Molecular modeling and analysis of the 3D pharmacophore structure of the selective PI3K-α inhibitors as antitumor agents

Abstract

The importance of the activation of phosphatidylinositol-3-kinase (PI3K) in the processes of tumor cell formation and growth has brought great interest in the development of new antitumor drugs that inhibit the PI3K-Akt signaling pathway. Recent studies have shown that in most tumors the activity of p110α isoform of PI3K kinase has been altered, and today special emphasis is placed on developing specific PI3K-α inhibitors. In a series of 92 PI3K-α inhibitors, whose data on experimentally determined inhibitory activity were collected from literature, 3D quantitative structure-activity relationship studies (3D-QSAR) were carried out. All molecular structures were previously optimized using semiempirical PM3 and ab initio Hartree-Fock / 3- 21G methods, and modeling was performed using PLS regression analysis. Calculated parameters of internal (R2=0.84; Q2=0.67) and external (R2pred=0.681; r2m=0.594; Δr2m=0.00039)validation indicate on good reliability and predictive power of created 3D-QSAR model. The analysis of the variables enabled pharmacophore structure determination, which implies: presence of hydrogen bond donor and hydrogen bond acceptor at a distance of 18-18.4Å or 12- 12.4Å, presence of hydrophobic domain and hydrogen bond donor at a distance of 15.2-15.6Å, and presence of steric hotspot at optimal distance from hydrogen bond donor and acceptor. These results will have the relevance in selecting the lead compounds for targeted structural modifications in order to design new selective PI3K-α inhibitors as potential antitumor drugs.

Keywords

• 3d-qsar
• pharmacophore
• pi3k-α
• inhibitor
• tumor