Background

The function of biomolecules is rooted in their three-dimensional structure, dynamics, and interaction with other molecules. For example, proteins and oligonucleotides adopt structures that provide interaction sites or binding pockets for metal ions, organic ligands, and other oligonucleotides or proteins. Upon complex formation, the structure of the biomacromolecule may change, and it is this structural change that is the basis for function; thus, methods are needed that enable resolution of the structures along the trajectory of the conformational change. X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy, electron microscopy (EM), small-angle X-ray scattering (SAXS), atomic force microscopy (AFM), optical trap, and magnetic tweezers are powerful methods for obtaining the structures of biomolecules at atomic resolution (Salas et al., 2015; Lottspeich and Engels, 2018; Geffroy et al., 2018); yet, they also have their limitations. X-ray crystallography requires the biomolecule to be crystallized and is, as EM, limited in providing information on dynamics and conformational intermediates. NMR spectroscopy can follow the dynamics of biomolecules since it is performed in liquid solution, but is limited with respect to the size of the biomolecule. Complementary to these techniques, fluorescence microscopy (FM), Förster resonance energy transfer (FRET), and electron paramagnetic resonance (EPR) spectroscopy are biophysical methods that allow resolution of the dynamics of structural changes without size restriction and in solution (Yang et al., 2012; Lottspeich and Engels, 2018; Kuzhelev et al., 2018). In the toolbox of EPR spectroscopy (Schweiger et al., 2001; Goldfarb et al., 2018), a set of pulse sequences summarized under the term Pulsed Dipolar EPR Spectroscopy (PDS; Schiemann et al., 2007) measures the dipolar coupling between the spins of unpaired electrons. This coupling encodes the inter-spin distance, from which structural and dynamic information can be derived. Owing to the high sensitivity of PDS, measurements at biomolecular concentrations down to the nanomolar range are feasible (Fleck et al., 2020). Out of all PDS techniques, Pulsed Electron-Electron Double Resonance (PELDOR or DEER) is the most prominent and works in a distance range from 1.5 to 16 nm upon deuteration of the buffer and the whole protein (Schmidt et al., 2016) with Angstrom precision (Jeschke, 2012; Tsvetkov et al., 2019). PELDOR imposes no size restriction (Malygin et al., 2019), and can be performed on biomolecules in liquid (Yang et al., 2012) or frozen solution (Duss et al., 2014), in membranes (Dastvan et al., 2019), and within cells (Theillet et al., 2016). However, since PELDOR requires unpaired electrons and biomolecules are usually diamagnetic, techniques are needed for site-directed spin labeling (Shelke and Sigurdsson, 2014). Focusing on RNA oligonucleotides, there are two principal strategies for spin labeling (Ward and Schiemann, 2014): the first is the phosphoramidite approach, where a spin-labeled phosphoramidite is incorporated into the RNA strand during solid-phase synthesis (Beaucage et al., 1992; Lottspeich and Engels, 2018) and the second is the post-synthetic method, where a functionalized RNA strand is labeled after synthesis (Kerzhner et al., 2016). The major drawbacks of the phosphoramidite approach are the rather laborious synthesis of the spin-labeled phosphoramidite and the easy reduction of the label during RNA synthesis, leading to an EPR-silent state. For the post-synthetic strategy, RNA strands modified with a unique functional group can be obtained commercially and are then reacted with a spin label carrying the complementary functional group. However, labeling with high yields and efficient purification with minimal product loss can be challenging. Even though the length of commercially available RNA oligonucleotides is limited, there are well-established methods for obtaining longer RNA strands, e.g., via enzymatic ligation (Duss et al., 2014; Kerzhner et al., 2018).

In this protocol, built on the publications Kerzhner et al., 2018, Wuebben et al., 2019 and 2020, we first introduce a workflow for the post-synthetic spin labeling of RNA oligonucleotides based on the “Click” reaction. Here, the “Click” reaction is performed as a copper-(I)-catalyzed [2+3] cycloaddition between an alkyne-substituted 5’-uridine in the RNA strand and an azide group on a nitroxide label (Figure 1). The reaction is usually quantitative and fast, and the purification of the labeled RNA can be easily performed via reversed-phase High Performance Liquid Chromatography (HPLC). Secondly, the protocol outlines how to set up and perform the PELDOR measurement. Thirdly, it provides a workflow for data analysis and the transformation of the distances into structures. Importantly, possible pitfalls are highlighted and tips provided on how to overcome experimental difficulties, which are rarely found in the literature.

Figure 1. Reaction scheme of the “Click” reaction used in this protocol.The azide-functionalized spin label (azide group in blue; nitroxide moiety and 1,1,3,3-tetraethyl-5-azoisoindolin-2-oxyl backbone in red) is attached to a 5-ethynyl-2′-deoxyuridine nucleotide (ethynyl group in green) in the RNA strand (black). Copper, in its oxidation state (I), forms a complex with tris ((1-hydroxy-propyl-1H-1,2,3-triazol-4-yl) methyl) amine (THPTA), catalyzing this reaction in aqueous solution, which leads to the formation of a triazol linker (cyan) covalently connecting the nitroxide moiety to the RNA.

Beyond this application, labeling via “Click” chemistry is restricted neither to RNA nor to nitroxides or spin labels in general. It can be applied to e.g., DNA (El-Sagheer and Brown, 2010) and proteins (Nikić et al., 2015) and can be used for fluorescent labels (Liang et al., 2019). For protein labeling, one would make use of unnatural amino acids e.g., 4-ethynyl-L-phenylalanine (Widder et al., 2019). Note, however, that the reaction conditions will have to be adapted to the particular system under study. The workflow for the Q-band PELDOR experiment is universally applicable to measurements of nitroxide-labeled biomolecules. Likewise, the data analysis and distance-to-structure transformation can be adapted for other PDS techniques or other types of spin labels.