In Silico Discovery of a Potential PI3Kα Inhibitor Incorporating an Imidazole Moiety for Breast Cancer Treatment

Abstract

Excessive activation of PI3Kα is prevalent in breast cancer, aberrantly activating the PI3K/Akt/mTOR pathway to drive tumor progression, thus rendering PI3Kα a vital therapeutic target. Two PI3Kα inhibitors (alpelisib and inavolisib) have been approved for clinical application. Inavolisib exerts a mutation-selective degradation mechanism and functions by directly eliminating aberrant proteins. To develop novel PI3Kα inhibitors, we employed multiple computational strategies including scaffold hopping based on the structure of inavolisib, -CDOCKER docking, Lipinski’s rule of five, ADMET prediction, molecular dynamics (MD) simulation and GLORYx metabolic prediction. These tools were used to assess candidate compounds for their predicted binding modes and binding affinity with PI3Kα, physicochemical profiles, pharmacokinetic properties, metabolic pathways and generated metabolites. Ultimately, M-288, a computationally prioritized candidate compound and imidazole derivative with favorable predicted properties, was identified. Its alanine moiety formed a H-bond with the key residue Gln859 in the binding pocket, and the oxygen atom on its carboxyl group established a salt bridge with non-conserved Arg770 to enhance binding affinity. M-288 also formed H-bonds with Asp933, Tyr836 and Val851, facilitating its stable binding to PI3Kα. This compound was predicted to occupy the ATP-binding pocket. Compared with other screened candidates, it exhibited low predicted toxicity and no inhibitory activity against major hepatic CYP450 isoforms, while satisfying drug-likeness criteria and possessing favorable hepatic metabolic clearance. Molecular dynamics simulations verified its stable binding profile. This study indicates that M-288 serves as a potential PI3Kα inhibitor and provides valuable references for subsequent discovery of PI3Kα inhibitors.