Discovery of novel selective PI3Kγ inhibitors through combining machine learning-based virtual screening with multiple protein structures and bio-evaluation

Abstract

Introduction: Phosphoinositide 3-kinase gamma (PI3Kγ) has been regarded as a promising drug target for the treatment of various diseases, and the diverse physiological roles of class I PI3K isoforms (α, β, δ, and γ) highlight the importance of isoform selectivity in the development of PI3Kγ inhibitors. However, the high structural conservation among the PI3K family makes it a big challenge to develop selective PI3Kγ inhibitors.

Objectives: A novel machine learning-based virtual screening with multiple PI3Kγ protein structures was developed to discover novel PI3Kγ inhibitors.

Methods: A large chemical database was screened using the virtual screening model, the top-ranked compounds were then subjected to a series of bio-evaluations, which led to the discovery of JN-KI3. The selective inhibition mechanism of JN-KI3 against PI3Kγ was uncovered by a theoretical study.

Results: 49 hits were identified through virtual screening, and the cell-free enzymatic studies found that JN-KI3 selectively inhibited PI3Kγ at a concentration as low as 3,873 nM but had no inhibitory effect on Class IA PI3Ks, leading to the selective cytotoxicity on hematologic cancer cells. Meanwhile, JN-KI3 potently blocked the PI3K signaling, finally led to distinct apoptosis of hematologic cell lines at a low concentration. Lastly, the key residues of PI3Kγ and the structural characteristics of JN-KI3, which both would influence γ isoform-selective inhibition, were highlighted by systematic theoretical studies.

Conclusion: The developed virtual screening model strongly manifests the robustness to find novel PI3Kγ inhibitors. JN-KI3 displays a specific cytotoxicity on hematologic tumor cells, and significantly promotes apoptosis associated with the inhibition of the PI3K signaling, which depicts PI3Kγ as a potential target for the hematologic tumor therapy. The theoretical results reveal that those key residues interacting with JN-KI3 are less common compared to most of the reported PI3Kγ inhibitors, indicating that JN-KI3 has novel structural characteristics as a selective PIK3γ inhibitor.

Keywords: ADMET, absorption, distribution, metabolism, excretion, and toxicity; AKT, protein kinase B; AUC, area under receiver operations characteristic curve; Badapple, bioactivity data associative promiscuity pattern learning engine; CADD, computer-aided drug design; CDRA, confirmatory dose–response assays; DMEM, Dulbecco’s Modified Eagle Medium; DS3.5, discovery studio 3.5; FBS, fetal bovine serum; GPCR, G protein-coupled receptors; H-bond, hydrogen bond; Hematologic malignancies; IMDM, Iscove’s Modified Dulbecco’s Medium; Ionic, ionic interactions; JN-KI3; MD, molecular dynamics; MM/GBSA, molecular mechanics/generalized born surface area; Molecular dynamics simulation; NBC, naive Bayesian classifier; PAGE, polyacrylamide gel electrophoresis; PAINS, pan-assay interference compounds; PARP, poly ADP-ribose polymerase; PDB, protein data bank; PI3K, Phosphoinositide 3-kinase; PI3Kγ; PSA, primary screening assays; REOS, rapid elimination of swill; RMSD, root-mean-squared-deviation; RMSF, root-mean-squared-fluctuation; ROC, receiver operations characteristic; RTK, receptor tyrosine kinases; SD, standard deviation; SMILES, simplified molecular input line entry specification; SP, standard precision; Selective inhibitor; VS, virtual screening; Virtual screening; Water Bridge, hydrogen bonds through water molecular bridge; XP, extra precision.

Graphical abstract

Fig. 1

The workflow of this study.

Fig. 2

(A-J) Distributions of the docking scores of the inhibitor (red)/noninhibitor (green) for each PI3Kγ protein with the best discrimination power; (K) the ROC curve of the naive Bayesian classifier based on every single docking score; (L) the ROC curve of the naive Bayesian classifier based on the combined docking scores.

Fig. 3

(A) The cell-free PI3Kγ inhibitory activities of 49 hits and two positive inhibitors, IPI-145 and LY294002; (B) the anti-proliferation effects towards multiple tumor cells of 49 hits and two positive inhibitors, IPI-145 and LY294002; (C) the anti-proliferation effects towards multiple tumor cells of JN-KI3 and IPI-145; (D) the anti-proliferation effects towards multiple malignant tumor cells of JN-KI3 with the gradient of concentrations.

Fig. 4

JN-KI3 inhibited PI3K/AKT signaling pathway and induced apoptosis of (A) K562 and (B) RPMI-8226 cell lines at different concentrations from 0 to 20 μM after 24 h incubation. JN-KI3 inhibited PI3K/AKT signaling pathway and induced apoptosis of (C) K562 and (D) RPMI-8226 cell lines at different times from 0 to 24 h for 20 μM concentration. JN-KI3 induced apoptosis of hematologic malignancies illustrating by flow cytometry. Apoptosis of (E) K562 and (F) RPMI-8226 cell lines treated with different concentrations of JN-KI3 (from 0 μM to 10 μM) for 24 h.

Fig. 5

2D interaction patterns and protein–ligand occupancy histogram of (A) PI3Kα/JN-KI3; (B) PI3Kβ/JN-KI3; (C) PI3Kδ/JN-KI3; (D) PI3Kγ/JN-KI3 complexes.

Fig. 6

(A) the 2D-structure labeled with atomic numbers; (B) the ligand RMSF value of JN-KI3; (C) the 3D-interaction diagrams between JN-KI3 and PI3Kγ before MD simulation (H-bond colored in green); (D) the 3D-interaction diagrams between JN-KI3 and PI3Kγ after MD simulation (H-bond colored in green).