High-throughput functional profiling and evolutionary covariation analysis of entire riboswitch sequences.

Riboswitches are useful models for revealing how some RNA molecules undergo dynamic rearrangements of their structures to perform cellular functions. A great deal is known about the structure of riboswitch ligand-binding aptamer domains through evolutionary sequence covariation analysis. However, covariation analysis has been more difficult to apply to riboswitch expression platforms given their large range in cellular functions, and their large sequence diversity. Here, we develop an approach to identify whole transcriptional riboswitch sequences starting from their conserved aptamer domains. We then generate covariation models for the entire riboswitch including the aptamer domain and the expression platform. The method consists of first bioinformatically extending identified aptamer domains to include downstream sequence that could contain an expression platform. Filtering is then performed using either a computational prediction algorithm to identify bacterial intrinsic terminator sequences in the expression platform, or a high-throughput functional assay that uses massive parallel oligo synthesis and next generation sequencing to characterize transcriptional termination of riboswitch candidates as a function of ligand. Filtered sequences are then used to develop full riboswitch sequence covariation models. We developed this approach in the context of the fluoride riboswitch, characterizing 1901 fluoride riboswitch sequences using our high-throughput assay. We find that the prediction filtering approach results in few false positives to identify novel, highly functional fluoride riboswitch variants. Finally, we employ the computational approach to develop covariation models of the ZTP, lysine, and TPP riboswitches and find covariation support for previously published rearrangement mechanisms. Overall, our method represents a new hybrid computational and high throughput experimental approach to characterize large numbers of riboswitch sequences and to generate new covariation models of complete riboswitch sequences, which should expand our understanding of riboswitch mechanism and the evolution of RNA structure dynamics.