Assistant Professor University of Michigan Ann Arbor, Michigan, United States
Introduction:: Synthetic biology toolkits for model organisms allow precise control of gene expression and have been crucial for studying and engineering these microbes. Genetic tools for organisms like E. coli and B. subtilis enabled fundamental insights into metabolism, cell division, and population-level dynamics. However, understudied bacteria lack these highly characterized genetic toolkits since individual genetic parts are sensitive to changes in sequence context and cannot be transferred between species. As a further complication, interactions between expression cassette elements affect gene expression levels, but no effort has been made to systematically quantify these effects in non-model bacteria -- and for good reason. Estimating interaction effects requires a combinatorial approach with a massive number of experiments that must be repeated for each species.
Here we present a generalizable experimental pipeline for the Rapid Assembly and Characterization of a Combinatorial Expression Toolkit (RACCET). We used RACCET to systematically investigate inter-element interactions within thousands of gene expression cassettes. We simultaneously used RACCET to build a catalog of genetic parts offering a >260-fold range of transcriptional strengths and a >425-fold range of translational strengths in the non-model bacterium and human pathogen, Streptococcus mutans. Our results demonstrate how RACCET provides a fast and scalable method for assembling and characterizing synthetic biology toolkits for understudied bacteria.
Materials and Methods:: RACCET begins with the assembly of an uncharacterized collection of promoter sequences, ribosome binding sites, and reporter ORFs. Using a pooled Golden Gate assembly approach, we constructed a large S. mutans expression library whereby each strain harbored a single, chromosome-integrated expression cassette containing one of > 850,000 possible promoter-RBS-ORF combinations. An NGS barcode counting assay was then utilized to assess the transcriptional strengths of >30,000 expression strains. Following selection of a validated set of promoters spanning the observed transcriptional range, a Golden Gate assembly pipeline integrated with liquid handling robotics was used for the individual construction and characterization of >1000 unique S. mutans fluorescent reporter strains. The combined results of the transcriptional and translational characterization experiments allowed us to quantify the global contribution of interaction effects on both processes while building a characterized gene expression toolkit containing high-precision parts offering control over a wide range of gene expression levels.
Results, Conclusions, and Discussions:: At the transcriptional level, promoter effects dominate, accounting for >90% of the explainable variance in observed transcript abundances. Interestingly, inter-element interactions dominate at the translational level, particularly between the promoter and the RBS. These interactions account for over 40% of the explained variance in protein expression. We used these data to select high-precision parts for a gene expression toolkit in S. mutans. This final toolkit contains 12 promoters offering a >260-fold range of transcriptional strengths, and 15 promoter-RBS pairs offering a >425-fold range of translational strengths. Based on these results, RACCET offers researchers an experimental framework for rapidly building gene expression toolkits in a way that properly accounts for the interactions between genetic parts. Increasing the availability of characterized genetic parts for understudied microbes would be a major step in addressing the current, severe knowledge disparity between understudied bacterial species and their model counterparts.