Few or none of the transcription factors encoded in the human genome act in isolation to regulate the expression of even a single human gene. The relevant physiological effectors are not single proteins, but rather multi-protein complexes in which the sequence-specificity of one or more DNA-binding protein is modulated by combinatorial interactions with structurally unrelated transcription factors that may or may not be DNA bound. These multiprotein complexes possess greater specificity than any single transcription factor, emphasizing the essential role of protein.protein interactions in defining the true specificity of gene expression. The long term goal of this program is to understand the structural and energetic basis for protein.protein interactions that regulate transcription in eukaryotes and deregulate transcription in a diseased human cell. This proposal represents the first step toward that goal, by focusing on the molecular mechanisms by which HTLV-I Tax and human HMG I(Y) interact with eukaryotic bZIP transcription factors to modify their intrinsic DNA sequence preferences.
In Aims 1 & 2 we explore the energetic benefit of bZIP.Tax and bZIP.HMG I(Y) interactions. We employ mobility shift analysis, fluorescence spectroscopy, and bZIP peptide models to apportion the effects of Tax and HMG I(Y) on bZIP dimerization and DNA binding. We will also use DNA minicircles to determine whether the increase in bZIP.DNA stability with HMG I(Y) is the result of induced DNA distortion. Taken together, these experiments will reveal whether Tax and HMG I(Y) function by stabilizing bZIP.bZIP interactions, or bZIP.DNA interactions, or both, and will provide fundamental information on the assembly mechanisms of viral and human initiation complexes.
In Aim 3 we propose alanine scanning mutagenesis experiments to identify which bZIP residues are most important for Tax and HMG I(Y) recognition.
In Aim 4 we propose NMR experiments to explore the structural basis for the enhancement in DNA binding provided to bZIP proteins by HMG I(Y) and Tax. Identification of those bZIP residues that contact Tax and HMG I(Y) will place the thermodynamic analyses described in Aim 3 in a defined structural context and provide important information for the design of transcriptional inhibitors. More generally, this information will deepen our understanding of protein.protein interactions and, we hope, provide the very first glimpse of a transcriptional accessory factor in action.
