Molecular modeling of potential PET imaging agents for adenosine receptor in Parkinson’s disease

Abstract

The adenosine receptors have appeared as potent and selective drug target in various diseases especially for central nervous system diseases. Adenosine receptor A2A antagonists have been known as potential treatment for Parkinson’s disease (PD). Radiolabeled A2AR antagonists can be used as positron emission tomography (PET) tracers and diagnostic tools for PD. In the present investigation, we perform the quantitative structure–activity relationship (QSAR) analysis and docking studies of a series of PET tracers as ligands and Adenosine receptors (A2AR) binding affinity, to elucidate the structural properties required for A2AR antagonist in treatment for PD. Several variable-selection methods were used to choose the descriptors that would lead to good QSAR model. Among several models developed, the best model was a five-variable multiple linear regression (MLR) equation with statistical parameters of squared correlation coefficient R 2 = 0.90 ± 0.01 and cross-validated correlation coefficient Q 2 = 0.84 ± 0.02. The QSAR models were also constructed for A2AR selectivity to tracer ligands, that yielded a four-variable model with R 2 = 0.94 ± 0.01 and Q 2 = 0.89 ± 0.02. The most important variables contributed in models construction involved: partial charge, hydrophobic atoms, rotatable bonds, polar van der Waals surface area, potential energy, and conformation-dependent charge descriptors. Finally, molecular docking analysis was carried out to better understand the interactions between ligands and Adenosine receptors. The importance of π-π stacking interactions between aromatic moiety of the ligands and triazine core of A2AR antagonist was confirmed.