Fluorescent End-Labeling and Encapsulation of Long RNAs for Single-Molecule FRET-TIRF Microscopy

Summary

This article describes the dual-color labeling of long RNAs at termini positions and their surface immobilization via encapsulation in phospholipid vesicles for single-molecule FRET TIRF microscopy applications. Combining these techniques enables precise visualization and analysis of RNA dynamics at the single-molecule level.

Abstract

Single-molecule Förster Resonance Energy Transfer (smFRET) excels in studying dynamic biomolecules by allowing precise observation of their conformational changes over time. To monitor RNA dynamics with smFRET, we developed a method to covalently label RNAs at their termini with a FRET pair of fluorophores. This direct end-labeling strategy targets the 5'-phosphate by carbodiimide (EDC)/N-hydroxysuccinimide (NHS) activation and the 3'-ribose by periodate oxidation, which can be adapted to other RNAs regardless of their size and sequence to study them independently of artificial modifications. Furthermore, the 5'-EDC/NHS activation is of general interest to all nucleic acids with a 5'-phosphate. The use of commercially available chemicals eliminates the need to synthesize RNA-specific probes.

Total Internal Reflection Fluorescence (TIRF) microscopy requires the surface-immobilized molecules of interest to be within the evanescent field to be illuminated. A sophisticated way of keeping the RNA molecules within the evanescent field is to encapsulate them in phospholipid vesicles. Encapsulation benefits from the best of both worlds, tethering the molecule to the surface while enabling free diffusion of the molecule. We ensure that each vesicle contains only a single RNA molecule, enabling single-molecule imaging. Upon dual-end labeling and encapsulation of the RNA of interest, smFRET measurements offer a dynamic and detailed view of RNA behavior.

Introduction

Förster Resonance Energy Transfer (FRET) is a powerful and sensitive technique for studying inter- and intramolecular interactions of biomolecules at the nanoscale. It is based on the non-radiative energy transfer from an excited donor molecule to a nearby acceptor molecule, which occurs over distances typically between 1 nm and 10 nm. The distance between the donor and acceptor dyes determines the efficiency of this energy transfer, making FRET an invaluable tool for studying molecular dynamics, conformational changes, and interactions in a wide range of biological systems^1,2, including RNAs. Total Internal Reflection Fluorescence (TIRF) microscopy has proven to be a powerful technique for smFRET investigations, as it selectively illuminates molecules only near the surface, allowing FRET dynamics of individual molecules with high spatial and temporal resolution. However, before performing smFRET-TIRF experiments, the molecule of interest must first be fluorescently labeled with an appropriate FRET pair and then immobilized on the microscopy surface. The smFRET-TIRF protocol described here was validated using the wild-type group II intron Sc.ai5γ from Saccharomyces cerevisiae mitochondria, flanked by its two exonic sequences (915 nucleotides)³. For a more detailed view of fluorescent labeling group II introns and their immobilization for smFRET TIRF microscopy, refer to our review⁴.

An ideal site-specific RNA labeling strategy would allow for the precise incorporation of donor and acceptor dyes at predetermined positions without altering the RNA's structure or function, ensuring accurate and efficient FRET measurements. This is challenging due to the chemical similarities among the four nucleobases, which complicates selective labeling. End-labeling attaches donor and acceptor dyes to the RNA ends by targeting the 5'-phosphate and the 3'-ribose. This approach offers a minimally invasive approach while still providing valuable insights into structural dynamics and interactions. Group II intron's ability to self-splice in the presence of Mg²⁺ restricts the use of metal ion-dependent enzymes. Here, we present an approach to dual end-label long and catalytically active RNAs (ribozymes) that bypass the need for enzymes or the synthesis of specialized probes.

A common approach to tether RNA molecules to the surface for TIRF microscopy is to covalently link a biotin moiety directly to the RNA or to hybridize a biotin-carrying antisense oligonucleotide (ASO)^5,6,7. However, this direct immobilization method can introduce artifacts due to RNA-surface interactions, potentially resulting in misfolded RNAs⁸. An elegant solution to mitigate these immobilization artifacts is to encapsulate RNA in surface-attached nanoscale phospholipid vesicles^9,10,11. These vesicles, which are approximately 100 nm in diameter, are anchored to the surface through a biotin-streptavidin linkage^12,13,14, allowing the RNA to diffuse freely inside while permitting the exchange of ions across the lipid membrane¹⁰. After covalently labeling a large functional RNA³, we present an approach to encapsulate such RNAs in phospholipid vesicles by combining established protocols for surface passivation and vesicle encapsulation, adapted to preserve RNA functionality^10,11,14. This dual-end labeling and encapsulation approach achieves a high rate of mono-encapsulation of functional RNAs for smFRET TIRF microscopy.

