The recent progress in crystallography of G-protein coupled receptors opens an unprecedented venue for structure-based GPCR drug discovery. To test efficiency of the structure-based approach, we performed molecular docking and virtual ligand screening (VLS) of more than 4 million commercially available "drug-like" and ''lead-like'' compounds against the A 2A AR 2.6 Å resolution crystal structure. Out of 56 high ranking compounds tested in A 2A AR binding assays, 23 showed affinities under 10 µM, eleven of those had sub-µM affinities, and two compounds had affinities under 60 nM. The identified hits represent at least 9 different chemical scaffolds and are characterized by very high ligand efficiency (0.3-0.5 kcal/mol per heavy atom). Significant A 2A AR antagonist activities were confirmed for 10 out of 13 ligands tested in functional assays. High success rate, novelty and diversity of the chemical scaffolds and strong ligand efficiency of the A 2A AR antagonists identified in this study suggest practical applicability of receptor-based VLS in GPCR drug discovery.

Many G protein-coupled receptors (GPCRs), including the adenosine A(1) receptor (A(1)AR), have been shown to be allosterically modulated by small molecule ligands. So far, in the absence of structural information, the exact location of the allosteric site on the A(1)AR is not known. We synthesized a series of bivalent ligands (4) with an increasing linker length between the orthosteric and allosteric pharmacophores and used these as tools to search for the allosteric site on the A(1)AR. The compounds were tested in both equilibrium radioligand displacement and functional assays in the absence and presence of a reference allosteric enhancer, (2-amino-4,5-dimethyl-3-thienyl)-[3-(trifluoromethyl)phenyl]methanone, PD81,723 (1). Bivalent ligand N(6)-[2-amino-3-(3,4-dichlorobenzoyl)-4,5,6,7-tetrahydrothieno[2,3-c]pyridin-6-yl-9-nonyloxy-4-phenyl]-adenosine 4h (LUF6258) with a 9 carbon atom spacer did not show significant changes in affinity or potency in the presence of 1, indicating that this ligand bridged both sites on the receptor. Furthermore, 4h displayed an increase in efficacy, but not potency, compared to the parent, monovalent agonist 2. From molecular modeling studies, we speculate that the allosteric site of the A(1)AR is located in the proximity of the orthosteric site, possibly within the boundaries of the second extracellular loop of the receptor.

The crystal structure of the human A 2A adenosine receptor bound to the A 2A receptor-specific antagonist, ZM241385, was recently determined at 2.6-Å resolution. Surprisingly, the antagonist binds in an extended conformation, perpendicular to the plane of the membrane, and indicates a number of interactions unidentified before in ZM241385 recognition. To further understand the selectivity of ZM241385 for the human A 2A adenosine receptor, we examined the effect of mutating amino acid residues within the binding cavity likely to have key interactions and that have not been previously examined. Mutation of Phe-168 to Ala abolishes both agonist and antagonist binding as well as receptor activity, whereas mutation of this residue to Trp or Tyr had only moderate effects. The Met-177 3 Ala mutation impeded antagonist but not agonist binding. Finally, the Leu-249 3 Ala mutant showed neither agonist nor antagonist binding affinity. From our results and previously published mutagenesis data, we conclude that conserved residues Phe-168(5.29), Glu-169(5.30), Asn-253(6.55), and Leu-249(6.51) play a central role in coordinating the bicyclic core present in both agonists and antagonists. By combining the analysis of the mutagenesis data with a comparison of the sequences of different adenosine receptor subtypes from different species, we predict that the interactions that determine subtype selectivity reside in the more divergent "upper" region of the binding cavity while the "lower" part of the binding cavity is conserved across adenosine receptor subtypes.

The adenosine A(1) receptor is a promising therapeutic target for neurological disorders such as cognition deficits and is involved in cardiovascular preconditioning. Classically adenosine receptor agonists were all derivatives of adenosine, and thought to require a D-ribose moiety. More recently, however, the discovery of non-adenosine agonists for the human adenosine A(1) receptor (hA(1)R) has challenged this dogma (Beukers et al., 2004). In this study we characterize the tritiated form of one of these compounds, [(3)H]LUF5834, as the first non-ribose partial agonist radioligand with nanomolar affinity for the hA(1)R. Due to its partial agonist efficacy, [(3)H]LUF5834 labeled both G protein-coupled and uncoupled receptors with a similar high affinity. Using [(3)H]LUF5834 we performed competition binding experiments to characterize a range of A(1)R ligands varying in efficacy from the full agonist CPA to the inverse agonist DPCPX. Surprisingly, in the control condition both agonists and inverse agonists displayed biphasic isotherms. With the addition of 1mM GTP the high affinity isotherm of agonists or the low affinity isotherm of inverse agonists was lost revealing the mechanism of action of such inverse agonists at the A(1)R. Consequently, [(3)H]LUF5834 represents a novel high affinity radioligand for the A(1)R and may prove a useful tool to provide estimates of inverse agonist efficacy at this receptor.