Abstract
Natural evolution can be artificially accelerated and fine-tuned in a process called directed evolution (DE) to engineer biocatalysts with desired properties. Diversification, screening, and selection methods are user-defined, and the resultant engineered enzymes have been employed in a broad range of applications, from biotechnology to advancing understanding of molecular mechanisms. Historically, the focus of DE endeavors was on output, and comparatively little research has investigated, in detail, the mechanisms by which adaptive mutations modulate an enzyme’s desired function. As protein engineers have been repeatedly confounded by fitness plateaus, there is growing appreciation for the interplay between genotype and phenotype during adaptive walks on fitness landscapes. This has led not only to a diversification of methods to engineer better enzymes but has motivated the exploration of the molecular determinants that describe the link between primary sequence, enzyme function, and phenotypic fitness. In the first chapter of this thesis, I present work that tested and verified a new hypothesis that ancestrally reconstructed sequences serve as more evolvable starting points for directed evolution because they are more intrinsically phenotypically robust. The second chapter describes preliminary investigation into the mechanism of catalysis by an artificial gold biocatalyst. This work revealed that binding and catalysis are incredibly slow, but also highly dynamic, and that the enzyme is prone to oligomerization. Taken together, the work presented in this thesis outlines an exciting new method for design of more efficient directed evolution campaigns and underscores the value in fully characterizing biocatalysts to understand how they perform under selection pressure.