Protocol

1. RNA dual end-labeling

NOTE: The following protocol describes the site-specific labeling of RNAs with a FRET pair of fluorophores by covalent attachment of a donor dye (sCy3) to the 5'-phosphate and an acceptor dye (Cy5) to the 3'-ribose. A catalytically active long RNA, the group II intron ribozyme, is chosen as the RNA of interest. Table 1 and Figure 1 summarize this dual-end labeling protocol. Perform all steps involving fluorophores under dark conditions.

Day 0	▪ Aliquot 50-75 µg of RNA to a total volume of 55 µL per 1.5 mL tube.
Day 1	5′-Phosphate activation
	▪ Add 45 µL of freshly prepared EDC-NHS, pH 6.0 solution to the RNA in ddH2O to a final volume of 100 μL, mix well and incubate for 4 h at 25 °C and 500 rpm.
	▪ Purification round 1: Overnight EtOH precipitation.
Day 2	▪ Precipitate the 5′-activated RNA, wash, and dry.
	5′-Dye attachment
	▪ Resuspend in 95 µL of 100 mM MOPS, pH 7.5.
	▪ Add 5 μL of 2 mM amine-functionalized dye solution.
	▪ Mix well and incubate for 16 h at 25 °C and 500 rpm.
Day 3	▪ Purification round 2: EtOH precipitation.
Day 4	▪ Precipitate the 5′-activated RNA, wash, and dry.
	Blocking step
	▪ Resuspend in 100 μL of 100 mM Tris–HCl, pH 7.5, and incubate for 2 h at 25 °C and 500 rpm.
	▪ Purification round 3:  Centrifugal filtration.
	→ Elute the 5′-labeled RNA.
Day 5	3′-Periodate oxidation
	▪ Incubate the RNA with 20 mM NaIO4 in 50 mM NaOAc buffer, pH 5.5 in a final volume of 100 μL  for 2 h at 25 °C and 500 rpm.
	▪ Quench the excess periodate: Add 30 μL of 50% glycerol, mix well, and incubate for 30 min at 25 °C and 500 rpm.
	▪ Purification round 4: Overnight EtOH precipitation.
Day 6	▪ Precipitate the 3′-oxidized RNA, wash, and dry.
	3′-Dye attachment
	▪ Resuspend in 95 µL of 50 mM NaOAc, pH 6.0.
	▪ Add 5 μL of 2 mM hydrazide-functionalized dye solution. Mix well and incubate for 16 h at 25 °C and 500 rpm.
Day 7	▪ Purification round 5: EtOH precipitation.
Day 8	▪ Precipitate the labeled RNA, wash, and dry.
	▪ Centrifugal filtration.
	→ Elute the dual-end labeled RNA.

Table 1: Protocol summary for RNA dual-end labeling. Please click here to download this Table.

Figure 1: The experimental flow of dual-end labeling by targeting the 5'-phosphate and the 3'-ribose sugar. The 5'-phosphate is activated using EDC in the presence of NHS and is subsequently coupled with the amine-functionalized dye. The 3'-diol moiety of RNA is oxidized by periodate activity to dialdehyde, which further reacts with the hydrazide-functionalized dye. For dual-end labeling, it is important to start with the 5'-labeling to prevent cross-labeling, followed by 3'-labeling with an intermediate blocking step. Abbreviations: EDC = carbodiimide; NHS = N-hydroxysuccinimide; MOPS = 3-morpholinopropane-1-sulfonic acid; NaOAc = sodium acetate. Please click here to view a larger version of this figure.

