While the technology for designing and implanting active sites on protein scaffolds has recently been developed, a significant problem in de novo protein design remains: how to best engineer the remainder of the protein to both support the active site and retain a singular three dimensional structure. Our recent work suggests that native-like proteins can be most simply created using simple hydrophobic/hydrophilic, binary patterned, helical protein libraries which include interhelical polar interactions such as ligand binding that impose orientational specificity. We are examining cofactor-binding protein design by first creating libraries of binary patterned proteins in which each helix donates one ligand to divalent zinc, thereby fixing its orientation upon metal binding. Then, to examine the more complicated binding of larger catalytic cofactors, we plan to make similarly patterned heme-binding proteins which incorporate heme binding sites which have been statistically derived from the Protein Data Bank. Both experiments are predicted to result in assemblages of proteins which will be predominantly molten globules in the absence of cofactors, but which will have a significant proportion of native-like proteins in their presence.

The proportion of native-like proteins in the presence and absence of these cofactors will provide the first quantitative description of the probability of generating a uniquely structured core using binary patterning alone. Furthermore, proteins that exhibit this binding-coupled structural change are being used to examine the thermodynamics of these phase transitions. Promising members of these libraries will serve as starting points for the development of new heme-, porphyrin- and nonheme iron-based ‘green’ industrial catalysts.