This is a synthetic, structure-based drug design proposal directed toward finding novel peptidemimetics that inhibit aspartic and other proteases. Peptidemimetics are low molecular weight non-peptide molecules that bind to the target enzyme or receptor in a fashion close to that of the natural peptide ligand. Since many putative peptidemimetics are structurally related to classical, heterocyclic drugs that are well absorbed, true peptidemimetics are expected to possess desirable pharmacodynamic properties superior to the natural peptides due to their improved oral activity and duration of action. The potential benefits to clinical medicine are enormous if this goal can be achieved. Several non-peptide inhibitors of pepsin, penicillopepsin and other aspartic proteases will be designed, synthesized, and evaluated as part of a long term, collaborative effort to find ways to obtain non-peptide (or minimally peptide) inhibitors of proteases. Compounds will be designed in collaboration with Dr. Regine Bohacek, CIBA, by use of GrowMol, Dr. Bohacek's computerized structure-generating program that generates novel structures in the active site of a target enzyme atom-by- atom. Selected analogs will be synthesized and assayed in my laboratory for potency against the target enzyme. The X-ray crystal structure will be obtained for pepsin-inhibitor complexes and penicillopepsin-inhibitor complexes when suitable inhibitors are identified; this work will be done in collaboration with Dr. Hazel Holden, Associate Professor at UW- Madison. We have chosen the aspartic protease pepsin to carry out the initial studies because of the large body of structure-activity data that will help guide our selection of early synthetic targets, because of my extensive experience synthesizing inhibitors of this system, and because of the importance of several aspartic proteases, including HIV protease, renin and cathepsin D, as potential therapeutic targets. Critical to the successful evaluation of structure based drug design using GrowMol is our ability to determine the actual mode of binding by X-ray crystallography so as to evaluate how well the computer-generated predictions work.
Showing the most recent 10 out of 11 publications
