RNA systems for NMR studies in vitro and in vivo

Abstract

NMR spectroscopy is an excellent tool to study the structure-function relationship of RNA. Such measurements are usually performed in vitro, which requires large amounts of isotope- labeled sample in high purity and can give access to individual atoms, structure of the molecule and conformational dynamics. This is contrasted by measurements in living cells, where researchers struggle with low signal intensity, line-broadening and rapid sample degradation. In this work, we developed sample preparation methods for NMR studies to expand the range of RNA constructs that are accessible for NMR studies in vitro and in cells.

Firstly, we improved yield and purity of in vitro transcription of short RNA constructs by transcribing several repeating target sequences from a tandem template, and cleaving them to the target length with RNase H. This abolishes issues with suboptimal initiation sequences and creates higher purity due to the high sequence-specificity of RNase H guided by a chimeric oligo. We demonstrated the high yield and purity of several such RNA molecules and incorporated the protocol into a workflow for studies of conformational dynamics with relaxation dispersion NMR.

Secondly, we demonstrated the site-specific incorporation of a 13 C/15 N-labeled adenosine into a 46 nt RNA molecule with the use of purely enzymatic methods. Such site-specific labeling is an effective approach to overcome resonance overlap in larger RNAs, which can preclude further structural and dynamics studies. We showed the facile production of such a sample and reported on a second conformation which would in a uniformly labeled sample be hidden by overlapping resonances.

Lastly, we furthered method development for in-cell NMR methods by exploring transfection strategies, cell culture methods and RNA systems. We adapted a protocol for the production of circular RNA at high concentration in HEK293T cells to generate the first in-cell NMR spectra of intracellular expressed RNAs. Furthermore, we produced the same circular RNAs by in vitro transcription and ligation to assess their improved stability against cellular exonucleases. As circular RNA model systems, we used the fluorescent aptamer Broccoli and a small hairpin RNA, called GUG, which proved useful for relaxation dispersion NMR measurement previously. The expression of both circular constructs at was possible at micromolar concentration in HEK283T cells and both constructs could be transcribed and circularized in vitro. In-cell NMR of the expressed circular RNA did however not yield detectable signals, indicating that either the intracellular concentration is too low, or the location of the expressed RNA precludes free tumbling.