This proposal concerns the creation of novel enzymes by computational methodologies. Building on successes of designed metal-binding sites, receptor binding for a wide range of ligands, and de novo proteins, enzymatic activity was designed in a protein scaffold that lacked previous catalytic character. Despite achieving strong enough enzymatic activity for biological function, this novel computationally designed enzyme did not attain the catalytic efficiencies observed for natural enzymes. Here we propose to extend the enzyme design method by improving the accuracy and scope of the method's preliminary success and augment this design process with complementary directed evolution techniques. These methods will allow us to develop enzyme designs to test specific mechanistic hypotheses. We propose to explore and improve the accuracy of the enzyme design for enhanced catalytic efficiencies in the previously designed TIM enzymes, examine stereoelectronic control within the model by computationally designing the divergently evolved enzyme methylglyoxal synthase, and extend the scope of reactions approached by this computational method by designing a class I (Schiff base) aldolase enzyme.
