Design, Construction, and Validation of Dynamic Gene Circuits in Yeast

Abstract

Cells are dynamical systems that constantly change and evolve following underlying rules governed by their genetic programs and interactions with a changing environment. Gene circuits that exhibit dynamic behavior, such as bistable switches and oscillators, are designed, constructed, and validated in synthetic biology with the aim of understanding the dynamics of biological systems and gaining the ability to engineer them. However, it remains challenging to construct even the most basic dynamic gene circuits such as bistable switch or oscillator in yeast, the model eukaryote that is of great interest to synthetic biology. Most academic labs use the typical design, build, and test cycle to build dynamic gene circuits, but the build and test components are often old fashioned, highly reliant on expert knowledge, error-prone, low-throughput, and difficult to debug. These challenges greatly hinder progress in engineering dynamic gene circuits. In this thesis, we present the design, analysis, and results of engineering dynamic gene circuits in the yeast Saccharomyces cerevisiae and software approaches that improve the build and test process. Specifically, we designed, built, and tested a synthetic bistable switch in yeast. To demonstrate the applicability of the switch as a functional module, we built a rewritable antibiotic resistance memory, which employs the switch as a component of its circuit architecture. To develop an oscillator in yeast, we designed an auxin-sensitive relaxation oscillator, performed mathematical modeling and analysis, and proposed a high-throughput method called Time-seq to tune the oscillator. To address the challenges inherent in the build and test processes, we developed and implemented a software system called Aquarium, a smart wetlab operating system that enables human-in-the-loop automation, where researchers submit orders and technicians execute experiments following instructions from Aquarium. For the build process, we developed two core cloning process in Aquarium: plasmid construction and yeast cloning. Through Aquarium, researchers view the build process as an input-output abstraction so they can instead focus on high-level designs and leave the build details to the software and technicians. For the test process, we developed a yeast cytometry process in Aquarium, where researchers submit orders to perform time course cytometry experiments with customizable media inputs, time points, and dilution rates. Researchers directly receive the cytometry results without the struggle of having to actually perform the cytometry experiment themselves. To streamline the downstream analysis, we developed a Python package named Cowfish to automatically perform data retrieval, labeling, and pre-processing of the cytometry results from Aquarium. These build and test processes in Aquarium provide a modern way of cloning and testing with repeatable, easy-to-debug, and high-throughput software-enabled human-in-the-loop automation.