Exploring divergent enzyme evolution using droplet-based microfluidics

Project Summary
Almost all biochemical processes are controlled and accelerated by enzymes. Enzymes are among the most proficient catalysts known with regard to rate accelerations and specificity. This proficiency is the result of finely tuned networks of amino acid residues that emerged over millennia of Darwinian evolution and work in concert to promote challenging biochemical processes.
Directed evolution is a technique that mimics Darwinian evolution in the laboratory, and allows optimization of protein properties even in the absence of precise molecular or mechanistic understanding. This technology is therefore particularly useful for engineering enzymes, as sequence-to-function relationships of enzymes are complex, and our current quantitative understanding of enzyme catalysis is still rudimentary. Directed evolution is an iterative process composed of three major steps: 1) mutagenesis 2) identification of mutants with improved traits 3) isolation of desired mutants. It is therefore essentially an experimental random search algorithm in protein sequence space. Protein sequence space is however inconceivably large - e.g. the combinatorial diversity of a small 100 amino acid long protein is > 10130. For comparison the estimated number of atoms in the known universe is ca. 1082. Therefore, to efficiently optimise an enzyme by directed evolution, maximising the number of analysed mutants (throughput) is essential.
This project aims at implementing lab-on-a-chip technologies to drastically increase throughput of directed evolution experiments. Specifically, we will perform single cell experiments using droplet-based microfluidics, where picolitre sized droplets serve as reaction vessels and provide a physical link between genotype and phenotype. Importantly, droplets can be sorted according to a fluorescence readout at speeds of around 103 droplets per second, to allow evaluation of ca. 106-107 enzyme variants per day. Here, fluorogenic enzyme substrates, which produce fluorescent products allow monitoring of reaction progress. In comparison, traditional plate-based assay formats typically have a throughput of 103-104 mutants per day.
These ultra-high throughput assays will be used to explore the evolvability computationally designed enzymes promoting valuable chemical transformations that are unknown in nature. In particular, we will evolve a common starting enzyme towards two mechanistically unrelated reactions, mimicking natural divergent evolution. This will lead to new enzymes, whose activity will be far above that of most state-of-the-art artificial enzymes and approach that of natural enzymes. In depth analysis of the divergent evolutionary trajectories will reveal new catalytic motifs and show how these motifs emerged during directed evolution.
More importantly, analysis of the trajectories will also uncover how mutations that are far apart the active site contribute to enzyme activity. This point is of particular importance as current enzyme design methodologies are focused on active site design alone. Overall, it is anticipated that we will gain a better understanding of enzyme catalysis and enzyme evolvability through this project, which will be ultimately important for design and engineering of the next generation of artificial enzymes.