Search

The present study concerns the use of site-directed mutagenesis experiments to optimize a three-dimensional model of the neutral protease of Bacillus subtilis (NP-sub). An initial model of NP-sub was constructed using the crystal structures of the homologous neutral proteases of Bacillus thermoproteolyticus (thermolysin) and Bacillus cereus as templates. The largest portion of NP-sub could be modelled satisfactorily, using standard techniques, but several surface-located regions could only be modelled with a high degree of uncertainty. In order to make the model more reliable in these regions a 'model building by mutagenesis' approach was adopted. Mutations were designed such that their effect on thermal stability could indicate how their local environment should be modelled. This approach provided insight in the local structure of several regions in NP-sub that were hard to model on the basis of homology with the two known structures alone.

Using genetically engineered mutants of the neutral protease from Bacillus stearothermophilus (BsteNP), it had been shown that the surface-exposed structural motif constituted by Phe63 embedded in a four amino acid hydrophobic pocket is critical for the thermal stability of the thermophilic neutral proteases from Bacilli. To measure the stabilizing contribution of each hydrophobic interaction taking place between Phe63 and the hydrophobic pocket, we grafted this structural motif in the neutral protease from the mesophile Bacillus subtilis (BsubNP). This was accomplished by first creating the Thr63-->Phe mutant of BsubNP and then generating a series of mutants in which the four amino acids which in thermolysin surround Phe63 and form the hydrophobic pocket were added one after the other. By analysing the thermal stability of each mutant it was found that the 2 degrees C destabilizing effect of the Thr63-->Phe substitution was completely suppressed by the addition of the four amino acid hydrophobic pocket, each replacement providing a stabilizing contribution of approximately 0.8-1 degrees C. These results are discussed in the light of the peculiar mechanism of thermal inactivation of proteolytic enzymes.

Using genetic techniques the contribution of surface loops to the thermal stability of Bacillus subtilis neutral protease (NP-sub) was studied. Mutations were designed to make the surface of NP-sub more similar to the surface of more thermostable neutral proteases such as thermolysin (TLN). The mutations included the replacement of an irregular loop by a shorter variant and the introduction of a ten-residue beta-hairpin. In general, these drastic mutations had little effect on the production and activity of NP-sub, indicating the feasibility of major structural rearrangements at the surface of proteins. In the most stable mutant, exhibiting an increase in thermal stability of 1.1 degree C, approximately 10% of the surface of NP-sub was modified. Several NP-sub variants carrying multiple mutations were constructed. Non-additive effects on thermal stability were observed, which were interpreted on the basis of a model for thermal inactivation, that emphasizes the importance of local unfolding processes for thermal stability.

Oligonucleotide-directed mutagenesis has been used to replace glycine residues by alanine in neutral protease from Bacillus subtilis. One Gly to Ala substitution (G147A) was located in a helical region of the protein, while the other (G189A) was in a loop. The effects of mutational substitutions on the functional, conformational and stability properties of the enzyme have been investigated using enzymatic assays and spectroscopic measurements. Single substitutions of both Gly147 and Gly189 with Ala residues affect the enzyme kinetic properties using synthetic peptides as substrates. When Gly replacements were concurrently introduced at both positions, the kinetic characteristics of the double mutant were roughly intermediate between those of the two single mutants, and similar to those of the wild-type protease. Both mutants G147A and G189A were found to be more stable towards irreversible thermal inactivation/unfolding than the wild-type species. Moreover, the stabilizing effect of the Gly to Ala substitution was roughly additive in the double mutant G147A/G189A, which shows a 3.2 degrees C increase in Tm with respect to the wild-type protein. These findings indicate that the Gly to Ala substitution can be used as a strategy to stabilize globular proteins. The possible mechanisms of protein stabilization are also discussed.