||Raloxifene was approved in 2007 by the FDA for the chemoprevention of breast cancer in postmenopausal women with osteoporosis and in postmenopausal women at high risk for invasive breast cancer. The FDA's decision was based, in part, on a more favorable safety profile for raloxifene, as compared to the standard treatment of tamoxifen. However, recent studies have demonstrated the ability of raloxifene to form reactive intermediates and to act as a mechanism-based inactivator of cytochrome P450 3A4 (CYP3A4). The reactive raloxifene intermediate was theorized to be produced either through CYP3A4-mediated dehydrogenation to a di-quinone methide or through the more common oxygenation route to an arene oxide. However, the operative mechanism was never identified. The goal of this dissertation was to characterize the reactive intermediate formation from raloxifene, specifically the ability of CYP3A4 to catalyze the dehydrogenation of raloxifene to a di-quinone methide, using a combination of convent•i onal b•i oc«h emi•c ali and-i computati• onal techn•i ques. In the current woir k,1 8O incorporation studies were utilized to differentiate CYP3A4-mediated oxygenation versus dehydrogenation of raloxifene. These studies established that 3'-hydroxyraloxifene is produced exclusively via CYP3A4-mediated oxygenation with molecular O2. These studies also provided convincing evidence for the mechanism of CYP3A4-mediated dehydrogenation of raloxifene to a reactive di-quinone methide, and excluded the alternative arene oxide pathway. Furthermore, it was demonstrated that 7-hydroxyraloxifene, which was previously believed to be a typical 02-derived metabolite of CYP3A4, is in fact produced by a highly unusual hydrolysis pathway from a putative ester, formed by conjugation of a carboxylic acid moiety of CYP3A4 or another protein, with the di-quinone methide of raloxifene. The use of molecular modeling in conjunction with site-directed mutagenesis has extensively been used to study substrate orientation within the P450 active sites, and to identify potential residues involved in positioning and/or catalysis of these substrates. However, the effectiveness of these studies is highly dependent on the selection of the most accurate enzyme crystal structure. In the current work, we compared the ability of two commonly used CYP3A4 crystal structures, 1TQN and 1 WOE, to predict the sites of metabolism of two known CYP3A4 substrates, indapamide and raloxifene. Indapamide was used to evaluate the accuracy of the molecular model, while raloxifene was used to investigate the effects of adding partial charges to the P450 heme moiety to improve predictions of metabolic pathways by computation methods. The results demonstrated that while docking studies with both 1TQN and 1 WOE crystal structures could accurately predict the sites of indapamide metabolism, the interactions between the substrate and active site residues were different for each crystal structure. The addition of partial charges to the heme moiety of 1 WOE dramatically increased the predictive power of the model, and dramatically increased the accuracy of the model, regarding substrate/active site residue interactions. To determine the validity of our models, active site residues involved in the positioning and/or catalysis of the substrates were identified, and site-directed mutagenesis studies were conducted. Mutations based on 1TQN and 1 WOE (without heme partial charges) models did not alter CYP3A4-mediated dehydrogenation of raloxifene, demonstrating that these models were not accurate enough to correctly predict enzyme/substrate interactions. However, the 1 WOE (with heme partial charges) model identified the Phe215 residue as an important component of the enzyme tertiary structure that would interact with raloxifene. Site-directed mutagenesis experiments with Phe215 confirmed that this residue positions raloxifene to favor the dehydrogenation pathway, most likely through a combination of n bond T-stacking and steric interactions. These results validate the use of the 1 WOE (with heme partial charges) model for docking studies of raloxifene, and demonstrate the usefulness of molecular modeling in conjunction with site-directed mutagenesis to study the mechanisms of P450-mediated metabolism.