Deep Sequencing-Based Protein Engineering to Optimize Functional Enzyme Expression in Synthetic Metabolic Pathways

Synthetic biologists primarily optimize enzyme expression levels in designed metabolic pathways by modifying transcription and translation. However, many enzymes can not be rescued by such methods as they are intrinsically unstable, thus requiring changes to the protein sequence itself. In my presentation I will detail different deep sequencing-based approaches to enable the rapid and parallel determination of sequence effects on protein function for complete gene-encoding sequences. The first approach, called Fluxscan, is able to resolve the sequence-flux relationship of an enzyme in a growth-based selection. I will show application of this method to determine the function of 93% of all possible single point mutants (over 8,000) in a levoglucosan utilization pathway implanted in Escherichia coli. I will describe how computational protein design augmented by these experimental datasets resulted in an designed metabolic pathway that supported 15-fold improvement in growth rate on levoglucosan as the sole carbon source, and describe how this advance improves hybid thermochemical/biochemical processing of renewable biomass. The main shortcoming of FluxScan is that the method relies on a coupling of enzyme activity to growth rate, yet many pathways cannot be so coupled. In the second portion of my talk I will describe an alternative deep sequencing-based tool to improve enzyme expression, for which I will supply datasets from several model enzyme systems.