Turbocharge RNA Device Engineering with Data-Rich Strategies and Automation

Engineered biological systems hold great potential to transform technologies in health, medicine and sustainability. Challenges in the design of biological devices include our incomplete understanding of the sequence-function relationships underlying the design space, and the limited throughput of experiments required to characterize individual devices. To significantly improve the efficiency of engineering highly functional gene-regulatory RNA devices, we 1) developed a massively parallel FACS-seq screen to assay the activity of hundreds of thousands of ribozyme switch library variants in yeast and mammalian cells, 2) adapted and automated a similarly high-throughput in vitro Cleave-seq assay on a liquid handler for parallelization and rapid iteration of the design-build-test cycle, and 3) employed computational analyses to extract design principles from the large datasets to inform future rational design.

To assay their activity in yeast or mammalian cells, ribozyme switches - a class of RNA devices with ligand-controllable self-cleavage activity - are integrated into the untranslated region of a fluorescent reporter, such that cleavage directly modulates fluorescence levels. Combining fluorescence-activated cell sorting (FACS) and next-generation sequencing (NGS), we extensively interrogated the effect of tertiary loop I-loop II interaction motifs on ribozyme activity, characterizing 10⁵ variants in a single FACS-seq experiment¹. Cleave-seq measures in vitro cleavage activities of the ribozyme switches at a similar throughput using NGS, and automation of this method allowed us to quickly iterate through different library designs, while obtaining insight into differences in conserved sequence motifs between in vitro experiments and inside cells. With these high throughput techniques, we generated new RNA switches to a number of small molecule therapeutics and metabolites, which can serve as gene-control devices and live cell genetic sensors. Our analysis of the large datasets also identified a set of general sequence motifs that result in enhanced ribozyme cleavage activity, highlighting the importance of pseudoknot-like base pairs at the base of the ribozyme stem loops to mediate tertiary interactions. These datasets serve as a rich resource to further enable the discovery of novel hammerhead ribozymes with bioinformatics approaches and streamline the design of RNA switches for synthetic biology applications.

¹Townshend, B., Kennedy, A. B., Xiang, J. S., & Smolke, C. D. (2015). High-throughput cellular RNA device engineering. Nature methods, 12(10), 989-994.