We can use the model structures to predict the roles that the mutations may play in NK function. As shown in Fig. 3b, many mutations were far away from the active TGF beta inhibitor pocket, and two mutations (V150 and T224) were close to the active site. The hydrophobic pocket of the active site

was broadened as a result of all of the amino acid substitutions (Fig. 3c), which may lead to changes in the protein structure and the catalytic activity. In the current study, we investigated how to improve the fibrinolytic activity of NK using directed evolution to broaden its medical or commercial applications. In vitro molecular evolution strategies are the most efficient methods for creating proteins with improved or novel properties. We generated a library of NK variants by the shuffling of genes encoding subtilisin NAT (NK), BPN′ and Carlsberg. To screen large libraries, the NK variants were expressed in E. coli. BL21(DE3)pLysS using a prokaryotic signal peptide, PelB, for efficient secretion. NK variants were selected based on zone-forming activity on agar plates with skim milk or fibrin. A mutant NK showed

a 2.3-fold increase in fibrinolytic activities compared Selleck HSP inhibitor to the wild-type NK from Bacillus natto. The further sequence and structural study of the mutant enzyme will offer some insight into the structure-function relationship of NK. The amino acid sequence alignment of the three parents and the mutant enzyme revealed that the catalytic triad and the substrate-binding site were conserved. Nine amino acid substitutions were derived from SB, and Baf-A1 cost the rest from SB or SC. No new mutations were introduced into the mutant enzyme sequence (Fig. 3a). To understand the functions of the amino acid substitutions, the identified

mutations in the selected mutant was distributed throughout the model of the mutant structure based on the three-dimensional model of NK that was previously constructed by our lab (Zheng et al., 2005). The three-dimensional structure showed that the strictly conserved residues of the catalytic centre (D32, H64, S221) and the substrate-binding sites (S125, L126, G127) were positioned in the pocket, which comprised two α-helices and seven β-strands (Fig. 3b). However, in the current study, none of the mutations was located in those strictly conserved regions throughout the mutant. Most of the mutations were located in the surface regions and far away from the pocket, with the exception of the substitutions A150V and T224S (Fig. 3b), which were very close to the Ser221 in the catalytic centre of the enzyme. This change may not be involved in hydrogen bonding with other residues. However, the combination of this change with other substitutions may result in the formation of a larger active-site pocket to improve the catalytic efficiency (Fig. 3c).