Computational insight into crucial binding features for metabolic specificity of cytochrome P450 17A1

Abstract

In present study we aimed to explore the possible structural feature of Cytochrome P450 (CYP) 17A1 that contributes to the metabolic specificity. The predicable 3D-QSAR (Quantitative Structure-Activity Relationships) models were first developed with CoMFA (Comparative Molecular Field Analysis) and CoMSIA (Comparative Molecular Similarity Indices Analysis) methods based on a training set of 76 non-steroid inhibitors, then verified by a test set of 20 inhibitors. Our data demonstrates a cross-validation correlation coefficient q2s of 0.534 and 0.545, as well as a non-cross-validation correlation coefficient r2s of 0.904 and 0.889, respectively. The contours were generated to indicate the specific inhibitor feature. Further, molecular docking was used to probe the interacting feature between CYP17A1 and its non-steroid inhibitors or its natural substrates. We found the crucial binding features for the non-steroid inhibitor selection can be described as the suitable molecular length and the ability to fulfill the active pocket of CYP17A1, the hydrophobic parent body and the H-bond formed with special residues. Whereas, the distances between reaction site and the oxidative or the per-oxidative center played an important role in the substrate metabolism of 17α-hydroxylase and 17,20-lyase. This study helps to design and screen potential candidates for therapy of prostate cancer and other androgen-dependent diseases. Keywords: CYP17A1, Metabolic specificity, Inhibitors and substrates, 3D-QSAR, Docking