specific human cancers and prob ably contribute to one of several events that cause these cancers (9-11); (b) a study of the mechanisms of action of these viral oncogenes will provide fundamental information about the control of cell growth. There is already good evidence that some viral transforming gene products act upon normal cellular pro tooncogene products and alter their activities or levels (12,13). Other viral oncogenes act to bypass the normal cellular control mechanisms regulated by the protooncogenes and their prod ucts. A little consideration and several key experiments lead to the conclusion that a study of the DNA viral oncogenes will inevitably bring us back to the cellular protooncogenes. The goal of this "Perspective" will be to focus upon the DNA