Dual-colour imaging of RNAs using quencher- and fluorophore-binding aptamers

In order to gain deeper insight into the functions and dynamics of RNA in cells, the development of methods for imaging multiple RNAs simultaneously is of paramount importance. Here, we describe a modular approach to image RNA in living cells using an RNA aptamer that binds to dinitroaniline, an efficient general contact quencher. Dinitroaniline quenches the fluorescence of different fluorophores when directly conjugated to them via ethylene glycol linkers by forming a non-fluorescent intramolecular complex. Since the binding of the RNA aptamer to the quencher destroys the fluorophore-quencher complex, fluorescence increases dramatically upon binding. Using this principle, a series of fluorophores were turned into fluorescent turn-on probes by conjugating them to dinitroaniline. These probes ranged from fluorescein-dinitroaniline (green) to TexasRed-dinitroaniline (red) spanning across the visible spectrum. The dinitroaniline-binding aptamer (DNB) was generated by in vitro selection, and was found to bind all probes, leading to fluorescence increase in vitro and in living cells. When expressed in E. coli, the DNB aptamer could be labelled and visualized with different-coloured fluorophores and therefore it can be used as a genetically encoded tag to image target RNAs. Furthermore, combining contact-quenched fluorogenic probes with orthogonal DNB (the quencher-binding RNA aptamer) and SRB-2 aptamers (a fluorophore-binding RNA aptamer) allowed dual-colour imaging of two different fluorescence-enhancing RNA tags in living cells, opening new avenues for studying RNA co-localization and trafficking.


Preparation of the affinity resin
NHS-activated sepharose (GE Life Sciences) was washed with 2 volumes of ice-cold Millipore water. The resin was resuspended in 100 mM HEPES buffer (pH 7.4). Dinitroaniline-PEG 3 -Amine (12.1 mg, 38.5 µmol, for synthesis see section B1) dissolved in 500 µL of DMSO was added into the resin dropwise with vigorous shaking to ensure homogenous functionalization of the resin, and then incubated at 25°C for 3 hours. Further, the resin was incubated with 0.5 M ethanolamine at 25°C for 2 h to react with any free NHS-activated sites. The resin was washed thoroughly and stored in 100 mM Tris buffer (pH 7.5) at 4°C. The coupling efficiency was determined by measuring the absorbance of unreacted dinitroaniline in the flow-through. Using this strategy, we estimated that the resin contained 7 µmol of quencher per ml of resin. Mock-resin was also prepared by using the same approach where only DMSO was added to the resin instead of Dinitroaniline-PEG 3 -Amine.

Library design and preparation
To prepare a partially structured RNA library, a DNA oligonucleotide was synthesized that contained two fixed primer binding sites flanking a 64-nucleotide region. This region consisted of two 26-base random stretches that were separated by a 12-base constant region designed to form a stable CCGU stem-loop in transcribed RNA (Supplementary Figure S1B). The single-stranded oligonucleotide was synthesized in 1 µmol scale and phosphoramidites for the random regions were mixed in a ratio of 3:3:2:2 (A:C:G:T). The randomized singlestranded oligonucleotide was PAGE-purified, and 1.6 nmol of the oligonucleotide were amplified in a 30 mL PCR reaction for 6-cycles by using the forward and reverse primers (see Supplementary Table S3 for sequences) to yield double-stranded DNA template for transcription of the library. The PCR product was precipitated with sodium acetate and ethanol after phenol:chloroform:isoamyl alcohol (25:24:1) extraction. The DNA pellet was dissolved in water and directly used for in vitro transcription reaction.

Random mutagenesis of round 15 pool (first SELEX) and clone 5 (best aptamer identified from round 15)
Random mutagenesis was performed by a mutagenic PCR protocol (1), that was developed to increase mutation frequencies involving high Mg concentration, unequal dNTPs ratios, and addition of MnCl 2 . For PCR, the following reagents were added: 1X PCR buffer (Rapidozyme; 67 mM Tris, pH 8.

Structure predictions and truncation studies
The RNA sequences obtained from sequencing the pool were subjected to secondary structure prediction using Mfold software (2). The active clones were truncated by deleting the constant primer binding regions from both ends. Mutated and truncated aptamers were created by ordering single-stranded DNA templates (Integrated DNA Technologies) with desired mutations and PCR-amplified to form double-stranded templates. PCR products were purified with a PCR purification kit (QiaGen) and used as templates for in vitro T7 transcription reactions.
S4 B) Synthesis of the compounds

Synthesis of Dinitroaniline-PEG3-Amine (DN-PEG3-Amine)
To a stirring solution of 2,2'-(ethylenedioxy) bisethylenediamine (3.98 g, 27 mmol) in 20 mL of dichloromethane (DCM) at 0°C was added a solution of dinitrofluorobenzene (0.50 g, 2.7 mmol) in 10 mL of DCM, dropwise. After the addition was complete, the temperature was brought to room temperature and the mixture was stirred for 30 minutes. Then, the reaction mixture was mixed with 100 mL of water and the organic phase was recovered. The organic phase was mixed with 100 mL of 0.1 M HCl and the product was taken into the acidic aqueous phase and the DCM phase was discarded. Finally, the pH of the aqueous phase was adjusted to 12-13 with NaOH and the product was extracted into the DCM phase using 2 x 100 mL of solvent. Organic phases were combined, washed with brine (100 mL), dried over Na 2 SO 4 , filtered and evaporated to yield DN-PEG3-Amine (0.77 g, 91%). The product was used without further purification for the next step. HR-ESI (positive): calculated 315.1299, found 315.1310 for C 12 H 19 N 4 O 6 .

Synthesis of Dinitroaniline-PEG3-SS-PEG4-Biotin (DN-SS-Biotin)
To a mixture of DN-PEG 3 Table S1: Sequences of the colonies picked from round 15 pool of SELEX 1