ABSTRACT Human cytochrome P450 enzymes are dynamic, often promiscuous, monooxygenases. Some function in the biosynthesis of critical endogenous compounds and are frequent drug targets. Others are dominant factors in drug metabolism, dictating drug clearance and/or prodrug activation. For both, understanding P450 interactions with substrates, inhibitors, and their catalytic partner proteins provides substantial useful information in drug design. While X-ray structures have provided numerous insights into drug binding to key human P450 enzymes, there are many gaps that cannot be filled by this approach. Conformational changes that P450s must undergo to channel ligands to the active site and for a single P450 to accommodate many different small molecule scaffolds only randomly become apparent in comparing the X-ray structures achievable. Despite substantial efforts, some key human P450 enzymes have not yielded to crystallization. Many drug substrates have lower active site affinity and/or multiple orientations not suitable to determining clear X-ray structures. There are no structures of human P450 enzymes with reductase or cytochrome b5 catalytic partners. As a result, drug design is limited by available structural information. We propose employing solution NMR as a newly-viable orthogonal method to obtain the requisite amino acid- level structural information needed to understand human P450/ligand and P450/protein interactions. While the size, stability, and the absence of information relating individual NMR resonances to the corresponding amino acid have all thus far prevented the determination of any human P450 structures by NMR, we have the combined expertise and preliminary data to demonstrate that this feat is now technically possible. The Pochapsky lab previously developed the expertise to determine solution NMR structures of slightly smaller, soluble bacterial P450 enzymes. The Scott lab developed the capacity to generate human membrane P450 enzymes in the isotopically-labeled forms, with the amounts and with the stability required for NMR experiments. Thus, based on substantial preliminary data, we propose to 1) advance strategies for determining human P450 structures by solution NMR while determining the human steroidogenic CYP17A1 structure and 2) apply solution NMR strategies to the two most important human drug-metabolizing P450 enzymes, CYP3A4 and CYP2D6. Successful completion of aim 1 will not only further establish the feasibility of NMR structures for human membrane P450 enzymes, but will do so for an important prostate cancer drug target for which additional structural information is essential to further drug design. Results of both aims will provide a reference set of amino acid assignments that can then be readily used by a wide range of non-NMR experts, in much simpler experiments, to quickly determine where and how drugs and other proteins bind to three clinically-important P450 enzymes, as well as deciphering the changes in protein conformation required for these events.

PROJECT NARRATIVE Cytochrome P450 enzymes break down most drugs and are the target of other drugs for cancers and infections. To treat diseases more effectively, we need better and more images of how these enzymes work. The proposed research generates baseline information necessary to rapidly determine how drugs bind using a new technology.