RNA origami scaffolds facilitate cryo-EM characterization of a Broccoli-Pepper aptamer FRET pair

Abstract

Cryogenic electron microscopy (cryo-EM) is a promising method for characterizing the structure of larger RNA structures and complexes. However, the structure of individual aptamers is difficult to solve by cryo-EM due to their low molecular weight and a high signal-to-noise ratio. By placing RNA aptamers on larger RNA scaffolds, the contrast for cryo-EM can be increased to allow the determination of the tertiary structure of the aptamer. Here we use the RNA origami method to scaffold two fluorescent aptamers (Broccoli and Pepper) in close proximity and show that their cognate fluorophores serve as donor and acceptor for FRET. Next, we use cryo-EM to characterize the structure of the RNA origami with the two aptamers to a resolution of 4.4 Å. By characterizing the aptamers with and without ligand, we identify two distinct modes of ligand binding, which are further supported by selective chemical probing. 3D variability analysis of the cryo-EM data show that the relative position between the two bound fluorophores on the origami fluctuate by only 3.5 Å. Our results demonstrate a general approach for using RNA origami scaffolds for characterizing small RNA motifs by cryo-EM and for positioning functional RNA motifs with high spatial precision.

Figure 1.

FRET between Broccoli and Pepper aptamers. (A) Structural model of Broccoli and Pepper aptamers shown in cartoon format with their cognate fluorophores DFHBI-1T (green) and HBC620 (red) shown as spheres. Excitation, energy transfer and emission illustrated as wavy lines. (B) Measured excitation and emission spectra of DFHBI-1T and HBC620 in complex with their cognate aptamers. (C) Depiction of RNA origami tiles with different arrangements of the fluorogenic aptamers. (D) FRET output measured after 30 min upon addition of the fluorophores (1 μM) to the RNA origami tiles (100 nM). Measured fluorescence spectra at 460 nm excitation. Data corresponds to three technical replicates, shown as mean ± SD. (E) Calculated absolute FRET output measured at 460 nm excitation and 620 nm emission. Data corresponds to three technical replicates, shown as mean ± SD.

Figure 2.

Cryo-EM structure of Pepper and Broccoli aptamers in apo and bound states. (A) Overlay of the cryo-EM maps of the apo (red) and ligand bound (blue) 1,2-B12P12. (B) Atomistic model built into the ligand bound cryo-EM map of 1,2-B12P12 showing the Brocolli (green) and Pepper (red) aptamer locations. (C) Close up view of the Brocolli aptamer in the apo and DFHBI bound cryo-EM maps. (D) Close up view of the Pepper aptamer in the apo and HBC620 bound cryo-EM maps.

Figure 3.

SHAPE probing of the Pepper aptamer in apo and bound states. (A) Secondary structure blueprint for Pepper with the labelling used in the text. (B) SHAPE gel analysis of pepper aptamer in the apo and HBC485, HBC497 and HBC620 bound states. Grey marking in sequencing lane indicates compressed area. (C) Quantitative per-nucleotide SHAPE reactivity analysis for Pepper aptamer in the apo and ligand bound states. Signals are normalized by the signal at the non-binding C28 position. (D) Tertiary structure of the Pepper aptamer (PDB ID: 7EOP). Structure diagram showing tertiary elements: Base pairs are shown as grey lines with Leontis-Westhof annotation of non-Watson–Crick base pairs. Base pair planes are indicated by horizontal alignment. Stacking is indicated by vertical alignment. Protection is marked as colors on nucleotides. (E) Atomic structure shown with protection colored on nucleotides showing that the whole binding pocket gets stabilized upon ligand binding. HBC 620 shown in red sphere representation.

Figure 4.

Cryo-EM 3D variability analysis of Pepper and Broccoli aptamers. (A) Representative structures are shown for the principal component analysis (PC0, PC1 and PC2). Arrows indicate the most prominent movements. Gaussian distribution of particles along two principal reaction coordinates (PC1, PC2) determined by 3DVA (right). (B) Intermediate reconstructions using particle subsets from the extremes of PC0 and PC1 show the positional variability of broccoli and pepper aptamers. (C) HBC ligands from pepper were aligned while maintaining the spatial relationship with the DFHBI-1T ligand from Broccoli from the extrema reconstructions. Distances from HBC to the center of each DFHBI-1T were measured as well as the furthest distance between DFHBI-1T fluorophores.