Identification of a novel P2X7 antagonist using structure-based virtual screening

P2X4 and P2X7 receptors are ATP-gated ion channels, which play important roles in neuropathic and inflammatory pain, and as such they are important drug targets in diseases of inflammatory origin. While several compounds targeting P2X4 and P2X7 receptors have been developed using traditional high-throughput screening approaches, relatively few compounds have been developed using structure-based design. We initially set out to develop compounds targeting human P2X4, by performing virtual screening on the orthosteric (ATP-binding) pocket of a molecular model of human P2X4 based on the crystal structure of the Danio rerio receptor. The screening of a library of approximately 300,000 commercially available drug-like compounds led to the initial selection of 17 compounds; however, none of these compounds displayed a significant antagonist effect at P2X4 in a Fluo-4 ATP-induced calcium influx assay. When the same set of compounds was tested against human P2X7 in an ATP-stimulated Yo-Pro1 dye uptake assay, one compound (an indeno(1,2-b)pyridine derivative; GP-25) reduced the response by greater than 50% when applied at a concentration of 30 µM. GP-25 displayed an IC(50) value of 8.7 μM at human P2X7 and 24.4 μM at rat P2X7, and was confirmed to be active using whole-cell patch clamp electrophysiology and not cytotoxic. Schild analysis suggested that mode of action of GP-25 was orthosteric. Screening of a further 16 commercially available analogues of GP-25 led to the discovery of five additional compounds with antagonist activity at human P2X7, enabling us to investigate the structure-activity relationship. Finally, docking of the R- and S-enantiomers of GP-25 into the orthosteric pocket of molecular models of human P2X4 and human P2X7 revealed that, while both enantiomers were able to make multiple interactions between their carboxyl moieties and conserved positively charged amino-acids in human P2X7, only the S-enantiomer of GP-25 was able to do this in human P2X4, potentially explaining the lack of activity of GP-25 at this receptor.