The coiled coil is a widespread and biologically important structural motif that mediates specific protein-protein interactions. Although coiled coils had previously been assumed to be parallel, recent structural studies have highlighted the importance of antiparallel-coiled coils in nature. Due to the prevalence of coiled-coil domains and to the recent explosion in genome sequence information, the ability to predict coiled-coil function from sequence data would be extremely valuable. Because the orientation within a coiled coil affects strand-pairing specificity, an understanding of the interactions that affect helix orientation is necessary to accurately predict interaction partners. We have recently developed a genetic method for selecting coiled coils that associate with a given relative helix orientation from a randomized pool of proteins. We will apply this approach to identify sequence features that specify one orientation over the other. In addition, because more than half of the structurally characterized coiled coil domains are intramolecular, we will use a combination of rational design and selection methods to probe the structural determinants for helical hairpin formation. Such information will aid in the prediction of coiled coil function, as the helical hairpin appears to be a versatile motif for protein-protein or protein-nucleic acid interactions. We will also probe the interactions of three proteins involved in the maintenance of muscle integrity that appear to contain helical hairpin domains: dystrophin, dystrobrevin, and dysbindin. Finally, one of the most surprising recent findings in this area is that structural maintenance of chromosome (SMC) proteins contains two antiparallel-coiled coils that are several hundred amino acid residues in length. We propose a combination of biophysical and functional studies that explore the importance of coiled-coil length and sequence for function of bacterial SMC proteins in vivo.