1. 5'-end labeling: Targeting the 5'-phosphate by EDC/NHS activation
   NOTE: This 5'-labeling method applies not only to RNA but to any single-stranded nucleic acid containing a 5'-phosphate. The specificity of the reaction is governed by pH dependence, allowing only the 5'-phosphate to be site-specifically targeted, despite the presence of multiple phosphates in the RNA backbone. At pH 6.0, the 5'-phosphate's unique reactivity toward carbodiimides is due to its specific protonation state, where two oxygens are deprotonated, and one remains protonated. This makes the 5'-phosphate reactive, while the fully deprotonated backbone phosphates remain unreactive, enabling selective labeling of the 5'-end via EDC/NHS targeting.
   1. Prepare the EDC-NHS-NaOAc, pH 6.0 solution. Mix 1.5 mg of EDC and 2.0 mg of NHS per aliquot in 35 µL of ddH₂O and 10 µL of 0.5 M NaOAc, pH 6.0 (pH adjusted with glacial acetic acid).
   2. Aliquot the RNA of interest to have approximately 50-75 µg RNA per tube in 55 µL of ddH₂O. Add 45 µL of the EDC-NHS-NaOAc, pH 6.0 mixture to the RNA, reaching a total volume of 100 µL with the final concentrations of 78 mM EDC, 174 mM NHS, and 50 mM NaOAc, pH 6.0. Incubate for 4 h at 25 °C while shaking at 500 rpm.
      NOTE: Here, 70 µg corresponds to 250 pmol of an in vitro transcribed 915 nt RNA and 2.5 µM in the final reaction volume of 100 µL.
   3. Purification round 1: Purify the 5'-phosphate-activated RNA by EtOH precipitation.
   4. Prepare a 2 mM sulfonated Cyanine3 amine (sCy3-amine) solution in water.
   5. Resuspend the activated RNA pellet in 95 µL of 100 mM 3-morpholinopropane-1-sulfonic acid (MOPS) buffer, pH 7.5 (pH adjusted with NaOH). Add 5 µL of 2 mM sCy3-amine solution to the activated RNA. Couple the fluorophore to the activated 5'-phosphate by incubation for 16 h at 25 °C, while shaking at 500 rpm.
      NOTE: Store MOPS buffer at 4 °C in the dark. NaOH is chosen for pH adjustment to keep the group II intron ribozyme inactive. The fluorophore should be at least 40-fold excess of the RNA.
   6. Purification round 2: Increase the volume by adding 200 µL of ddH₂O to improve the separation. Purify the 5'-phosphate-labeled RNA by EtOH precipitation and repeat until the supernatant is colorless (usually two rounds are needed).
   7. Blocking step: Resuspend the RNA in 100 µL of 100 mM Tris-HCl, pH 7.5, and incubate for 2 h at 25 °C and 500 rpm. For 5'-end single-labeling only, skip the blocking step, resuspend the RNA in ddH₂O, and proceed with step 1.1.8. However, for simplicity, if dual-labeling, separate the 5'end single-labeled control only after step 1.1.9.
      NOTE: This step serves to block the activated 5'-phosphates that have not been coupled to an amine-functionalized fluorophore by reacting with a relatively smaller primary amine source (e.g., Tris) to minimize the risk of cross-labeling with the hydrazide-functionalized fluorophore that is used for the 3'-end labeling protocol.
   8. Purification round 3: Remove the free dyes by washing the labeled RNA by centrifugal filtration with a total of at least 10 mL of ddH₂O, then elute in ddH₂O.
