Control of Hap1-DNA site recognition through the interplay of multiple distinct intermolecular interactions

Hap1 belongs to the Zn(2)Cys(6) zinc binuclear cluster family of transcription factors that typically bind as dimers to symmetric DNA sites containing two CGG triplets separated by spacer DNA. The cluster domain binds CGG while an adjoining C-terminal linker and dimerization helix specifies the length of spacer DNA recognized. Hap1 is unusual in binding a direct repeat of CGG triplets, in contacting a TA in the spacer DNA, and in making direct dimer contacts between its cluster domains. Binding of Hap1 fragments to different DNA sites was tested to determine how these interactions control Hap1-DNA recognition. The spacer TA contacts were found to facilitate monomer binding of Hap1 to a single CGG. When the spacer-binding residues were deleted, binding was still specific for the direct repeat but was much weaker and appeared to require dimerization. When the dimerization helix and all subsequent C-terminal residues were deleted, the remaining linker, cluster domain, and spacer-binding residues still dimerized on DNA. The energy of this dimerization was comparable to that of the Hap1-spacer TA interaction. Moving the TA from the spacer to a position following the second CGG maintained Hap1 monomer binding but greatly weakened dimerization. This suggested that binding a TA after the second CGG triplet required a geometry that impaired dimerization with a Hap1 molecule on the first CGG. The geometric restraints for optimal TA binding and dimerization thus drive Hap1 selectivity for CGG direct repeat sites that contain an asymmetrically positioned spacer TA following the first CGG triplet.