In Vitro Ultra-High Throughput Screens for Directed Evolution

Biocatalyst development is an important area of interest in synthetic biology. The complexity of enzyme structure strongly impedes our ability to design enzymes de novo, forcing biocatalyst development to build from natural, functional enzymes through enzyme engineering. The evolvability of these natural enzymes can be exploited to improve or alter enzyme function without any understanding of structure-function relationships through a process called directed evolution. Directed evolution mimics natural evolution; iterative rounds of mutagenesis and screening or selection yield mutants with the desired activity.;New screening formats that minimize sample size and increase analysis rate (throughput) greatly reduce the cost and time burden of directed evolution. However, to date each of these formats lack broad utility. Here, a potentially broad utility ultra-high throughput screening platform was developed that uses fluorescence activated cell sorting (FACS), in vitro compartmentalization (IVC) in emulsion microdroplets, a previously unexploited fluorogenic natural product glycosylation reaction, and an extensive, modular coupled enzyme biosensor system enabling detection of a variety of metabolites and cofactors.;In Chapter 2, an IVC-FACS screen was developed based on a natural reaction catalyzed by a natural product glycosyltransferase UGT72B1: the selective glucosylation of esculetin to yield esculin. Prior to this advance, screening platforms based on IVC-FACS suffered from poor protein expression yields. Here, this problem is resolved using a strategy of emulsion compartmentalization of E. coli cells followed by cell lysis. Robust biocatalyst expression permits exploitation of otherwise sluggish fluorogenic enzymes, such as UGT72B1. The crucial genotype-phenotype link necessary for directed evolution was verified by enriching model populations of UGT72B1 and an inactive mutant. Confirmation of enrichment provided strong evidence for a fully functional directed evolution screen.;In Chapter 3, the UGT72B1 catalyzed fluorogenic reaction was coupled to several upstream enzymatic reactions to produce coupled enzyme biosensors (CEBs) capable of detecting several natural biomolecules. Endpoint fluorescence correlated with analyte concentration in both the IVC-FACS lysate format as well as in microtiter plates. This CEB system is highly modular and engineerable, providing the potential for customized detection of nearly any metabolite of interest.;In Chapter 4, mathematical models were derived which can be used to predict optimal mutation rates in error-prone polymerase chain reaction (epPCR) for directed evolution. Modeling of a process called neutral drift, which can be performed with our CEB-IVC-FACS screens or other ultra-high throughput screens, including a previously published universal FACS screen for protein folding and stability, improves the guidelines for epPCR library mutation rates for neutral drift. The results of these models provide strong support for neutral drift as an advantageous preliminary step in all directed evolution experiments and suggest improved library quality compared to other prominent library quality enhancing techniques.;In Chapter 5 a mutant library of UGT72B1 was generated by epPCR and was subjected to neutral drift using the IVC-FACS screen from Chapter 2. Subsequent analysis of the enriched library by FACS failed to indicate the presence of a mutant displaying activity with non-native donor substrates. Further efforts to identify mutants with improved catalytic properties from the enriched library by screening mutants in 96-well microtiter plates is described.;In Chapter 6, the potential for a CEB-IVC-FACS screen to be used for engineering polyketide synthase (PKS) activities is discussed. The work of Dr. Irina Koryakina to better understand PKS specificities toward non-natural substrates has helped place the Williams lab at the forefront of PKS engineering, but the lack of a suitable screen hinders efforts to apply this knowledge toward altering PKS activities.;Progress in the field of synthetic biology is highly dependent on biocatalyst engineering and is therefore hindered or enhanced according to advances in directed evolution. The CEBIVC- FACS screening format has great potential as a broadly applicable directed evolution tool, and the mathematical models presented in this work enhance our understanding of strategic choices in directed evolution library preparation.