Abstract
Human d-amino acid oxidase (h-DAAO) can effectively act on d-serine, which has been actively explored as a novel therapeutic target for treating schizophrenia. In this study, 37 h-DAAO inhibitors based on a 6-hydroxy-1,2,4-triazine-3,5(2H,4H)-dione scaffold were obtained to construct the optimal comparative molecular field analysis (CoMFA, q (2) = 0.613, r (2) = 0.966) and comparative molecular similarity index analysis (CoMSIA, q (2) = 0.669, r (2) = 0.985) models. The results indicate that the models have good predictability and strong stability. Furthermore, contour maps of the three-dimensional quantitative structure-activity relationship (3D-QSAR) revealed the relationships between the structural features and inhibitory activity. A total of nine new h-DAAO inhibitors were designed, which exhibited good predicted pIC(50) values. Through molecular docking and molecular dynamics simulation, four essential residues (i.e., Gly313, Arg283, Tyr224 and Tyr228) were considered to interact with the inhibitor. The hydrogen bonds produced by the triazine structure with protein and the hydrophobic interactions with the residues (i.e., Leu51, His217, Gln53 and Leu215) play an important role in the stability of the inhibitor at the binding site of the protein. Additionally, the compounds D1, D3 and D8, with higher predicted activities, were selected by ADME and bioavailability prediction. The present study could offer a reliable theoretical basis for future structural optimisation, design and synthesis of effective antipsychotics.