Ligand-Receptor Interactions and Machine Learning in GCGR and GLP-1R Drug Discovery

Abstract

The large amount of data that has been collected so far for G protein-coupled receptors requires machine learning (ML) approaches to fully exploit its potential. Our previous ML model based on gradient boosting used for prediction of drug affinity and selectivity for a receptor subtype was compared with explicit information on ligand-receptor interactions from induced-fit docking. Both methods have proved their usefulness in drug response predictions. Yet, their successful combination still requires allosteric/orthosteric assignment of ligands from datasets. Our ligand datasets included activities of two members of the secretin receptor family: GCGR and GLP-1R. Simultaneous activation of two or three receptors of this family by dual or triple agonists is not a typical kind of information included in compound databases. A precise allosteric/orthosteric ligand assignment requires a continuous update based on new structural and biological data. This data incompleteness remains the main obstacle for current ML methods applied to class B GPCR drug discovery. Even so, for these two class B receptors, our ligand-based ML model demonstrated high accuracy (5-fold cross-validation Q2 > 0.63 and Q2 > 0.67 for GLP-1R and GCGR, respectively). In addition, we performed a ligand annotation using recent cryogenic-electron microscopy (cryo-EM) and X-ray crystallographic data on small-molecule complexes of GCGR and GLP-1R. As a result, we assigned GLP-1R and GCGR actives deposited in ChEMBL to four small-molecule binding sites occupied by positive and negative allosteric modulators and a full agonist. Annotated compounds were added to our recently released repository of GPCR data.

Drug Repositioning For Allosteric Modulation of VIP and PACAP Receptors

“…Moreover, an allosteric mode of ticagrelor inhibition of VIP receptors is in agreement with recently released PDB structures of class B GPCRs which we described in ( 50 ). Namely, recently determined class B structures included only small molecule agonists in the orthosteric TMD binding site but not inhibitors.…”

“…To elucidate details of ticagrelor binding modes and the basis of its receptor subtype selectivity we performed molecular dynamics (MD) simulations. Obtained results were consistent with so far confirmed basis of molecular recognition of other class B GPCRs by their negative allosteric modulators binding to the extrahelical active site located in a hydrophobic environment of a lipid bilayer (48)(49)(50). Altogether, our results suggest that ticagrelor acts as a negative allosteric modulator of VIP and PACAP receptors with weak selectivity for the VPAC2 receptor sub-type.…”

“…We focused on two binding sites of VPAC1: orthosteric (TMD and ECD) and allosteric (TMD) – for negative allosteric modulators. So far, only structures of class A GPCRs, not class B GPCRs, included antagonists in their orthosteric binding sites ( 50 ). Out of ca.…”

“…In Supplementary Figures S6– S8 we presented discarded, alternative binding modes of ticagrelor to orthosteric TMD and ECD binding sites. Yet, as we mentioned above, these alternative binding modes of ticagrelor seem much less probable as none among class B GPCR structures with a small molecule orthosteric antagonist has been released in PDB so far ( 50 ).…”

DataSheet_1.pdf

“…At the time of the beginning of this study, cryo-EM structures of class B GPCRs with positive allosteric modulators (50,64) were not known, so we have not included them in virtual screening. Following our recent results described elsewhere (41) TMD and ECD regions forming the orthosteric binding site were treated separately in VS. Namely, we screened ZINC15 compounds with the VIP-binding orthosteric site of VPAC1 divided into the TMD part and the ECD part.…”