Multi-target QSAR modeling for the identification of novel inhibitors against Alzheimer's disease

Abstract

Alzheimer's disease (AD) is an age-related neurodegenerative disorder, which is the most common cause of dementia in elderly individuals. It is characterized by selective neuronal cell death that affects the brain area related to memory and learning. So far, various computational research targeting AD have been reported, but we are still far from finding a precise treatment strategy for AD. It appeared of interest to us to carry out a two-dimensional quantitative structure-activity relationship (2D-QSAR) analysis against multiple targets of AD using large datasets to determine the essential structural features which are responsible for the inhibition of the enzymes/targets. In the present research, we have implemented 2D-QSAR modeling against twelve major targets (AChE, BuChE, BACE1, β-amyloid, 5-HT6, CDK-5, Gamma-secretase, Glutaminyl Cyclase, GSK-3β, MAO-B, NMDA and Phosphodiester (PDE10A) enzymes) of AD for the identifications of novel multitarget inhibitors. The models were used to check the applicability domain of a pool of ∼19 million compounds obtained from the four chemical drug-like databases (ZINC12, Asinex, NCI, and InterBioscreen databases) and provided prioritized compounds for experimental detection of their performance as anti-Alzheimer's drug. Additionally, we have also developed the quantitative structure activity-activity relationship (QSAAR) and selectivity-based models to explore the most important features contributing to the dual inhibition against the respective targets. Furthermore, we have also performed chemical Read-Across predictions using the Read-Across-v3.1 tool (https://dtclab.webs.com/software-tools), the results for the external validation metrics were found to be better than the 2D-QSAR-derived predictions. Furthermore, molecular docking experiments have been performed to understand the molecular interactions between ligands and enzymes at the atomic level, and the observations are compared with the structural features acquired from QSAR models that justified the mechanistic aspect of binding phenomena. The proposed models and read-across hypotheses could be used as potential tools to identify essential molecular features for designing suitable drug(s) for Alzheimer's therapy using rational design of multi-target inhibitors.