DESCRIPTION (provided by applicant): This Program Project is designed to address fundamental issues in mutagenesis relevant to the root causes of cancer, in accordance with the mission of NCI. We propose to investigate the molecular basis of DNA polymerase accuracy, relating theory to experiment and vice versa, using human DNA polymerase beta as a model system. Pol beta plays a key role in the avoidance of cancer, because its loss of regulation or disruption by mutation induces chromosome instability and tumorigenesis. Our primary goals are focused on understanding the principles of polymerase fidelity defined by the detailed interactions between specific amino acid side chains, primer/template bases and dNTP substrates at the Pol active site. The Program Project contains three research projects, structural (Project 1), theoretical computational (Project 2), kinetics (Project 3) and three core facilities, a Biochemical Synthetic and Analysis Core (Core B), a Computational Core (Core C) and an Administrative Core (Core A). The goal of Project 1 is to obtain high-resolution structural data for normal and mutant forms of pol ? using a new class of nucleotide analogs designed in Project 3 and synthesized in Core B. These analogs will be used in drug design and delivery strategies to establish their potential use as anticancer agents in mouse and cultured cell model systems, in a translational approach to target bone tumors. A unique and timely aspect of the PPG is the application of theoretical and computer-modeling approaches to structure/function analysis of catalytic efficiencies in polymerase active sites, as proposed in Project 2. The modeling analysis calculates free energies, which are used to predict individual contributions of amino acid side chains to fidelity, including substrate binding and catalysis in the polymerase active site. The theory serves as the intellectual framework with which to marry structural analysis with kinetic mechanistic analyses described in Project 3. It is usually atypical for the experimentalist to test a priori computational predictions. Thus, a defining aspect of this PPG is its bidirectional interplay, where computational predictions are tested experimentally and new experimental data are used to refine the theory.