| Understanding genomic regulation is central to determining mechanisms of both normal cellular function as well as disease onset and progression. The 'recruitment model' proposes that the transcription factors activate genes by tethering the transcription machinery to target genes. Specifically, transcription factors recognize target gene promoters using DNA binding domains and interact with the transcription machinery through their activation domains. In these studies, we have focused on the 'acidic activators' whose activation domains are biased towards the presence of acidic and bulky hydrophobic residues. Examples of acidic activators such as the mammalian p53, the viral VP16 and the yeast Gcn4 appear to function by a conserved mechanism and are functional in both yeast and mammalian systems. In addition, these activators share the common features that they contain tandem activation domains that are able to both work in isolation and bind diverse target proteins. Using Gcn4 as a model activator, I sought to understand the basis for both the redundancy between the two activation domains and the logic behind having numerous binding targets present in the transcription machinery. First, photo-crosslinking was used to identify targets of the Gcn4 N-terminal activation domain as Gal11 (Mediator), Tra1 (SAGA/NuA4), Taf12 (SAGA/TFIID) and Sin4 (Mediator). Next, a hydroxyl-radical cleavage assay in combination with in vitro binding studies led to the identification of two regions in Gal11 and one region of Taf12 that directly interacted with Gcn4. Binding measurements were then conducted to determine both the specificity and affinity of binding. In vivo studies then demonstrated that the Gcn4-Gal11 interaction was uniquely required at only a subset of Gcn4 target genes. Finally, a large amount of effort was put into structural studies, which through collaboration with NMR spectroscopists led to the determination of a structure between the Gcn4 central activation domain and a region within Gal11. Overall, these studies demonstrate the basis for activation domain redundancy, reveal a logic for the presence of multiple activation domain binding partners and provide atomic level detail on the mechanism of function of the acidic activation domain. |
