Discovery of drug‐like acetylcholinesterase inhibitors by rapid virtual screening of a 6.9 million compound database

Cholinesterase inhibitors remain the mainstay of Alzheimer's disease treatment, and the search for new inhibitors with better efficacy and side effect profiles is ongoing. Virtual screening (VS) is a powerful technique for searching large compound databases for potential hits. This study used a sequential VS workflow combining ligand‐based VS, molecular docking and physicochemical filtering to screen for central nervous system (CNS) drug‐like acetylcholinesterase inhibitors (AChEIs) amongst the 6.9 million compounds of the CoCoCo database. Eleven in silico hits were initially selected, resulting in the discovery of an AChEI with a Ki of 3.2 µM. In vitro kinetics and in silico molecular dynamics experiments informed the selection of an additional seven analogues. This led to the discovery of two further AChEIs, with Ki values of 2.9 µM and 0.65 µM. All three compounds exhibited reversible, mixed inhibition of acetylcholinesterase. Importantly, the in silico physicochemical filter facilitated the discovery of CNS drug‐like compounds, such that all three inhibitors displayed high in vitro blood‐brain barrier model permeability. Virtual screening and physicochemical filtering of the 6.9 million compound CoCoCo database led to the discovery of three acetylcholinesterase inhibitors with low to sub‐micromolar in vitro potency. Further investigation showed a reversible mechanism of inhibition with mixed‐model kinetics. The in silico physicochemical filters applied during virtual screening ensured the compounds were CNS drug‐like, with high parallel artificial membrane permeability assay‐blood‐brain barrier permeability.