HinZip, a designed frankenprotein that combines Hin recombinase and FosW, mimics the structure and DNA-binding function of the HD-Zip plant transcription factor family

Small bespoke proteins that bind a desired DNA sequence in a cell's genome could be a powerful tool in various applications. Examples include using genetically encoded tools to control gene circuits in synthetic biology, or serve as a protein drug that inhibits a disease network impacting human health. We designed HinZip to bind a specific target comprising at least 24 base pairs with high affinity and DNA sequence specificity, because a larger DNA sequence is likely to be unique in a genome, thereby minimizing off-target effects. We took inspiration from the HD-Zip, a transcription factor family only found in plants. No high-resolution structures exist for HD-Zip: genome-wide analyses indicate that they use a homeodomain to bind DNA and leucine zipper for dimerization. HinZip is a fusion of the Hin recombinase DNA-binding domain and FosW leucine zipper. HinZip binds cooperatively as a dimer to DNA targets comprising two 12 base-pair hixC half-sites. Using bacterial one-hybrid and quantitative electrophoretic mobility shift assays, we tested spacings of 0-9 base pairs between half-sites to optimize the orientation and space parameters that FosW needs for coiled-coil dimerization. HinZip binds cooperatively to a 29 base-pair inverted hixC palindrome with Kd 17 nM and no binding to nonspecific DNA up to 2 µM protein. Hin—which lacks ability to dimerize—was previously shown to bind hixC with Kd 34 nM, showing similar binding to half- or full-sites and no cooperativity. HinZip/LA, where the Leu residues responsible for dimerization were replaced with Ala, showed virtually no benefit from cooperative binding at any full-site, and bound first as a monomer, and then as two monomers. Circular dichroism and dynamic light scattering indicate that only HinZip is capable of forming a coiled-coil dimer. HinZip demonstrates that even in the absence of guidance from structural information, small frankenproteins designed from cut-and-paste of unrelated protein modules can specifically target long DNA sequences with broad applications toward orthogonally controlling gene networks in a wide variety of organisms.