Oncogenes and human cancer: cause or consequence?

During the last five years we have witnessed a series of scientific developments that are making possible for the first time the unveiling of molecular events involved in the onset of human neoplasia. Identification and subsequent characterization of oncogenes in human tumors occupies the most prominent place among these scientific developments. The concept of oncogene emanated from work started almost 20 years ago with acute transforming retroviruses (see 1, 2 for reviews). Genetic and biochemical evidence unquestionably defined the existence of a small region in the viral genome whose expression was sufficient to trigger carcinogenesis. At least 20 retroviral oncogenes have been identified and characterized so far, particularly since the advent of recombinant DNA technology (2). However, many investigators have remained skeptical about the significance of oncogene research in regard to our understanding of human cancer. After all, the concept of oncogene as defined by retroviruses is in serious conflict with the multi-stage nature of human malignancies. The realization that retroviral oncogenes are of cellular origin (3,4) aJerted scientists to the possibility that related dominant oncogenes may also exist in non-virally induced neoplasms. It was reasoned that cellular genes (proto-oncogenes) may become activated by somatic mutations that mimic the changes imposed upon these loci during retroviral transduction. Testing this hypothesis required technology capable of transferring single copy genes from one mammalian cell to another. In 1973, Graham and van der Eb developed a calcium precipitation technique which allowed them to introduce subgenomic fragments of adenovirus DNA into rodent cells (5). Four years later, this technique was successfully utilized to introduce single copy genes into mammalian cells utilizing either metaphase chromosomes or deproteinized genomic DNA (6,7).