Probing The Relationship Between Structure And Function Of Nattokinase From Bacillus Natto

A three dimensional structural model of nattokinase (NK) from Bacillus natto was constructed by homology modeling. High resolution X-ray structures of Subtilisin BPN' (SB), Subtilisin Carlsberg (SC), Subtilisin E (SE) and Subtilisin Savinase (SS), four proteins with sequential, structural and functional homology were used as templates. Initial models of NK were built by MODELLER and analyzed by the PROCHECK programs. The best quality model was chosen for further refinement by constrained molecular dynamics simulations. The overall quality of the refined model was evaluated. The refined model NKCl was analyzed by different protein analysis programs including PROCHECK for the evaluation of Ramachandran plot quality, PROSA for testing interaction energies and WHATIF for the calculation of packing quality. This structure was found to be satisfactory and also stable at room temperature as demonstrated by a 300ps long unconstrained molecular dynamics (MD) simulation. Further docking analysis promoted the coming of a new nucleophilic catalytic mechanism for NK, which is induced by attacking of hydroxyl rich in catalytic environment and locating of S221.Hydrogen Bonds around the catalytic triad (D32, H64, S221) and the oxyanion hole (N155) are very important to catalysis of peptide bond hydrolysis in all proteases. For subtilisin nattokinase (NK), a bacterial serine protease, the three-dimensional structural model was analyzed to suggest that several hydrogen bonds from S33, D60, S62, and T220 stabilize the transition state of hydrolysis reaction. And four mutants ofS33A, D60A, S62A, and T220A were respectively constructed by site-directed mutagenesis to remove these hydrogen bonds. The results of enzyme kinetics indicate that removal of these hydrogen bonds will increase the transition state energy (AAGT). And these hydrogen bonds are more important for catalysis not substrate binding, because virtually all of the effects are on kca and not Km, in which the substrate Suc-AAPF-pNA was used. We also presented the results of free energy perturbation (FEP) calculations on binding and catalysis of another substrate, H-D-VLK-pNA, by native NK, and variant mutants (S33A, D60A, S62A, and T220A). The calculated difference in the free energy of catalysis and binding also suggests that these four residues are more important for catalysis than substrate binding. And these calculated values compared well with the experimental values in enzyme kinetics studies. The results of molecular dynamics (MD) simulations further demonstrate that removal of these hydrogen bonds will partly release D32, H64 and N155 so that the stabilization of transition state is destroyed.