The Simulated Binding of (±)-2,3-Dihydro-5,6-dimethoxy-2-[[1-(phenylmethyl)-4- piperidinyl]methyl]-1H-inden-1-one Hydrochloride (E2020) and Related Inhibitors to Free and Acylated Acetylcholinesterases and Corresponding Structure−Activity Analyses

Abstract

The simulated binding profiles of acetylcholine, ACh, and the inhibitor (+/-)-2,3-dihydro-5,6- dimethoxy-2-[[1-(phenylmethyl)-4-piperidinyl]methyl]-1H-inden-1-on e hydrochloride (E2020), 1, and some of its analogs to acetylcholinesterase, AChE, were determined using full force field energetics and allowing complete conformational flexibility in both the ligand and receptor. A new mode of binding of ACh to AChE was found which involves the carboxyl oxygen of ACh interacting with Gly 118 and 119. Multiple modes of binding of 1 and some of its analogs were found which include alignment models observed in previous more restricted modeling studies. The key ligand-receptor interactions identified, and the corresponding energetics, are consistent on a relative basis, with observed binding constants for both the individual isomers of each of the inhibitors, as well as among the inhibitors themselves. The multiple modes of binding of 1 to AChE arises from small changes in binding at a single subsite and also from multiple subsite changes. Thus, an independent subsite model for ligand-receptor binding holds for some modes of binding, but not for others. A comparison of the simulated AChE-1 (and analog inhibitors) binding models to the receptor-independent 3D-QSARs previously developed for this class of inhibitors reveals extensive mutual consistency. The findings from these two modeling studies provides greater guidelines for inhibitor design than can be realized from either one. The combined docking and 3D-QSAR studies permit a detailed understanding of the SAR of more than 100 compound 1 analog inhibitors. A simple molecular recognition model can also be gleaned from the docking studies. A cylindrical "plug" (the inhibitor) having a large dipole moment must sterically fit into a cylindrical hole (the active site gorge of AChE), the lining of which also has a large dipole moment. Our simulations suggest that the dynamic "back door" to the active site of AChE does not form a large enough opening for sufficiently long time periods so as to be an effective entrance/exit pathway.