Computational and combinatorial design of protein-based inhibitors of human tyrosyl-DNA phosphodiesterase

Computer programs emulating protein-protein interactions were implemented to design novel protein inhibitors (designer proteins) of human tyrosyl-DNA phosphodiesterase (Tdp1, target protein). In the first of two calculations, the Geometric Recognition Algorithm was used to find orientations of high surface complementarity between designer and target proteins. In the second calculation, protein design algorithms were used to computationally mutate specific positions within the designer protein at the interfacial boundary such that protein-protein interactions were optimized. Molecular biology tools were used to physically generate target and designer proteins and test their specific binding properties. Phage display methods were also used to select improved designer proteins with higher binding affinity. Computationally-generated variants had increased binding affinity to the target compared to the wild type parent designer protein. A phage variant had increased binding affinity to the target compared to the computationally-generated parent.; Designer proteins were computationally docked to the target protein such that their positioning would prevent the target from association with its substrate. Therefore, designer proteins were tested for their inhibitory properties by applying them to a tyrosyl-DNA phosphodiesterase catalytic activity assay. Inhibition was not observed. However, tyrosyl-DNA phosphodiesterase processes the in vitro substrate utilized in a diffusion-limited manner, which renders any findings herein inconclusive.; This project was a test of computer-assisted protein design methods and allowed for a comparison of the relative merits of rational and combinatorial protein design techniques. The methods described in this work are intended as a first round of rational and combinatorial design. Designed protein variants with improved binding properties can be used as input to a second round of computational design. Methods can be refined and repeated in an iterative process, each iteration producing a protein variant with improved binding characteristics.