Computational Design of Inhibotor Specificity

The development of drug therapies has been an essential approach to the treatment of infectious disease and cancer. High rates of error-prone replication for certain infectious agents and cancer cells can lead to drug resistance on a relatively short time scale. This project will pursue a new approach to the development of enzyme inhibitors that are less prone to the emergence of target resistance, namely the substrate envelope hypothesis, will develop and apply inverse computational design methods for small-molecule ligands, and will test the substrate envelope hypothesis extensively in the context of HIV-1 protease through a collaborative effort with experimental groups expert in organic and medicinal chemistry, enzyme assays, protein crystallography, and virology. HIV protease has been selected as a case study due to the large amount of prior information regarding resistance mutations that have been selected in patient populations and cell culture under the selective pressure of clinical compounds.

The substrate envelope hypothesis maintains that inhibitors that reside within the volume shared by substrates are less susceptible to resistance mutations, because such mutants must still turn over substrates. Through our computational approaches we will develop inhibitors for HIV protease with robust binding properties to panels of resistance mutants, and through collaborative work these inhibitors will be synthesized, assayed, and characterized. Preliminary work has demonstrated some success and raised some questions regarding the substrate envelope hypothesis. The proposed project involves the further development of our computational ligand design methodology to achieve the goals of the current work, but the developments are of broad applicability, including the efficient treatment of target site flexibility, the incorporation of additional implicit solvent energy function forms, and the implementation of more efficient search algorithms. The proposed project will stringently test the substrate envelope hypothesis through the fine- scale design of inhibitors that do and do not respect the substrate envelope, including tests to rescue inhibitors that succumb to resistance mutations and that violate the substrate envelope, through the design of variants that respect the envelope.

This collection of otherwise identical inhibitors that differ in their adherence to the substrate envelope will be a crucial resource for relating resistance profiles to the envelope hypothesis, which I will study with my experimental collaborators. The proposed project will also design and study the properties of inhibitors predicted to bind broadly to multiple members of a panel of resistance mutants, and compare them to properties of inhibitors predicted to bind narrowly to a single target. In this way, new principles for robust binders and improvements to the substrate envelope hypothesis are likely to result. A particular advantage to implicit design approaches like the substrate envelope hypothesis is that they do not require prior explicit knowledge of drug resistance mutations.

PUBLIC HEALTH RELEVANCE: Current medical drug therapy for infectious disease and cancer is limited by the emergence of resistance, in which a previously effective therapy loses its effectiveness, often through mutations in the target. This project aims to study methods for developing new therapies that prevent, or at least significantly delay, the emergence of resistance. Initial work will target the HIV protease, which is the target of some current therapies, but for which the emergence of resistant strains remains a significant problem.