      NOTE: The molecular cut-off of the filter should be less than half the size of the nucleic acid. Buffer of choice can be used instead of ddH₂O. Elution should follow the manufacturer's instructions, ensuring the sample is not spun to complete dryness.
   9. Determine the RNA and conjugate dye concentrations by UV-Vis spectroscopy.
2. 3'-end labeling: Targeting the 3'-ribose by periodate oxidation
   1. Aliquot the 5'-end labeled RNA in ddH₂O to have approximately 50-75 µg of RNA per tube in 90 µL. If the elution volume of the previous step was high, resulting in a low concentration, concentrate the RNA by precipitating and resuspending the pellet in ddH₂O. Add 5 µL of 1.0 M NaOAc buffer, pH 5.5 (corresponding to 50 mM NaOAc, pH 5.5 for 100 µL reaction volume).
      NOTE: Similar to the 5'-end single-labeled RNA, we routinely prepare 3'-end single-labeled RNA as a control. To do this, aliquot the unlabeled RNA to have approximately 50-75 µg of RNA per tube in 86 µL of ddH₂O.
   2. Add 4 µL of freshly prepared 500 mM sodium meta-periodate (NaIO₄) stock solution (corresponding to 10 mM NaIO₄ for 100 µL reaction volume). For 3'-end single labeling, add 8 µL of NaIO₄ stock solution (corresponding to 20 mM NaIO₄ for 100 µL reaction volume). Mix thoroughly and incubate for 2 h at 25 °C while shaking at 500 rpm under dark conditions as NaIO₄ is light-sensitive.
      NOTE: Do not change the order of addition or further increase the incubation time, as this may cause photobleaching of the attached donor dye. The concentration of NaIO₄ used is decreased by half for dual-end labeling to minimize the quenching of the already attached dye at the 5'-end.
   3. Stop the reaction by adding 30 µL of 50% glycerol. Incubate for 30 min at 25 °C while shaking at 500 rpm.
      NOTE: Glycerol serves as a diol to quench the excess of periodate.
   4. Purification round 4: Add 400 µL of ice-cold EtOH-NaOAc precipitation mixture and perform ethanol precipitation.
   5. Resuspend the oxidized RNA pellet in 95 µL of 50 mM NaOAc, pH 6.0.
   6. Hydrazide-dye attachment: Prepare the fluorophore solution by dissolving a few crystals of Cyanine5 hydrazide (Cy5-hydrazide) in DMSO and then diluting with ddH₂O to the concentration of 2 mM. If the fluorophore of choice is water-soluble, prepare the solution in ddH₂O. Add 5 µL of 2 mM Cy5-hydrazide solution to the oxidized RNA. Mix well and incubate for 16 h at 25 °C while shaking at 500 rpm.
      NOTE: The fluorophore should be at least 40-fold excess of the RNA.
   7. Purification round 5: Purify the dual-end-labeled RNA (or 3'-end single labeled RNA) by EtOH precipitation and centrifugal filtration, similar to steps 1.2.6 and 1.2.8, respectively, and elute in ddH₂O.
   8. Calculate the RNA and conjugate dye concentrations by UV-Vis spectroscopy.
   9. Characterize the labeled RNA using analytical gel-based assays, ensemble fluorescence spectroscopy (see the Representative Results section), and/or analytical HPLC as shown elsewhere³.

