Developing genetic tools for model and non-model bacterial species through synthetic biology and protein engineering

Genetic engineering has long faced a reproducibility crisis, largely due to the complexity of biological systems where small variations in genetic sequence or context between strains or DNA constructs can result in opaque reasons for failed experiments. Synthetic biology seeks to overcome these challenges by establishing structured, top-down frameworks for engineering, enabled by genetic tools that perform predictably, robustly, and reproducibly. Such tools allow standardized design rules for genetic circuits and simplify systems so failures can be understood and rationally resolved. Central to this effort are modular, hierarchical Golden-Gate plasmid assembly systems, which break down individual transcriptional units into discrete genetic parts that can be combined in diverse configurations with predictable effects. However, ribosome binding sites, essential for protein production in bacteria, remain difficult to standardize in genetic circuits due to their context specific nature. Bi-cistronic design RBSs (BCDs) are one solution to this challenge as they mitigate sequence dependence but were not historically designed for compatibility with modular cloning. To address this, I developed a series of BCDs that behave robustly across genetic contexts and are compatible with modern modular-cloning frameworks. Much of the work on genetic tool development has occurred in E. coli, however, a key goal of synthetic biology is to extend our understanding and control of biological systems into new bacterial species with novel traits. To bring the genetic tools for non-model bacteria up to the same standard as E. coli, I adapted a modular cloning system from E. coli for environmental Actinomycetes and developed a suite of tools for three industrially relevant genera. Finally, I show how modular assembly of genetic circuits can be used to provide fine control for selection schemes used for high-throughput protein engineering. I use quantitative sequencing to map the phenotypic landscape of a deep mutational scanning library of the TetR-family transcriptional repressor CymR, uncovering novel phenotypes for biosensor development. Using the methods developed for CymR, we have extended this deep mutational scanning workflow to aminoacyl-tRNA synthetases to better understand how these enzymes can discriminate between canonical and non-canonical amino acids. Taken together, these efforts promote improved reproducibility in genetic engineering across circuit design, non-model Actinomycetes, and protein engineering.