Item

Abstract

Bacillus licheniformis a-amylase (BLA) is a highly thermostable enzyme which is widely used in biotechnological processes. Although it is produced by a non-thermophilic bacterium, it remains active for several hours at temperatures over 90degreesC under conditions of industrial starch hydrolysis. It is also far more thermostable than the alpha-amylases from B. stearothermophilus and B. amyloliquefaciens despite the strong sequence similarities between these three proteins. BLA provides therefore an interesting model for protein engineers investigating enzyme thermostability and thermostabilization. Over the last decade, we have performed an extensive mutational and structural analysis on BLA in order to elucidate the origin of its unusual thermal properties and, if possible, increase its thermostability even further. Before the three-dimensional structure was known, we had used "blind" mutagenesis and identified two critical positions where amino-acid substitutions could either increase or decrease significantly the rate of irreversible thermoinactivation. Once a detailed X- ray structure of BLA was solved, structure-based mutagenesis was used to probe the role of residues involved in salt- bridges, calcium-binding or potential deamidation processes. Our results revealed the key role of domain B and its interface with domain A in determining the overall thermostability of BLA. Most of the mutations we introduced in this region modify the stability in one way or another by influencing the network of electrostatic interactions entrapping a Ca-Na-Ca metal triad at the domain A/B interface. In the course of this mutational study we have constructed over 500 BLA variants bearing single or multiple mutations, among which many were found to be either highly detrimental or slightly beneficial to the stability. The cumulative effect of the mutations enabled us to modulate the enzyme stability over a 50degreesC temperature range without perturbing significantly the amylolytic function. Although a full understanding of the origin of BLA natural thermoresistance has not yet been reached, our study demonstrated that it is not optimized and that it can be increased or decreased artificially by several means.