Novel approaches for metabolic engineering of yeast at multiple scales

Living systems contain enormous potential to solve many pressing engineering problems, including the production of usable energy, the synthesis and degradation of a variety of materials, and the treatment of disease. Metabolic engineering, as one approach to harness this potential, treats the behavior of a living system as the combined product of multiple interacting modules, each of which can be tuned to maximize performance. However, the scarcity of techniques for predictive or high-throughput engineering design of these modules, especially in eukaryotes, contributes to long strain development times and high research cost. In this work, we develop several new tools to expand our capabilities for predictive design and high-throughput engineering in yeast. At the transcriptional level, we develop a method which, for the first time, enables predictive strengthening endogenous yeast promoters and also the de novo design of strong synthetic promoters. At the translational level, we show that it is possible to exploit the context resulting from the arrangement of DNA parts in order to predictably increase or decrease gene expression. We also develop a powerful new approach for directed evolution of enzymes in yeast, termed in vivo continuous evolution, which enables the creation of library sizes orders of magnitude larger than can be obtained with the current state of the art using significantly less labor. Finally, we harness the programmatic inhibitory potential of RNA interference to optimize and demonstrate a system for rapid strain engineering with minimal genomic editing. Taken together, this work provides new techniques which enable a significant reduction in the development time of new yeast strains and informs future development of new tools for metabolic engineering.