I propose to computationally design for the first time an enzyme catalyzing a reaction that is not catalyzed by a naturally occurring enzyme. This project will serve as a proof of principle for the computational design of new catalysts, which has been speculated about for some time but has only recently been technically possible. The first step is to create ideal active sites consisting of a transition state model for the reaction, using quantum chemical methods, to be catalyzed and side chain functional groups placed in positions optimal for catalysis. The second step will be adapting the protein design and protein docking algorithms previously developed in the Baker lab to design a protein with high affinity for the transition state model and with the key residues identified in the first step in the correct orientation. I will then produce the computationally designed enzymes in the laboratory by synthesizing genes encoding the proteins and expressing them, and subsequently characterizing their catalytic efficiency, selectivity and three dimensional structure. Based on feedback from the calculations and experimental results, I will optimize the design protocol and use it to generate enzymes for additional reactions of importance to industry and medicine.
