Selective Deuteration Improves the Affinity of Adenosine A_2A Receptor Ligands: A Computational Case Study with...

We used a range of computational techniques to assess the effect of selective C-H deuteration on the antagonist istradefylline affinity for the adenosine A_2A receptor, which was discussed relative to its structural analogue caffeine, a well-known and likely the most widely used stimulant. The obtained results revealed that smaller caffeine shows high receptor flexibility and exchanges between two distinct poses, which agrees with crystallographic data. In contrast, the additional C8-trans-styryl fragment in istradefylline locks the ligand within a uniform binding pose, while contributing to the affinity through the C-H···π and π···π contacts with surface residues, which, together with its much lower hydration prior to binding, enhances the affinity over caffeine. In addition, the aromatic C8-unit shows a higher deuteration sensitivity over the xanthine part, so when both of its methoxy groups are d₆-deuterated, the affinity improvement is -0.4 kcal mol^-1, which surpasses the overall affinity gain of -0.3 kcal mol^-1 in the perdeuterated d₉-caffeine. Yet, the latter predicts around 1.7-fold potency increase, being relevant for its pharmaceutical implementations, and also those within the coffee and energy drink production industries. Still, the full potential of our strategy is achieved in polydeuterated d₁₉-istradefylline, whose A_2A affinity improves by -0.6 kcal mol^-1, signifying a 2.8-fold potency increase that strongly promotes it as a potential synthetic target. This knowledge supports deuterium application in drug design, and while the literature already reports about over 20 deuterated drugs currently in the clinical development, it is easily foreseen that more examples will hit the market in the years to come. With this in mind, we propose that the devised computational methodology, involving the ONIOM division of the QM region for the ligand and the MM region for its environment, with an implicit quantization of nuclear motions relevant for the H/D exchange, allows fast and efficient estimates of the binding isotope effects in any biological system.