BioSysBio:abstracts/2007/Hyesun Han

In Silico Prediction System of CYP450-mediated Metabolism Profile
Author(s): Won Seok Oh1, Ji Hoon Jung1, Hyesun Han1, Doo Nam Kim2 and Kyoung Tai No1 Affiliations: 1 Department of Biotechnology, College of Engineering, Yonsei University, Seoul, Korea              2 Chem and Bioinformatics Division, Bioinformatics and Molecular     Design Research Center, Seoul, Korea Contact: trueeagle@hanmail.net Keywords: 'Cytochrome P450', 'Classification' 'Statistical Model' 'Modeling'

Abstract
The Cytochrome P450 (CYP450) enzymes represent an ideal subject for the investigation of metabolic drug. drug interactions. CYP450 is considered to be the most important single enzyme family in drug metabolism. This superfamily of enzymes is thought to metabolize approximately 90% of all marketed drugs. It is essential for the purpose of reducing time and resources of drug design and development to predict metabolism occurrence and regioselectivity. However, the lack of quantitative experimental metabolic data and substrate specificity, and the fact that CYP450 system depends on both steric effect and electronic reactivity indicate that metabolism prediction is difficult. In this study, we tried three approaches using statistical methods to classify substrate, empirical model to predict the activation energy of CYP450 reaction and combination of docking method and semi-empirical molecular orbital calculations to determine the binding mode of CYP450 enzyme-substrate complex. Each model gives a lot of insights to describe CYP450-mediated system in vivo. Statistical method is useful to separate potential substrates and non-substrate with only two dimensional descriptors. Empirical model can explain aliphatic hydroxylation and aromatic hydroxylation which altogether constitute most reactions mediated by CYP450 using AM1 quantum mechanical calculations. Third model gives major metabolic positions of substrates in CYP450. Further, knowledge collection of appropriate ligand and its binding site of CYP450 will help to follow up pharmacokinetics of novel compounds.