Identification of potential PI3Kγ inhibitors among FDA approved drugs using integrated computational and quantum chemical approaches

Phosphoinositide 3-kinase gamma (PI3Kγ) is recognized as a promising target for inhibiting the activity of cancer cells, leading to increased efforts to develop PI3Kγ inhibitors as potential cancer therapies. In this research, Quantitative Structure-Activity Relationship (QSAR) modeling was performed using CORAL software on a dataset comprising 243 PI3Kγ inhibitor compounds. The model enabled the prediction of the endpoint (pIC₅₀) for different populations of chemical compounds targeting the PI3Kγ receptor. It demonstrated high predictive accuracy for split 1 and TF3, achieving R²_val = 0.8003, Q² = 0.7807, and MAE = 0.3410, using the optimized parameters WIIC = 0.2 and WCII = 0.5. A screening of FDA-approved drugs, which were not part of our dataset, identified 11 potential candidates as PI3Kγ receptor ligands (predicted pIC₅₀ ≥ 9). A literature review revealed that 8 of these compounds had previously been evaluated for PI3K receptor interactions. Among them, Epirubicin, Doxorubicin, and Daunorubicin; all anthracycline chemotherapy agents; are known to inhibit cancer progression by slowing or halting cell growth. Molecular dynamics simulations further validated the QSAR screening results, confirming that Epirubicin forms the most stable and tightly bound complex with PI3Kγ, with ΔG_bind values of -3.81 kcal/mol for PI3Kγ-M192 and -4.47 kcal/mol for PI3Kγ-Epirubicin. These findings indicate that the QSAR model could function as a targeted screening tool, facilitating the identification of drugs with strong affinity for the PI3Kγ receptor, and ultimately improving the hit rate compared to random screening approaches.