In silico-guided development of novel benzothiazole-linked diazepino derivatives with improved BBB penetration for Alzheimer's disease through AChE inhibition

Predicting chemical cytotoxicity and brain-targeted penetration remains a major challenge in developing therapeutics for neurological disorders. To address this challenge, we designed a series of novel benzothiazole-linked diazepino derivatives and evaluated their in silico ADMET and blood–brain barrier (BBB) permeability profiles. The designed compounds were predicted to exhibit partial BBB permeability and were identified as non-substrates of P-glycoprotein, suggesting a favorable pharmacokinetic profile for CNS drug development. In addition, in silico toxicity prediction and LD₅₀ estimation indicated a comparatively favorable safety profile for the designed molecules. Molecular docking studies revealed key ligand–protein interactions, including hydrogen bonding and π–π stacking with important active-site residues of acetylcholinesterase (AChE). The target compounds, namely 2-(phenyl)− 1-[6-(trifluoromethoxy)− 1,3-benzothiazol-2-yl]− 1,3-diazepane−4,7-diones, were successfully synthesized and evaluated for AChE inhibitory activity using Ellman's assay. Among the synthesized derivatives, compound DR22 exhibited the highest inhibitory activity with an IC₅₀ value of 0.9 × 10 3 μM, followed by compound DR21 with an IC₅₀ value of 0.88 × 10 3 μM. The in vivo efficacy of these compounds was further evaluated in a scopolamine-induced Alzheimer's disease rat model. Both DR22 and DR21 significantly improved memory performance by reducing escape latency in the Morris water maze compared with the disease control group (p < 0.001). Overall, these findings identify DR22 and DR21 as promising lead compounds for further optimization as potential anti-Alzheimer agents, supported by favorable computational pharmacoknetic predictions, comparatively low predicted acute toxicity, and encouraging in vitro and in vivo biological activities.