3D-QSAR, Molecular Docking, Molecular Dynamics Simulation, and Pharmacokinetic Prediction of...

The inhibition of cellular inflammatory factor secretion by phosphatidylinositol- 3-kinase δ (PI3Kδ) makes it a novel target for acute lung injury therapy. This study aimed to elucidate the structure-activity relationship of 1H pyrazolo [3, 4-d] pyrimidine derivatives as PI3Kδ inhibitors for novel drug design.

In this study, a three-dimensional quantitative conformational relationship (3DQSAR) model was constructed using COMFA and CoMSIA techniques with Sybyl X-2.0 software. In-silico ADME and toxicity predictions were also evaluated using SwissADME and pkCSM tools. Molecular docking and molecular dynamics (MD) simulations (100 ns) were conducted to confirm the interaction of the compounds with the target protein using the GROMACS 2021 software package.

Good predictability was assessed using the CoMFA model (Qcv² = 0.547; Rncv² = 0.991; Rpred² = 0.996) and the best CoMSIA model (Qcv² = 0.58; Rncv² = 0.992; Rpred² = 0.994). The A, B, and C rings were the basic skeleton for the potency of the inhibitors. The hydrogen bond acceptor field, electrostatic field, and hydrophobic field significantly influenced activity. The newly designed T03 showed stable RMSD/RMSF during 100 ns MD simulations and low predicted toxicity (LD50: 1579.4 mg/kg). Molecular docking revealed that T03 formed six hydrogen bonds with PI3Kδ, exhibiting a high binding affinity of -11.6 kcal/mol.

The structural features of 1H pyrazolo [3, 4-d] pyrimidine derivatives through 3D-QSAR modelling established a relationship between functional groups and their biological activity. Based on the constructed 3D-QSAR model, five novel 1H pyrazolo [3, 4- d] pyrimidine derivatives targeting PI3Kδ were designed with improved predicted activities. Furthermore, molecular docking and MD simulations demonstrated stability and revealed ligand-receptor interactions. These novel potent inhibitors were assessed for their ADMET properties.

The 3D-QSAR model exhibits great reliability and predictive power through ADMET, molecular docking, and molecular dynamic approaches. These results would provide a valuable insight into lead optimization for new drug discovery.