Studies on two transcription regulatory proteins : cellular c-Jun and adenovirus E1A

Abstract

Adenoviruses cause lytic infections in a range of mammalian cells, but can in certain cases also transform the host cell. The first adenovirus gene to be expressed upon infection is E1A. The E1A gene encodes two major proteins of 243 and 289 amino acids, respectively, both of which serve as key regulators during virus infection. The E1A proteins can redirect transcription and also reprogram the cell cycle to suit the need of the virus. In these processes different regulatory proteins from the host cell are bound by the E1A proteins via its conserved regions (CR1, CR2 and CR3). Studies of E1A function have thus proven very useful for learning about regulation of transcription and cell cycle control in the cell. The E1A proteins can both activate and repress transcription of specific cellular, and viral genes The repressing function of E1A has previously been shown only for RNA polymerase II transcribed genes. Using transient transfection experiments, we here demonstrate that E1A can repress transcription also of the polymerase III-driven VA RNA genes E1A deletion studies showed that CRI and the first five amino acids of CR2 were required for this repression.

We also discovered that another adenovirus encoded protein, EIB 19K, could counteract the E1A-mediated repression by activating transcription of the VA RNA genes The importance of E1B 19K for VA RNA expression was demonstrated also during a viral infection In a parallel project, we identified a previously undescribed transcription regulatory region located near the C-terminus of the E1A proteins. This region exactly corresponds to the binding site for the cellular protein CtBP. Binding of CtBP to E1A has previously been suggested to suppress the ability of E1A to transform host cells. We show that transcription mediated by the E1 A-CR I domain is regulated by the CtBP binding region. Moreover, we show that the CtBP-binding region is needed for efficient induction of some, but not all, E1A243R-responsive cellular genes. We therefore propose a novel mechanistic model for E1A243R-mediated transactivation.

Studies on transcription activation domains in mammalian cells are often done by fusing them to a heterologous DNA binding domain (DBD). Such fusion proteins expressing the DNA binding domain of the yeast transcription factor Gal4 can activate transcription from synthetic promoters with binding sites for the Gal4-DBD. This Gal4-fusion system is widely used since mammalian genes do not contain binding sites for Gal4. As a spin-off result from our studies with the CR1 activation domain, we however found that Gal4-DBD can bind to the cellular transcription factor c-Jun. This resulted in efficient transactivation in vivo of promoters containing binding sites for the c-Jun containing APl factor. This finding has important consequences for everyone working with Gal4 fusions in mammalian cells, since transactivation is not restricted to promoters containing binding sites for Gal4, as was previously thought.

In summary, the work presented in this thesis reveals new insight into the mechanisms used by the E1A proteins to modulate transcription. It also describes a previously unknown interaction between the yeast Gal4 and mammalian c-Jun transcription factors and the consequences of this interaction for the study of transcription initiation is discussed.