Updates‎ > ‎

Nature Scientific Reports publication: Computer-aided drug discovery à la carte

posted Jul 6, 2016, 12:39 AM by Albert Kooistra   [ updated Jul 6, 2016, 12:44 AM ]

Researchers from the division of Medicinal Chemistry have developed a new methodology, which not only allows for the in silico discovery of novel ligands for GPCRs, but also for the prediction of the functional effect of those ligands.

The Nature Scientific Reports publication virtual screening methodology combines a conventional energy-based scoring function with a molecular interaction fingerprinting (IFP) technique. This approach, together with the selection of an optimal reference X-ray structure, enabled the prospective identification of histamine H1 receptor antagonists and β2-adrenoceptor agonists.

Picture: Graphical summary of the workflow and application of the published computational method that enables the selective identification of novel GPCR ligands with the desired functional effect.

Dr. Albert J. Kooistra, Dr. Chris de Graaf, and colleagues utilized the growing amount of available GPCR X-ray structures to develop a computational method for the identification of GPCR ligands through virtual screening while simultaneously selecting for a specific functional effect. GPCRs form the largest family of transmembrane proteins encoded for by the human genome and are key drug targets due to their essential regulatory role in signal transduction and cell function. The published approach takes advantage of the increasing structural knowledge regarding the molecular determinants of ligand binding and, more specifically, function-specific binding of agonists, antagonists, and inverse agonists.

The combination of a conventional energy-based scoring function with a molecular interaction fingerprinting (IFP) technique and the optimal reference X-ray structure enabled the prospective identification of 19 histamine H1 receptor antagonists and 18 β2-adrenoceptor agonists. Systematic evaluation of X-ray structures and the optimized combined scoring approach allow, in a sense, to pick new GPCR ligands with high accuracy (hit-rates of 73% for H1R and 53% for β2R) from a database with millions of molecules and also to select the flavor of the ligand (i.e. agonist versus antagonist). Moreover, the performance of the consensus approach was also evaluated by experimentally validating the performance of the individual application of the energy-based as well as the IFP approach. In all cases the consensus methodology resulted in the highest hit-rates and identified compounds with the highest receptor binding affinity and potency.    

This study shows that the combination of cheminformatics approaches and the developments in the structural determination of GPCRs opens up new opportunities for targeted drug discovery. Ultimately, this approach could even be used to predict individual signaling pathways to design optimal biased signaling profiles for ligands of this pharmaceutically important protein family.

Click below to go to the open access article:

Kooistra, A.J., Vischer, H.F., McNaught-Flores, Leurs, R., de Esch, I.J.P., de Graaf, C. Function-specific virtual screening for GPCR ligands using a combined scoring method. Scientific Reports. 6 (2016): 28288.