Engineering stabilized enzymes via computational design and immobilization

Includes bibliographical references.2016 Fall.The realm of biocatalysis has significantly matured beyond ancient fermentation techniques to accommodate the demand for modern day products. Enzymatically produced goods already influence our daily lives, from sweeteners and laundry detergent to blood pressure medication and antibiotics. Protein engineering has been a major driving force behind this biorevolution, yielding catalysts that can transform non-native substrates and withstand harsh industrial conditions. Although successful in many regards, computational design efforts are still limited by the crude approximations employed in searching a complex energy landscape. Advancements in protein engineering methods will be necessary to develop our understanding of biomolecules and accelerate the next generation of biotechnology applications. Our work employs a combination of computational design and simulation to achieve improved enzyme stability. In the first example, an enzyme used in the production of cellulosic biofuels was redesigned to remain active at high temperature. An initial approach involving consensus sequence analysis, predicted point mutation energy, and combinatorial optimization resulted in a sequence with reduced stability and activity. However, by using recombination methods and molecular dynamics simulations, we were able to identify specific mutations that had a stabilizing or destabilizing effect, and we successfully isolated mutations that benefited enzyme stability. Our iterative approach demonstrated how common design failures could be overcome by careful interpretation and suggested methods for improving future computational design efforts. In the second example, a cellulase was designed to have a high net charge via selected surface mutagenesis. “Supercharged” cellulases were experimentally characterized in various ionic liquids to assess the effect of high ion concentration on enzyme stability and activity. The designed enzymes also provided an opportunity to systematically probe the protein-solvent interface. Molecular dynamics simulations showed how ions influenced protein behavior by inducing minor unfolding events or by physically blocking the active site. Contradictory to previous reports, charged mutations only appeared to alter the affinity of anions and did not significantly change the binding of cations at the protein surface. Understanding the different modes of enzyme inactivation could motivate targeted design strategies for engineering protein resilience in ionic solvents. In addition to the discussed computational design methods, immobilization strategies were identified for capturing enzymes within porous protein crystals. Immobilization offers a generic approach for improving enzyme stability and activity. Our preliminary studies involving horseradish peroxidase and other enzymes suggested protein scaffolds could be employed as an effective immobilization material. Co-immobilizing multiple enzymes within the porous material led to improved product yield via exclusion of off-pathway reactions. Although future studies will be required to assess the potential capabilities of this immobilization strategy in comparison to other materials, preliminary results suggest protein crystals offer a favorable, controlled environment for immobilizing enzymes. The diversity of approaches presented in this thesis emphasizes that there are many options for engineering enzyme stability. Extending the lessons learned from our cellulase engineering to the greater field of rational protein design promotes the concept of biomolecules as designable entities. By establishing the shortcomings of our designs and suggesting routes for improvement, we anticipate our design methods and immobilization strategies will procure continued interest from the biotechnology community. The toolsets we developed for cellulases can be directly transferred to other enzymes and have the potential to impact a range of protein engineering applications