Computer-aided discovery of novel non-peptide inhibitors against amyloid-beta (Aβ) peptide aggregation for treating Alzheimer's disease

A non-peptide inhibitor that is metabolically stable, orally active and capable of crossing the blood-brain barrier has been a popular option for treating Alzheimer's disease (AD). To identify novel non-peptide inhibitors for AD drug development, a structure-based pharmacophore model (SBPM) was developed using the representative docked conformation of the recently discovered peptide inhibitor PGKLVYA in the potential binding site on the Aβ(17-42) protofibril. The best SBPM, consisting of two hydrophobic, one hydrogen bond donor, and one positive ionisable feature, was further validated using ligand pharmacophore mapping studies. The well-validated SBPM was then used as the 3D query in virtual screening to identify potential hits from the National Cancer Institute database. These hits were subsequently filtered by toxicity prediction and molecular docking, and their binding stabilities and affinities were validated by 20-ns molecular dynamics simulations and molecular mechanics Poisson-Boltzmann surface area analysis, respectively. Finally, two Hits (NSC35984 and NSC102747) were identified as potential leads, which exhibited higher binding stability and affinity towards Aβ compared with PGKVYA. Our results also suggest that these two Hits have the ability to prevent Aβ adopting toxic β-sheet structures, and can be easily synthesised and have structural novelty, indicating that they are promising candidates for treating AD.