Artificial Selection Device to Expedite the Evolution of Metabolite-producing Microbes

Abstract

Although great progress has been made in rational metabolic engineering, strain improvement still requires the optimization of the biological system itself. An extension of directed evolution strategies to genome-wide perturbation increases the chance of obtaining improved phenotype, however, the combinatorial approach have been limited by the method isolating high metabolite-producing strains from among the library.

Among the various methods available currently, selection by the physiology associated with the growth is known to be the most efficient since high producer can be easily enriched simply by a serial culture. However, most physiological features to be improved such as biochemical productions are not easily associated with the growth and consequently, tedious screening process should be only technique to select high producer out of the library.

In this dissertation, to expedite the evolution of metabolite-producing microbes, artificial selection devices were developed by combining a riboswitch and a selection module that selectively sense inconspicuous metabolites. Using L-lysine-producing Escherichia coli as a model system, it was demonstrated that this RNA device could enrich pathway-optimized strains to up to 75% of the total population after four rounds of enrichment cycles. When used in conjunction with combinatorial mutagenesis for metabolite overproduction, our synthetic RNA device should facilitate strain improvement.