In silico discovery of potent and selective Janus kinase 3 (JAK3) inhibitors through 3D-QSAR, covalent docking, ADMET analysis, molecular dynamics simulations, and binding free energy of pyrazolopyrimidine derivatives

Abstract

Rheumatoid arthritis is a prevalent and debilitating chronic disease worldwide. Targeting Janus kinase 3 (JAK3) has emerged as a crucial molecular strategy to treat this condition. In this study, we employed a comprehensive theoretical approach that included 3D-QSAR, covalent docking, ADMET, and molecular dynamics to propose and optimize new anti-JAK3 compounds. We investigated a series of 28 1H-pyrazolo[3.4-d]pyrimidin-4-amino inhibitors and developed a highly accurate 3D-QSAR model using comparative molecular similarity index analysis (COMSIA). The model predicted with Q2 = 0.59, R2 = 0.96, and R2 (Pred) = 0.89, was validated using Y-randomization and external validation methods. Our covalent docking studies identified T3 and T5 as highly potent inhibitors of JAK3 compared to the reference ligand 17. Additionally, we evaluated the ADMET properties and drug similarity of our newly developed compounds and reference ligand, providing critical insights for further optimization of anti-JAK3 medications. Furthermore, MM-GBSA analysis showed promising results for the designed compounds. Finally, we validated our docking results using molecular dynamics simulations, which confirmed the stability of hydrogen bonding contacts with key residues required to block JAK3 activity. Our findings offer new chemical scaffolds and insights that could lead to the development of novel and effective JAK3 therapeutic targets for treating rheumatoid arthritis. Communicated by Ramaswamy H. Sarma