论文部分内容阅读
Background: Previous results indicated that the CDK2/Cyclin E1 protein complex, which plays a key role in regulating the cell cycle, could be disrupted by two short peptides TAALS and LAALS.Selective alanine mutagenesis on CDK2 suggests that Y180 and K178 are two possible residues that the peptides target in the CDK2-Cyclin E1 interface.To design even more potent peptide inhibitors, the detailed binding mechanism needs to be understood clearly.Methods: For the flexible T-loop (amino acids 150-165) on CDK2, an ensemble of ten different conformations was generated by Rosetta for both the inactive CDK2 (PDB 1E1X) and the active CDK2 (PDB 1FIN), respectively.Peptide docking by AutoDock onto the ensemble of CDK2 conformations thus mimics a flexible docking.The resulting CDK2-peptide complex decoys were ranked by three different strategies according to frequency analysis, AutoDock binding energy, or a knowledge-based potential score.Moreover, MD simulations on both inactive and active CDK2, as well as 16 CDK2-peptide decoys were carried out.Results: MD simulations showed that TAALS could dock onto CDK2 between its T-loop and two key residues Y180 and K178.This stabled a particular T-loop conformation that prevented cyclin from binding to CDK2.Both TAALS and LAALS bind to the CDK2 similarly based on docking and MD simulations.These findings agree with previous experimental data that TAALS and LAALS are potent inhibitors.More interestingly, a failed peptide inhibitor RAALF, also known from the previous experinents, was found to move away from the above key pocket sites in our simulations.This observation offered a means to select better peptides.Indeed, based on the methodology described above, six peptides are designed to interrupt CDK2/Cyclin E 1 interface and four of them are found to dissociate the CDK2/Cyclin E1 complex experimentally by our collaborator.Moreover, two such designed peptides (DAALT and YAALQ) inhibit the kinase activity of Cyclin E/Cdk2.Conclusions: A combined computational approach including loop modeling, protein-peptide flexible docking, and MD simulation is used to investigate the binding mechanism of CDK2-TAALS and to design new peptides.Our main finding is that the potency of small peptide inhibitors that bind to the novel pocket of CDK2 largely depends on the dynamical nature of the protein/peptide interaction, particularly, whether the peptide could stabilize the intermediate state of the T-loop.Such effects are hard to identify by docking simulations only without sampling the T-loop conformations .