Identification of Acetylcholinesterase Inhibitors Through a Pharmacophore-Guided Deep Learning Approach for Therapeutic Applications in Alzheimer's Disease

Abstract

Acetylcholinesterase (AChE) metabolizes the neurotransmitter acetylcholine (ACh), vital for signal transmission in neurons and the central nervous system (CNS). Decreased ACh leads to Alzheimer's disease (AD) and cognitive dysfunction syndromes. This research identified novel AChE inhibitors that deplete ACh, hindering AChE protein activity. Advanced chemoinformatics approaches, including de novo design, molecular docking, and pharmacokinetic analysis, were used to design potential AChE inhibitors. New chemical entities were generated from known drug pharmacophoric features, followed by molecular docking and pharmacokinetic analyses, resulting in four potential AChE molecules: AChE_M1, AChE_M2, AChE_M3, and AChE_M4. The molecular docking revealed binding energies of −10.80, −11.30, −9.80, and -10.70 kcal/mol for AChE_M1, AChE_M2, AChE_M3, and AChE_M4, respectively, which is better than the cocrystal ligand and Donepezil. Several binding interactions were observed between the proposed molecules and the AChE protein. All molecules exhibited acceptable pharmacokinetic profiles and were nontoxic. The MDS metrics indicated stability at the AChE active site. Low-energy basins and atomic mobility in principal component analyses of AChE bound to the final molecules confirmed their strong affinity for the protein, highlighting the molecules' potential. The final compounds may represent promising candidates for CNS-related healthcare, subjected to validation through in vitro and in vivo studies.