Identification Of CDK 9 Inhibitors Based On Virtual Screening And Biological Evaluation

Objective: Cancer and many proliferative diseases occur due to changes in the cell cycle.The cell cycle is controlled by the Cyclin-dependent kinases(Cyclin-dependent kinases,CDKs)family.The CDKs family binds to cyclins to activate cyclins.The activation and regulation of CDKs promote cell division,change the transition state and enter the mitotic stage.In addition to cell cycle control,some CDKs are also involved in the control of the transcription process.CDKs play different roles.For example,CDK 2 plays an important role in apoptosis,CDK 5 plays a role in neuronal cells,and CDK 7,CDK 8,and CDK 9 play a role in transcription.Cyclin-dependent kinase 9(Cyclin-dependent kinase 9,CDK 9),as a cancer treatment target,has attracted widespread attention.Inhibition of CDK 9 selectively reduces the up-regulated proteins in many cancers,thereby inhibiting proliferation and promoting apoptosis.It is one of the effective ways to treat cancer.Here,we aim to obtain new CDK 9 inhibitors by using a cross-over virtual screening strategy.Methods: Computer Aided Drug Design(CADD)is a method of designing and optimizing lead compounds based on computational chemistry,through computer simulation,calculation and budgeting of the relationship between drugs and receptor biomacromolecules.In this study,CADD was used to construct a cross-cutting virtual screening strategy to obtain new CDK 9 inhibitors.First,according to the Lipinski’s Rule of Five and ADMET(adsorption,distribution,metabolism,excretion and toxicity)characteristics of the preliminary screening of the compounds,the drug-like molecules have been obtained.Secondly,a 3D-QSAR pharmacophore model was constructed with the help of known CDK 9 inhibitors through the Hypo Gen algorithm,and the reliability of the pharmacophore was verified using the test set and Fischer randomization technology.The obtained pharmacophore model is used for large-scale screening to obtain the seed molecules.In addition,molecular docking studies are performed on the screened sprout molecules to obtain hit molecules with high computational simulation accuracy.Then,molecular dynamics(MD)simulations were performed on the complex of the hit molecule and CDK 9 to study the stability and binding mode of the hit molecule in the CDK 9 protein conformation.Finally,the CDK 9 kinase inhibitory activity test was performed on the obtained hit molecules by biochemical means to verify the reliability of the cross-virtual screening strategy designed in this study.Results:(1)We established a pharmacophore model Hypo 1 with 4 interaction characteristics through 3D-QSAR.According to the statistical parameters of Hypo 1,we concluded that Hypo 1 has the lowest total cost compared to other hypothetical models.And it is far from zero cost(351.435 bits)and fixed cost(75.5308 bits).This shows that Hypo 1 has a certain degree of reliability and can meet the requirements of coarse screens.(2)Based on the established pharmacophore,we designed a multi-level screening strategy.First,357,231 small molecules in the commercial database were pre-screened using the Lipinski’s Rule of Five and ADMET characteristics,and158,694 small molecules were obtained.(3)Use the constructed pharmacophore Hypo1 for coarse screening to obtain small molecules with a Fit value greater than the positive drug,and take the top 10% of the small molecules(16,120)to continue the screening.(4)Using the molecular docking program Libdock and CDOCKER combination for precise screening,7 hit compounds with docking scores greater than positive drugs and ranking the top 10% were retained.(5)In the 20 ns molecular dynamics simulation,except for the larger root mean square deviation(RMSD)amplitude of the Hit 5 and CDK 9 systems,the other 6 Hits systems all tended to be stable(wave range within 1 ?).During the entire kinetic simulation process,all Hits formed stable hydrogen bond interactions with CDK 9.The lowest potential energy value of all systems is low,centered at-119300 kj/mol.The MM/PBSA free binding energy of 6 Hits is stable,slightly lower than DRB,which indicates that all systems remain stable throughout the simulation process.(6)In addition,these 6 Hits have strong inhibitory activity on CDK 9 kinase.In particular,hit 3 showed the strongest inhibitory activity,with an inhibitory rate of 71% at 1 μmol/L.Conclusion: The multi-level virtual screening strategy designed in this research has successfully obtained 6 hit compounds that specifically inhibit CDK 9,which provides important reference value for the design and synthesis of new CDK 9 inhibitors.