Ligand‐binding affinity of alternative conformers of human β₂‐adrenergic receptor in the presence of intracellular loop 3 (ICL3) and their potential use in virtual screening studies

Abstract

This study investigates the structural distinctiveness of orthosteric ligand‐binding sites of several human β2 adrenergic receptor (β2‐AR) conformations that have been obtained from a set of independent molecular dynamics (MD) simulations in the presence of intracellular loop 3 (ICL3). A docking protocol was established in order to classify each receptor conformation via its binding affinity to selected ligands with known efficacy. This work's main goal was to reveal many subtle features of the ligand‐binding site, presenting alternative conformations, which might be considered as either active‐ or inactive‐like but mostly specific for that ligand. Agonists, inverse agonists, and antagonists were docked to each MD conformer with distinct binding pockets, using different docking tools and scoring functions. Mostly favored receptor conformation persistently observed in all docking/scoring evaluations was classified as active or inactive based on the type of ligand's biological effect. Classified MD conformers were further tested for their ability to discriminate agonists from inverse agonists/antagonists, and several conformers were proposed as important targets to be used in virtual screening experiments that were often limited to a single X‐ray structure.

Intrinsic Dynamics and Causality in Correlated Motions Unraveled in Two Distinct Inactive States of Human β2-Adrenergic Receptor

“…42 Thus, in this novel inactive state, it is expected that the packing of ICL3 and consequently its decreased mobility would play a significant role in the overall receptor's dynamics by preventing the two domains to alter their relative orientations and consequently reducing the overall basal activity of the receptor. Also, based on the distribution of driver−follower associations, which shifted from one domain involving H5−H6 to the other incorporating H1−H7, it is suggested that phase II might represent a more favorable conformational state for β-arrestin coupling as H7 is well known to play a role in arrestin binding as reported in a 19 F-NMR study by Liu et al, 43 which points to a significant level of decoupling of G protein and β-arrestin signaling pathways. Also, it was experimentally observed that the phosphorylation of H8, which is the small appendage of H7, and also the C-terminal tail recruited arrestin binding.…”

“…In each of these runs, ICL3 preserved its packed form, which displayed only minor local fluctuations. This novel inactive state, which displayed a distinct character in both its intra- and extracellular parts, was later used in virtual screening experiments and successfully discriminated antagonists from agonists …”

“…This novel inactive state, which displayed a distinct character in both its intra-and extracellular parts, was later used in virtual screening experiments and successfully discriminated antagonists from agonists. 19 Moreover, a theoretical approach, the Anisotropic Network Model (ANM) which is an elastic network model 20 was used in order to confirm whether the same type of ICL3 closure would still be observable using several different models and conformations of ICL3. 8 Indeed, ANM modes exhibited high overlap with the first mode (principal components) of MD runs, which clearly demonstrated that slow modes of ANM drove the closure of ICL3, independent of ICL3's initial model/conformation.…”

Distinctive communication networks in inactive states of β2‐adrenergic receptor: Mutual information and entropy transfer analysis

“…This novel inactive state also revealed a relatively expanded orthosteric ligand‐binding site mostly due to the outward tilt of transmembrane helix 5 (H5). Consequently, the expanded volume of the binding cavity efficiently accommodated large size antagonists and inverse agonists as opposed to small size agonists, which was reported in a prior virtual screening experiment 14 . The fully atomistic model of the receptor had the completed structure of ICL3 predicted via homology modeling, which was missing in all X‐ray structures 3,15‐17 .…”

Identification of Alternative Allosteric Sites in Glycolytic Enzymes for Potential Use as Species-Specific Drug Targets

“…Several well-established approaches exist to detect alternative allosteric sites. Some relies on static structures of proteins acquired from NMR or X-ray experimental studies, while others investigate large scale motions such as hinge bending via normal mode analysis (NMA) using coarse-grained elastic network model (Bahar and Rader, 2005;Tama and Brooks, 2006) or molecular dynamics simulations (Hornak et al, 2006;Lou and Cukier, 2006;Dilcan et al, 2019), since large scale motions involving large domains can be correlated with protein function. Moreover, large scale motions defined by the slowest frequency modes present an intrinsic feature of the protein (Tobi and Bahar, 2005) and also defines the distant couplings which is the nature of allostery.…”