Computational Redesign of Protein-Protein Interactions

Computational protein modeling predicts and manipulates the biophysical properties of proteins based on their amino acid sequences. Computational protein modeling has been applied to protein-protein and protein-peptide interactions in order to develop research tools as well as protein therapeutics. The intent of our work was to address three areas of protein interface design that are of special interest due to their potential applications. Affinity maturation has been used to improve biosensors as well as potential protein therapeutics. We developed a protocol to predict point mutations that will enhance the binding affinity of protein-protein interactions. Extending this work, we evaluated a protocol designed to increase protein-peptide binding specificity. Redesigning binding specificity can be used to isolate specific protein interactions within complicated signaling networks by limiting the interactions of redesigned proteins with their wild type counterparts and other natural binding partners. Finally, the de novo design of a peptide-protein interaction, or the design of a peptide that will bind a wild type protein, could enable the creation of biosensors or therapeutics from scratch. We take a step towards this goal by redesigning a portion of a peptide backbone in the context of its wild type binding partner