Directed evolution study of temperature adaptation in a psychrophilic enzyme.

We have used laboratory evolution methods to enhance the thermostability and activity of the psychrophilic protease subtilisin S41, with the goal of investigating the mechanisms by which this enzyme can adapt to different selection pressures. A combined strategy of random mutagenesis, saturation mutagenesis and in vitro recombination (DNA shuffling) was used to generate mutant libraries, which were screened to identify enzymes that acquired greater thermostability without sacrificing low-temperature activity. The half-life of seven-amino acid substitution variant 3-2G7 at 60 degrees C is approximately 500 times that of wild-type and far surpasses those of homologous mesophilic subtilisins. The dependence of half-life on calcium concentration indicates that enhanced calcium binding is largely responsible for the increased stability. The temperature optimum of the activity of 3-2G7 is shifted upward by approximately 10 degrees C. Unlike natural thermophilic enzymes, however, the activity of 3-2G7 at low temperatures was not compromised. The catalytic efficiency, k(cat)/K(M), was enhanced approximately threefold over a wide temperature range (10 to 60 degrees C). The activation energy for catalysis, determined by the temperature dependence of k(cat)/K(M) in the range 15 to 35 degrees C, is nearly identical to wild-type and close to half that of its highly similar mesophilic homolog, subtilisin SSII, indicating that the evolved S41 enzyme retained its psychrophilic character in spite of its dramatically increased thermostability. These results demonstrate that it is possible to increase activity at low temperatures and stability at high temperatures simultaneously. The fact that enzymes displaying both properties are not found in nature most likely reflects the effects of evolution, rather than any intrinsic physical-chemical limitations on proteins.

Subtilisin BPN' variants: increased hydrolytic activity on surface-bound substrates via decreased surface activity.

Site-directed mutagenesis and random mutagenesis were used to produce variants of subtilisin BPN' (Bacillus amyloliquefaciens) protease with variable surface adsorption properties. Protease adsorption and peptide hydrolysis rate were measured for these variants using a model substrate consisting of a peptide covalently bound to a surface. While most variants adsorb at a level very similar to that of native BPN', several variants were identified which adsorb either more or less. For surface-bound substrates we report a linear dependence between the concentration of adsorbed protease enzyme and substrate hydrolysis, similar to the linear dependence between enzyme solution concentration and hydrolysis of soluble substrates. On the basis of this knowledge we hypothesized that variants designed to adsorb at a higher level on a surface-bound peptide substrate would hydrolyze that surface-bound substrate faster. Contrary to our original expectations, the variants that adsorb more on the covalently bound peptide surface hydrolyze this substrate slower. In addition, variants of BPN' which adsorb at a lower level than native BPN' hydrolyze the surface-bound substrate faster. Enzyme adsorption and the subsequent peptide hydrolysis are altered by substituting amino acids that modify the surface charge or hydrophobicity of the native enzyme. This effect is most dramatic when the changes were made at surface-exposed sites around the binding pocket/active site of the enzyme. One mechanism that is consistent with the data is based on the relationship between the level of adsorption and the enzyme's affinity for the surface. In this mechanism weakly adsorbed enzymes are postulated to move more rapidly from site to site on the surface, thereby increasing substrate hydrolysis.

Enzyme behavior at surfaces. Site-specific variants of subtilisin BPN' with enhanced surface stability.

Enzyme adsorption and inactivation at the solid/liquid interface for subtilisin BPN' show a strong dependence on the nature of the solid surface. Adsorption of BPN' at the solid/liquid interface is considerably greater for a hydrophobic surface than for a hydrophilic one. Likewise, the rate of inactivation of the wild-type BPN' is over five times greater when equilibrated with a hydrophobic surface than with a hydrophilic surface. The rate data from these enzyme inactivation experiments performed at 50 degrees C are best fit by a second-order kinetic equation, suggesting a bimolecular pathway to inactivation. The role of increased surface adsorption on this bimolecular inactivation is discussed in terms of two different mechanisms. Several site-specific variants of subtilisin BPN' have been made in an attempt to alter the surface-inactivation of the wild-type enzyme. The extent of adsorption on the model surfaces is significantly lowered by certain lysine to phenylalanine changes in BPN'. Consequently, the surface autolytic stability shows a 4-fold improvement. The change in surface autolytic stability is achieved even though the basic kinetic parameters (kcat and KM) of the variant enzymes are not significantly different on a soluble substrate. The results provide insights into the use of mutagenesis to probe the mechanism of protein interactions with surfaces.