Grant: D/NZ2/05314

Development of a method for the structure prediction of GPCRs complexes with agonists and antagonists, including the ligand-induced alternations of the receptor structure

G protein-coupled receptors (GPCRs) constitute the largest and the most diverse superfamily of receptor proteins responsible for signal transduction across the plasma membrane. They are implicated in majority of physiological functions and control of pathological processes, thus, are very important pharmacological targets [1]. Currently, due to the difficulties in crystallization of these receptor proteins, there are x-ray crystallographic structures for only six members of the GPCR family available. Knowledge of the three-dimensional structure of a target protein is crucial for understanding of its mechanism of action and for the rational drug design.

Members of family A, GPCRs share the same topology – a bundle of seven transmembrane (TM) a‑helices – whereas the shapes and lengths of the N- and C-termini, as well as the cytoplasmic and extracellular loops, exhibit high variability. GPCRs interact with very diverse sets of ligands which bind to the TM segments and sometimes also to the extracellular domains of the receptor. Each receptor undergoes a series of conformational rearrangements, leading to the binding of a G protein, during the activation process [2]. Experiments suggest that an agonist binding and activation occurs through a series of conformational intermediates [3, 4]. Transition to these intermediate states involves the disruption of intramolecular interactions (“molecular switches”) that stabilize ground state of the receptor. Binding of structurally different agonists requires the disruption of different intramolecular interactions [3, 5], leading to different receptor conformations and differentiated effects on downstream signaling proteins. The dynamic character of GPCRs is likely to be essential for their physiological functions [3], and a better understanding of this molecular plasticity is of a great importance for the drug discovery.

Recently, there has been remarkable progress in theoretical methods aiming on more accurate and faster prediction of GPCR structure and function [6]. World-wide competition in structure  prediction of GPCR-ligand complexes [7], held in 2008 and 2010, demonstrated that computational methods guided by experimental data of various sort could be sometimes quite successful in the prediction of overall receptor structure and in the identification of the key protein ligand interactions. Unfortunately, none of the presented methods was capable of a fast and accurate conformation sampling of receptor-ligand complex in reasonable computational time. Moreover, correct predictions of the extracellular loops in GPCRs remains a challenging task. Knowledge of the loop conformations is crucial for comprehending ligand recognition [8] and other functional attributes of the structure, as well as for docking quality and Virtual Ligand Screening (VLS).

In this project a novel and accurate method for the structure prediction of GPCR-ligand complexes will be developed. This method will combine a coarse-grained modeling tool, CABS [9], adapted to the lipid membrane environment and expanded by allowing a coarse-grained representation of a variety of ligands and a proper knowledge-based scheme of the receptor-ligand interactions. The new method will employ a very efficient sampling of receptor conformations by means of the Replica Exchange Monte Carlo (REMC) algorithm. Possible ligand induced alternations of the receptor structure will be taken into account by the presence of the ligand inside the binding site during simulation. This innovative approach will also allow for loop prediction (especially the longest second extracellular loop), which is essential for accurate modeling of the binding cavity. Moreover, fast conformational sampling and a straightforward application of experimentally derived distance restraints will provide a framework for investigation of possible patterns of disulfide bonds present at the extracellular part of the protein.

The range of applicability and the accuracy of the developed method will be verified using the available GPCR crystallographic structures. Next, the method will be applied to structure prediction of a large set (~100) of members of subfamily A (“rhodopsin-like”) of GPCRs. To our knowledge, it will be the first attempt to predict structures of ligand-receptor complexes at such large scale. Obtained models will serve as a starting point for further investigation on GPCR selectivity, mechanisms of activation, identification of new molecular switches and design of new potent ligands by means of VLS or/and a structure-guided de novo design, selective toward a specific subtype of a receptor. Analysis of differences in binding of agonists and antagonists and investigation of probable mechanisms of activation for selected receptors will be carried out by means of molecular dynamics methods. The computed structural models of GPCR-ligand complexes will be also used for identification of thermo-stabilizing point mutations that may facilitate future crystallization of these receptors. Obtained structures of receptor-ligand complexes and the insight into molecular level mechanism of ligand binding, gathered during this project, will significantly increase our knowledge about GPCRs and will open new avenues for the computer‑aided rational drug design.


1. George, S.R., B.F. O'Dowd, and S.P. Lee, G-protein-coupled receptor oligomerization and its potential for drug discovery. Nat Rev Drug Discov, 2002. 1(10): p. 808-20.
2. Kobilka, B.K., G protein coupled receptor structure and activation. Biochim Biophys Acta, 2007. 1768(4): p. 794-807.
3. Kobilka, B.K. and X. Deupi, Conformational complexity of G-protein-coupled receptors. Trends Pharmacol Sci, 2007. 28(8): p. 397-406.
4. Kobilka, B., Agonist binding: a multistep process. Mol Pharmacol, 2004. 65(5): p. 1060-2.
5. Swaminath, G., et al., Probing the beta2 adrenoceptor binding site with catechol reveals differences in binding and activation by agonists and partial agonists. J Biol Chem, 2005. 280(23): p. 22165-71.
6. Katritch, V., et al., GPCR 3D homology models for ligand screening: lessons learned from blind predictions of adenosine A2a receptor complex. Proteins. 78(1): p. 197-211.
7. Michino, M., et al., Community-wide assessment of GPCR structure modelling and ligand docking: GPCR Dock 2008. Nat Rev Drug Discov, 2009. 8(6): p. 455-63.
8. Avlani, V.A., et al., Critical role for the second extracellular loop in the binding of both orthosteric and allosteric G protein-coupled receptor ligands. J Biol Chem, 2007. 282(35): p. 25677-86.
9. Kolinski, A., Protein modeling and structure prediction with a reduced representation. Acta Biochim Pol, 2004. 51(2): p. 349-71.