No abstract

A major challenge in the application of structure-based drug design methods to proteins belonging to the superfamily of G protein-coupled receptors (GPCRs) is the paucity of structural information (1). The 19 chemokine receptors, belonging to the Class A family of GPCRs, are important drug targets not only for autoimmune diseases like multiple sclerosis but also for the blockade of human immunodeficiency virus type 1 entry (2). Using the MembStruk computational method (3), we predicted the three-dimensional structure of the human CCR1 receptor. In addition, we predicted the binding site of the small molecule CCR1 antagonist BX 471, which is currently in Phase II clinical trials (4). Based on the predicted antagonist binding site we designed 17 point mutants of CCR1 to validate the predictions. Subsequent competitive ligand binding and chemotaxis experiments with these mutants gave an excellent correlation to these predictions. In particular, we find that Tyr-113 and Tyr-114 on transmembrane domain 3 and Ile-259 on transmembrane 6 contribute significantly to the binding of BX 471. Finally, we used the predicted and validated structure of CCR1 in a virtual screening validation of the Maybridge data base, seeded with selective CCR1 antagonists. The screen identified 63% of CCR1 antagonists in the top 5% of the hits. Our results indicate that rational drug design for GPCR targets is a feasible approach.Chemokines belong to a large family of small, chemotactic cytokines that regulate the trafficking of immune cells (5) by binding to cell surface receptors belonging to the GPCR 3 superfamily (5). CCR1, the first CC chemokine receptor to be identified, responds to a number of ligands, including MIP-1␣ (CCL3) and RANTES (regulated on activation normal T cell expressed and secreted) (CCL5) (6, 7). The strong association with a wide variety of autoimmune and pro-inflammatory diseases has made the CCR1 protein an attractive therapeutic target, and Berlex has developed a potent, specific, orally available antagonist, BX 471, currently in a Phase II clinical trial (8).The CCR1 antagonist program that yielded the clinical compound BX 471 followed a traditional drug discovery approach starting with high throughput screening of large compound libraries (9). Although high throughput screening is a main pillar of drug-finding programs in the pharmaceutical industry, it has recently been supplemented by in silico methods to maximize the probability of finding attractive novel leads. Structurebased in silico approaches have been challenging for GPCRs, because only one experimental GPCR structure, that of bovine rhodopsin, with only ϳ20% sequence identity to CCR1 (10), has been reported. Recent developments in GPCR structure prediction methods show great potential for structure-based drug design and identifying novel hits from virtual screens (11)(12)(13)(14)(15).In this communication we report a significant test of the computational method MembStruk by predicting the structure of CCR1. Further, we scanned the entire predicted struc...

BACKGROUND AND PURPOSE GPR35 is a poorly characterized G protein‐coupled receptor at which kynurenic acid has been suggested to be the endogenous ligand. We wished to test this and develop assays appropriate for the study of this receptor. EXPERIMENTAL APPROACH Human and rat orthologues of GPR35 were engineered and expressed and assays developed to assess interaction with β‐arrestin‐2, activation of Gα13 and agonist‐induced internalization. KEY RESULTS GPR35‐β‐arrestin‐2 interaction assays confirmed that both the endogenous tryptophan metabolite kynurenic acid and the synthetic ligand zaprinast had agonist action at each orthologue. Zaprinast was substantially more potent than kynurenic acid at each and both agonists displayed substantially greater potency at rat GPR35. Two novel thiazolidinediones also displayed agonism and displayed similar potency at each GPR35 orthologue. The three ligand classes acted orthosterically with respect to each other, suggesting overlapping binding sites and, consistent with this, mutation to alanine of the conserved arginine at position 3.36 or tyrosine 3.32 in transmembrane domain III abolished β‐arrestin‐2 recruitment in response to each ligand at each orthologue. CONCLUSIONS AND IMPLICATIONS These studies indicate that β‐arrestin‐2 interaction assays are highly appropriate to explore the pharmacology of GPR35 and that Gα13 activation is an alternative avenue of signal generation from GPR35. Arginine and tyrosine residues in transmembrane domain III are integral to agonist recognition and function of this receptor. The potency of kynurenic acid at human GPR35 is sufficiently low, however, to question whether it is likely to be the true endogenous ligand for this receptor.

No abstract

No abstract