Structure-Based Consensus Scoring Scheme for Selecting Class A Aminergic GPCR Fragments

详细信息    查看全文

• 作者：Ádám A. Kelemen ; Róbert Kiss ; György G. Ferenczy ; László Kovács ; Beáta Flachner ; Zsolt Lőrincz ; György M. Keserű
• 刊名：Journal of Chemical Information and Modeling
• 出版年：2016
• 出版时间：February 22, 2016
• 年：2016
• 卷：56
• 期：2
• 页码：412-422
• 全文大小：407K
• ISSN：1549-960X

文摘

Aminergic G-protein coupled receptors (GPRCs) represent well-known targets of central nervous-system related diseases. In this study a structure-based consensus virtual screening scheme was developed for designing targeted fragment libraries against class A aminergic GPCRs. Nine representative aminergic GPCR structures were selected by first clustering available X-ray structures and then choosing the one in each cluster that performs best in self-docking calculations. A consensus scoring protocol was developed using known promiscuous aminergic ligands and decoys as a training set. The consensus score (FrACS—fragment aminergic consensus score) calculated for the optimized protein ensemble showed improved enrichments in most cases as compared to stand-alone structures. Retrospective validation was carried out on public screening data for aminergic targets (5-HT₁ serotonin receptor, TA₁ trace-amine receptor) showing 8–17-fold enrichments using an ensemble of aminergic receptor structures. The performance of the structure based FrACS in combination with our ligand-based prefilter (FrAGS) was investigated both in a retrospective validation on the ChEMBL database and in a prospective validation on an in-house fragment library. In prospective validation virtual fragment hits were tested on 5-HT₆ serotonin receptors not involved in the development of FrACS. Six out of the 36 experimentally tested fragments exhibited remarkable antagonist efficacies, and 4 showed IC₅₀ values in the low micromolar or submicromolar range in a cell-based assay. Both retrospective and prospective validations revealed that the methodology is suitable for designing focused class A GPCR fragment libraries from large screening decks, commercial compound collections, or virtual databases.