Identification of potential ATP-competitive cyclin-dependent kinase 1 inhibitors: De novo drug generation, molecular docking, and molecular dynamics simulation. (March 2023)

• Record Type: Journal Article
• Authors: He, Fengming; Wang, Xiumei; Wu, Qiaoqiong; Liu, Shunzhi; Cao, Yin; Guo, Xiaodan; Yin, Sihang; Yin, Na; Li, Baicun; Fang, Meijuan
• Is Part Of: Computers in biology and medicine. Volume 155(2023)
• Publication Date: 2023-03
• Subjects: De novo drug generation -- Cyclin-dependent kinases 1 -- Rational drug design -- Molecular dynamics simulation -- Binding pose metadynamics -- ATP-Competitive inhibitors; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2023.106645
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Cyclin-dependent kinases 1 (CDK1) has been identified as a potential target for the search for new antitumor drugs. However, no clinically effective CDK1 inhibitors are now available for cancer treatment. Therefore, this study aimed to offer potential CDK1 inhibitors using de novo drug generation, molecular docking, and molecular dynamics (MD) simulation studies. We first utilized the BREED algorithm (a de novo drug generation approach) to produce a novel library of small molecules targeting CDK1. To initially obtain novel potential CDK1 inhibitors with favorable physicochemical properties and excellent druggability, we performed a virtual rule-based rational drug screening on our generated library and found ten initial hits. Then, the molecular interactions and dynamic stability of these ten initial hits and CDK1 complexes during their all-atom MD simulations (total 18 μs) and binding pose metadynamics simulations were investigated, resulting in five final hits. Furthermore, another MD simulation (total 2.1 μs) with different force fields demonstrated the binding ability of the five hits to CDK1. It was found that these five hits, CBMA001 (ΔG = −29.88 kcal/mol), CBMA002 (ΔG = −34.89 kcal/mol), CBMA004 (ΔG = −32.47 kcal/mol), CBMA007 (ΔG = −31.16 kcal/mol), and CBMA008 (ΔG = −34.78 kcal/mol) possessed much greater binding affinity to CDK1 than positive compound Flavopiridol (FLP, ΔG = −25.38 kcal/mol). Finally, CBMA002 and CBMA004 were identified as excellent selective CDK1 inhibitors in silico . Together, this study provides a workflow for rational drug design and two promising selective CDK1 inhibitors that deserve further investigation. Graphical abstract: Illustration of the identification of potential ATP-competitive Cyclin-dependent kinase 1 inhibitors using de novo drug generation, molecular docking, and molecular dynamics simulation. Image 1 Highlights: BREED algorithm is utilized to produce novel small molecules potentially targeting CDK1. Ten initial hits are obtained using a virtual rule-based rational drug screening strategy. MD and BPMD simulations identify five potential CDK1 ATP-competitive inhibitors. CBMA002 and CBMA004 exhibit the excellent selective for CDK1 in silico.