The primary goal of this project is to predict the biological properties of 800 to 1000 hypothetical proteins that have been identified by homology sequence analysis as probable members of the short chain oxidoreductase (SCOR) superfamily, and to test those predictions by X-ray crystal structure determination and biochemical analysis. The importance of conserved residues for proper enzyme folding, allosteric behavior, oligomerization, and specificity of cofactor, substrate, and inhibitor binding, as well as preference for catalyzing oxidation, reduction or epimerization reactions, will be determined. A second major goal is the design of selective inhibitors for the cholesterol esterase enzyme family, as well as the steroid-metabolizing subfamily of the STOR superfamily. This is a logical extension of an ongoing effort to achieve a molecular level understanding of the biological and physiological events surrounding steroid function. Since members of the SCOR superfamily are linked to cancer, hypertension, tuberculosis, Alzheimer's disease, infertility, ulcers, hypokalemia, and polycystic kidney disease, they are important targets for drug design. Unfortunately, drugs designed to combat a disease by inhibiting one family member may inadvertently disrupt other related metabolic pathways. By screening a battery of known inhibitors of individual STOR enzymes against representative members of each STOR subfamily, information will be obtained regarding their binding affinities and cross reactions that will be useful for the design of selective inhibitors. Successful achievement of these goals will require bioinformatic analysis of the SCOR proteome sequence in order to select good target proteins for study; cloning and overexpression of proteins; crystallization and X-ray diffraction data collection, 3D crystallographic structure determination and refinement; activity testing and inhibition studies, and the correlation of structural and functional data. Proof of the successful prediction of the structures of 20-30 hypothetical members of the SCOR proteome could provide a model for the prediction of protein folding on the basis of as little as 20% sequence homology.
Showing the most recent 10 out of 53 publications
