Cellular phenomena such as replication, recombination, differentiation, and cell growth are regulated at athe most fundamental level by transcription factors, proteins that bind DNA and regulate gene expression. The direct relationship between aberrant gene expression and human disease emphasized the importance of understanding, at the molecular level, the mechanisms by which transcription factors discriminate between DNA sites. This proposal builds on discoveries made in our laboratory over the past year to analyze in detail the mechanisms by which eukaryotic bZIP (basic segment leucine zipper) transcription factors discriminate between the CRE and AP-1 sites, sequences that differ by the presence or absence of a single base pair. Despite their sequence similarity, the CRE and AP-1 sites are athe nuclear end-points of two different signal transduction pathways. We discovered that these DNA sites are equated structurally by an intrinsic bend in the CRE site, and that the CRE-selective bZIP protein CRE-BP1 overcomes this intrinsic bend and straightens the DNA using residues in its basic segment. We propose (Specific Aim 1) to identify which residues in the basic segment are required for distortion of the CRE site, and thereby explore the relationship between induced distortion and specificity. The observation that CRE-BP1 contains basic segment residues that stabilize a distorted form of the CRE site explains why these proteins bind the CRE site, but it does not explain why they prefer it. We will address this issue (Specific Aim 2) by engineering DNA minicircles to contain a CRE or P-1 site pre-bent towards the minor groove, into a conformation suitable for CRE-BP1. By comparing the affinity of CRE-BP1 for minicircles containing pre-bent DNA with the corresponding linear DNA fragments, we will learn whether specificity results from differential bending energies or differential DNA contact energies, or both. In addition, we propose to assess the generality of our """"""""induced-straightening"""""""" model for the half-site spacing specificities of CREB/ATF proteins (Specific Aim 3) by examining other members of the family. We also propose a kinetic analysis of CRE/AP-1 discrimination by CRE-BP1 (Specific Aim 4). Finally, we propose an in vitro selection experiment (Specific Aim 5) to identify other DNA sequences that contain intrinsic major groove bends. Our long terms goals are to understand the thermodynamic basis for the specific protein.DNA and protein.protein interactions that orchestrate the precise control of gene expression. In a more specific sense, the relevance of these experiments to human medicine is straightforward: the three 21 base pair enhancer elements within the long terminal repeat of the human T-cell leukemia virus HTLV-1 each contain a CRE-like sequence, and the major T-cell proteins that bind the HTLVI21 bp repeats and mediate transactivation by the viral Tax transactivator are CREB/ATF family members. Therefore, results from the experiments described here will contribute directly to our thinking about the mechanisms of transcriptional activation by Tax.
