Development, characterization, and application of RNA catalysts for in situ labeling of target RNA molecules

Abstract

RNA molecules play a diverse set of crucial roles in biological systems. Studying various aspects of RNA biology such as RNA transport, localization, degradation, and structural dynamics is therefore of utmost significance. A prerequisite for studying such phenomena is availability of methods for labeling and visualization of the RNAs of interest. A wide array of RNA labeling methods has been developed over the years with promising results. Nevertheless, versatile, and efficient tools for covalent site- and sequence-specific labeling of RNA in live cell (in situ) are non-existent. One of the more recently emerged RNA labeling methods involves application of the 10DM24 deoxyribozyme. The recognition arms of this DNA catalyst recognize the desired labeling positions in the RNA of interest via Watson-Crick base pairing. Using a wide variety of 2'-modified GTP analogs, 10DM24 can label the target RNA at specific internal positions, efficiently and accurately. Despite the success achieved using this system, the deoxyribozyme has never been tested for RNA labeling within live cell. The dependency of the deoxyribozyme on metal ions and challenges regarding maintenance and delivery of DNA catalysts in the cell, has so far limited its application to labeling in vitro. Therefore, the aim of this study was to develop ribozymes for RNA labeling with the potential for cellular applications. We established a selection method through which selection of trans-acting, site-selective RNA labeling ribozymes was made possible. The selection involved a partially structured RNA pool with a bulged adenosine nucleotide as a predetermined modification-site. For the development and optimization of the selection strategy, biotinylated ATP was used as the selection substrate. The selection using this substrate led to the identification of adenylyltransferase ribozymes, denoted as FH ribozymes. The characterization of these catalysts confirmed the success of our selection method in dictating the modification site of ribozymes. The selected variants were effectively directed to modify the bulged adenosine nucleotide by forming a 2'-5'-branched phosphodiester bond with labeled adenosines. The FH ribozymes could be readily converted to trans-acting variants, and the ribozyme showed a broad substrate scope based on Watson-Crick base-pairing between target RNA and ribozyme binding arms. The selected ribozymes, especially the most efficient variant (FH14), efficiently accepted a wide range of N6-modified-ATP analogs as labeling substrates, including fluorophore-conjugated derivatives. FH14 was successfully applied in 2 site-specific labeling of large, heavily structured cellular RNA such as 5S, 16S, and 23S ribosomal RNA in total cellular RNA. In the next stage, we attempted to develop ribozymes that utilized a more bioorthogonal substrate such as the antiviral ATP analog tenofovir-diphosphate. Ligation of tenofovir analogs to the target RNA would also result in a more enzymatically stable linkage type, in which the branched 2'-5'-phosphodiester is replaced by a phosphonate ester. Biotinylated tenofovir-diphosphate was synthesized and applied in an in vitro selection process that led to the identification of tenofovir-transferase (FJ) ribozymes. These ribozymes showed similar properties to FH14 ribozyme, in terms of site-selectivity, substrate sequence generality, and broad labeling substrate scope. Moreover, these ribozymes were perfectly orthogonal to FH14 ribozyme, a feature that was exploited for dual-color RNA labeling. The ribozymes FJ1 and FH14 were successfully applied in simultaneous dual-color labeling of a synthetic transcript at two different positions. These ribozymes have also been used for simultaneous, site-specific labeling of 16S and 23S rRNAs using two different fluorophores. Although the cellular application of these ribozymes has not yet been demonstrated, they possess great potential for further optimization into efficient tools for RNA labeling in situ. Furthermore, the established selection process paves the way for future development of other RNA labeling tools, using non-nucleotide based selection substrates.