Combining structure-based pharmacophore and in silico approaches to discover novel selective serotonin reuptake inhibitors

Inhibition of human serotonin transporter (hSERT) has been reported to be a potent strategy for the treatment for depression. To discover novel selective serotonin reuptake inhibitors (SSRIs), a structure-based pharmacophore model (SBPM) was developed using the docked conformations of six highly active SSRIs. The best SBPM, consisting of four chemical features: two ring aromatics (RAs), one hydrophobic (HY), and one positive ionizable (PI), was further validated using Gunner-Henry (GH) scoring and receiver operating characteristic (ROC) curve methods. This well-validated SBPM was then used as a 3D-query in virtual screening to identify potential hits from National Cancer Institute (NCI) database. These hits were subsequently filtered by absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction and molecular docking, and their binding stabilities were validated by 20-ns MD simulations. Finally, only two compounds (NSC175176 and NSC705841) were identified as potential leads, which exhibited higher binding affinities in comparison with the paroxetine. Our results also suggest that cation-π interaction plays a crucial role in stabilizing the hSERT-inhibitor complex. To our knowledge, the present work is the first structure-based virtual screening study for new SSRI discovery, which should be a useful guide for the rapid identification of novel therapeutic agents from chemical database. In this study, we firstly combined the structure-based pharmacophore model with other CADD approaches, including virtual screening, ADMET prediction, molecular docking and MD simulation to search potent SSRIs from NCI database. Finally, only two compounds were identified as potential leads.