The adenovirus E1A oncogene encodes two major proteins of 289 and 243 aa involved in cell immortalization, transformation in cooperation with ras, and transcriptional activation and repression on both viral and cellular genes. The E1A protein binds specifically to different gene products, including the tumor suppressor gene product of the retinoblastoma, a related 107 kDa protein, a 300 kDa polypeptide, and a 60 kDa cyclin a subunit of the cdc2 type kinase. Since E1A is by itself unable to bind DNA and no consensus activator or repressor site has been detected, we investigated interactions between E1A and cellular proteins required for transcription. We have thus described a novel interaction between adenovirus E1A and the cellular TATA-box binding protein TBP. A region close to the basic domain of TBP and the Zn finger domain are involved in this interaction. Moreover, we have established that ATF2, a cellular DNA binding protein also is able to interact with TBP, probably via a third protein. To establish the functional relevance of these interactions, we propose to perform in vitro and in vivo assays using E1A and TBP mutants, as well as ATF2 recombinants on several viral promoters. In vitro assays will focus on the rate, efficiency and stability of transcription complex formation and transcription stimulation. In vivo assays will either alter the levels of these basal components or assay transdominance effects (i.e., E1A or TBP mutants which interfere with normal transactivation) in promoters that have been shown to respond to E1A via the TATA-box sequence. We will use protein-protein screening of expression libraries to look for additional sequences encoding proteins which may be, at least in part, responsible for the pleiotropic effects of the E1A oncogene. The role of these proteins in normal cellular processes in the absence of E1A will also be determined. These experiments will clarify the mechanism by which E1A stimulates or represses transcription, an important part of its oncogenic potential.
