Molecular dynamics guided design of coil-based therapeutics

The coiled-coil is a common protein tertiary structural motif that is composed of two or more alpha helices intertwined together to formed a supercoil. In biological systems, the coiledcoil motif often forms the oligomerization domain of various proteins including DNA binding proteins, structural and transport proteins, and cellular transport and fusion proteins. It was first described by Crick in the 1950s while describing the structure of α-keratin and has since that time been the subject of numerous engineering and mutation studies. This versatile motif has been adapted to a number of nonbiological applications including environmentally responsive hydrogels, crosslinking agents, the construction of self-assembling fibers for tissue engineering, and biosensor surfaces. In this dissertation, we test the applicability of computational methods to understand the underlying energetics in coiled-coils as we apply molecular modeling approaches in the development of pharmaceutics. Two studies are described which test the limits of modern molecular dynamic force fields to understand the structural dynamics of the motif and to use energy calculation methodologies to predict favorable mutations for heterodimer formation and specificity. The first study considers the increasingly common use of fluorinated residues in protein pharmaceutics with regard to their incorporation in coiled-coils. Many studies find that fluorinated residues in the hydrophobic core increase protein stability against chemical and thermal denaturants. Often their incorporation fails to consider structural, energetic, and geometrical differences between these fluorinated residues and their nonfluorinated counterparts. To consider these differences, several variants of Hodges' very stable parallel heterodimer coiledcoil were constructed to examine the effect of salt bridge lengths and geometries with mixed fluorinated and nonfluorinated packed hydrophobic cores. In the second study, we collaborated with an experimental laboratory in the development of a mutant Bcr monomer with designed mutations to increase specificity and binding to the oncoprotein Bcr-Abl for use as an apoptosis inducing agent in chronic myelogenous leukemia (CML) cells. The final chapters of this dissertation discuss challenges and limitations that were encountered using force fields and energetic methods in our attempts to use computational chemistry to model this protein motif.