Discovery Of Novel Nonpeptide Allosteric Inhibitors By Virtual Screening And Bioassays

Background: Cell cycle is the series of events that take place in a cell leading to its division and duplication(replication) that produces two daughter cells, which can be divided into two periods: interphase and the mitotic(M) phase. Cell cycle is a crucial part of the cell life activities, and proliferation, differentiation, senescence and apoptosis are also cell cycle-dependent. Much of the research has examined that deregulation of cell cycle and inactivation of apoptosis combined to result in the malignant transformation and tumor development. Therefore it has been taken the attitude that cancer is a cell cycle disease and becomes one of the the most fatal diseases threatening the health of characters worldwide. Research of anti-cancer targeted drugs has captured an army of scientists’ attention. The main characters of cancer cells include the regulation of dysfunction over apoptosis and the property of aberrant migration. In mammalian, cell cycle is governed by the sequential activation of Ser/Thr-specific cyclin-dependent kinases(CDKs), thereby driving the orderly transition of each phase and being core of cell cycle regulation. CDK2 is one of the members of CDK family, which has been studied rather deeply so far. Cyclin E combines with CDK2 to stimulate cell cycle by phosphorylating a number of substrates. On entry from G1 into S phase, cyclinE is abruptly destroyed by the proteasome to which it is targeted by ubiquitination, cyclin A then assoeiates with CDK2 to drive cells through S phase[1]. Abnormal upregulation of CDK2 activity is the major biological characteristic in tumor cells, and CDK2 therefore has been proven to be an important target for cancer therapy. With continual understanding of CDK2 crystal structure, increasing ATP-competitive inhibitors have been reported and viable drug candidates have been identified from clinical trials. However, these ATP-competitive inhibitors are faced with a lack of target selectivity[2]. Allostery, also referred to as allosteric regulation, is characterized by the regulation of the functional activity of a protein via the binding of an effector to a site distinct from the active site, defined as the allosteric site. Allosteric modulators target to allosteric sites, offering distinct advantages compared to traditional medicine(orthosteric ligands) that target to active sites, such as greater specificity, reduced side effects, and lower toxicity[3-9]. Allosteric modulators have therefore drawn increasing attention as potential therapeutic drugs in the design, optimization and development of new drugs.Objective: Based on the crystal structure of p-CDK2-cyclin A3(PDB code:2CCI), our project aims at attaining inhibitors as CDK2 novel lead structures targeting a noncatalytic pocket near the interface of the CDK2/cyclin A3 complex and inhibiting its kinase activity effectively.Methods: Our project is based on the crystal structure of p-CDK2-cyclin A3(PDB code:2CCI), combined with molecular docking, virtual screening and measurement of kinase activity in vitro. Besides, we also validated the binding modes of I4/CDK2 and I9/CDK2 and the noncompetitive inhibition of these two compounds. In vivo, we examined their anti-tumor activity. To further examine whether I4 and I9 could interrupt the interaction between CDK2 and cyclin A3, we employed a GST pull-down assay.Results:(1) We finally got I4(IC50 = 13.59 μM) and I9(IC50 = 52.12 μM) as the lead structures exhibiting superior CDK2/cyclin A3 inhibition activity;(2) The IC50 value of I4 and I9 did not change significantly under different concentrations of ATP or substrate peptide and still showed promising CDK2/cyclin A3 inhibition activity;(3) Based on the binding mode and site-directed mutagenesis kit, the IC50 values of I4 against p-R150A-cyclin A3(IC50 = 330.2 μM) and p-Y180A-cyclin A3(IC50 = 316.1 μM) were determined and the inhibition activity decreased significantly; The IC50 values of I9 against p-R150A-cyclin A3(IC50 = 884.7 μM) and p-Y180A-cyclin A3(IC50 = 151.4 μM) were also determined and the inhibition activity decreased significantly compared with wild type(IC50 = 52.12 μM);(4) I4 and I9 could both exhibit good antiproliferative activities against a broad spectrum of human cancer cell lines, A549, HepG2 and MDA-MB-231;(5) At 100 μM, I4 could effectively arrest the cell cycle of HepG2 at S phase and the percentage of HepG2 cells increased from 17.67% to 28.25%. When the concentration reached 200 μM, HepG2 cells entering G2 phase decreasd; At low concentrations(10-50 μM), I9 could weakly arrest the cell cycle of A549 at S phase. However, when the concentration reached 200 μM, I9 effectively arrested the cell cycle at S phase and the percentage of A549 cells increased from 11.53% to 41.30%;(6) The pull-down of cyclin A3 was examined by SDS-PAGE. GST-CDK2 bound to His-cyclin A3, indicating that the GST-CDK2 fusion protein can interact directly with cyclin A3 in solution. Once I4 and I9 was added, pull-down of cyclin A3 was significantly increased and exhibited a concentration-dependent trend. These results laid the foundation for the development of CDK2 allosteric anticancer drugs.Conclusions:(1) In this paper, we used molecular docking, the structure-based virtual screening, and biological activity tests to screen two lead structures with potent CDK2/cyclin A3 inhibition activity, I4(IC50 = 13.59 μM) and I9(IC50 = 52.12 μM);(2) The IC50 value of I4 and I9 did not change significantly under different concentrations of ATP or substrate peptide and still showed promising CDK2/cyclin A3 inhibition activity, which confirmed that maybe I4 is a noncompetitive inhibitor;(3) Based on the binding model and site-directed mutagenesis kit, we initially verified that I4 and I9 binds firmly into a pocket near the interface of the CDK2/cyclin A3 complex, which composed of Arg126, Arg150, Arg169 and Tyr180;(4) I4 and I9 exhibited concentration-dependent antiproliferative activities against a broad spectrum of human cancer cell lines, A549, HepG2 and MDA-MB-231, and I4 showed superior inhibition activity;(5) The cell cycle was blocked at S phase by compound I4 and I9;(6) I4 and I9 could both certainly interrupt the interaction between CDK2 and cyclin A3 and exhibited a concentration-dependent trend.