Combinatorial Library Design of Mutation- Resistant Hiv Protease Inhibitors

Title of Dissertation: COMBINATORIAL LIBRARY DESIGN OF MUTATIONRESISTANT HIV PROTEASE INHIBITORS. Sripriya Chellappan, Doctor of Philosophy, 2006 Dissertation directed by: Professor Michael K. Gilson Department of Molecular and Cellular Biology The emergence of HIV strains that are resistant to current HIV protease inhibitors in the past few years has become a major concern in AIDS treatment. The goal of this project is to design a combinatorial library of potential lead compounds that can bind to both the wild-type and mutant proteases and that can resist further mutations. A recent crystallographic study of complexes of HIV protease with its substrates has provided structural insights into the differential recognition of the substrates and inhibitors. It has been proposed that clinical resistance is a consequence of inhibitors failure to stay within the consensus substrate volume. In this work, we devised a quantitative indicator of the degree to which a candidate ligand falls outside the consensus substrate volume, and determined its correlation with the inhibitor’s sensitivity to clinically relevant resistant mutations. The validation of this hypothesis has encouraged us to use this strategy in our design of a combinatorial library of inhibitors. The compounds in a typical combinatorial library are built around a common structural scaffold possessing multiple connection points where substituents can be added by reliable synthetic steps. As the number of compounds encompassed by such a combinatorial scheme frequently exceeds what can actually be synthesized and tested, virtual screening methods are sought to shortlist the compounds. Even though these methods require only seconds to minutes of CPU time per compound, exhaustive screening of an entire virtual combinatorial library is computationally demanding. We therefore implemented a simple algorithm of combining substituents that have been optimized independently for the substituent sites. This method was compared with Genetic Algorithm, a global optimization method and was found equally efficient. This simple method was hence chosen for the design process. A combinatorial library based on these ideas and methods has been synthesized and tested. It includes four compounds with nanomolar inhibition constants. Two of them were shown to have retained affinity against a panel of treatment-resistant mutations. COMBINATORIAL LIBRARY DESIGN OF MUTATIONRESISTANT HIV PROTEASE INHIBITORS