2. Microfluidic chamber preparation

NOTE: We recommend handling six or eight chambers at a time. Sonication is performed at room temperature unless otherwise indicated. This protocol limits the use of organic solvents, such as acetone, to prevent the solubilization of trace impurities. For alternatives, refer to references^15,16.

1. Cleaning
   1. Drill four holes in the quartz slides with a diamond driller according to the scheme given in Figure 2A to form two channels.
      NOTE: Although they may break more easily during drilling, glass slides are a cost-effective alternative to quartz slides for objective-based TIRF microscopy setups where the coverslip is the imaging surface (Figure 2B). This is not an option for prism-based TIRF setups, where the microscopy slide is the imaging surface due to background fluorescence.
      NOTE: Alternatively, recycle the used microfluidic chambers¹⁶. For this, immerse the used chambers overnight in acetone (wrap the Coplin staining jar in aluminum foil to minimize evaporation) in the fume hood, and afterward sonicate for 5 min. Disassemble the chamber, and discard the coverslips and the stickers. If the disassembly does not work immediately, sonicate the chamber in acetone for 10 min more. Proceed with step 2.1.2.
   2. Place the drilled quartz slides and approximately twice as many coverslips (since they may break easily) in a glass Coplin staining jar with a cover. Rinse 3x with ddH₂O. Sonicate in ddH₂O for 5 min, then rinse 3x with ddH₂O.
   3. Sonicate in 10% laboratory glassware cleaning solution (see Table of Materials) for 30 min at 50 °C. Rinse at least 3x with ddH₂O until the detergent bubbles are gone. Sonicate in ddH₂O for 5 min. Rinse 3x with ddH₂O.
   4. Sonicate in 1 M KOH solution for 30 min, then leave overnight. Rinse 3x with ddH₂O. Sonicate in ddH₂O for 5 min. Rinse 3x with ddH₂O.
      NOTE: Although corrosion due to excessive incubation is a concern previously raised by Chandradoss et al.¹⁵, we recommend this long incubation that enables avoiding Piranha etching.
   5. Dry the slides and coverslips with N₂(g).
   6. Treat the dried slides and coverslips with oxygen plasma cleaner for 30 min according to the manufacturer's instructions.
2. Aminosilanization
   1. Prepare the 3% APTES-EtOH solution by thoroughly mixing 288.5 mL of absolute EtOH, 1.5 mL of ddH₂O, and 9 mL of (3-aminopropyl)triethoxysilane (APTES) in a 500 mL Erlenmeyer flask.
   2. Place the clean slides and coverslips in a Coplin staining jar, immerse in the 3% APTES-EtOH solution, sonicate for 1 min, and incubate for 30 min.
   3. Rinse 3x with absolute EtOH, then 3x with ddH₂O.
   4. Dry the slides and coverslips under N₂(g) flow.
3. Surface passivation and biotinylation
   1. Prepare a humid box by filling up an empty pipette tip box up to half with ddH₂O. Place the slides inside the box, with the side to be treated facing up.
   2. Freshly prepare the bPEG-mPEG mixture in a 1.5 mL sterile-clean microcentrifuge tube by gently mixing 2 mg of biotin-polyethylene glycol-succinimidyl valerate 5000 (biotin-PEG-SVA) and 80 mg of methoxy polyethylene glycol-succinimidyl valerate (mPEG-SVA) in 640 µL of 100 mM sodium bicarbonate (NaHCO₃) buffer, pH 8.3. Centrifuge the bPEG-mPEG mixture at 16,000 × g for 1 min to remove any air bubbles.
   3. Carefully remove the supernatant and add a 30 µL droplet to the center of the slide to cover both channels. Place an additional droplet on one side of the slide to have an extra PEGylated coverslip, as they break easily, and having a spare is useful. Finally, cover the drop with a clean coverslip, close the humid box, and perform the PEGylation overnight under dark conditions.
   4. Thoroughly rinse the PEG-passivated and biotinylated slides and coverslips with ddH₂O. Notice the change in the hydrophobicity of the treated surfaces. Dry under N₂(g) flow.
4. Microfluidic chamber assembly
   1. Cut a double-sided sticker to create the channels. Attach the sticker to the slide, ensuring it covers the area of interest. Carefully place the coverslip on top, aligning it so that the PEGylated surfaces face each other.
   2. Place each assembled chamber in a 50 mL centrifuge tube and fill the tubes with N₂(g), and store at -20 °C for up to 1 month.

Figure 2: Single-molecule FRET-TIRF microscopy. (A) Microfluidic chamber for TIRF imaging. (B) FRET-labeled RNA is encapsulated in a biotinylated phospholipid vesicle and immobilized on a steptavidin-coated glass surface. This keeps the molecule of interest within the evanescent field (gray gradient), which is created by the incident light that is totally reflected at the critical angle in TIRF microscopy. Here, both fluorophores are subsequently excited with ALEX scheme. Abbreviations: FRET = Förster Resonance Energy Transfer; TIRF = Total Internal Reflection Fluorescence; ALEX = alternative laser excitation. Please click here to view a larger version of this figure.

3. Phospholipid vesicle encapsulation

1. Lipid cake preparation
   NOTE: Always handle chloroform under a fume hood.
   1. Using a sterile needle, poke holes in the lid (from the inside out) of sterile-clean 2.0 mL microcentrifuge tubes, as shown in Figure 3A.
   2. Prepare the 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-cap biotinyl (bPE) stock solution by dissolving at least 2 mg of bPE in the appropriate amount of chloroform to a final concentration of 1 mg/mL.
   3. Prepare the 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) stock solution by dissolving at least 10 mg of DMPC in the appropriate amount of chloroform to a final concentration of 10 mg/mL.
   4. Prepare the bPE-DMPC mixture (keeping the 1:99 w/w ratio) by mixing 100 µL of bPE and 990 µL of DMPC stock solutions. Distribute 109 µL of the bPE-DMPC mixture per 2 mL tube.
   5. Place a cardboard cell partition insert of a microtube storage box in a Schlenk flask, serving as a tube holder. Using long tweezers, place the tubes containing the lipid mixture into the tube holder in a 500 mL Schlenk flask (Figure 3B).
      NOTE: Make sure that they are upright and not tilted.
   6. Evaporate the chloroform under a low flow of N₂(g) overnight (or for at least 2 h until the solvent is completely evaporated).
   7. Seal the tubes with Parafilm to cover the holes. Store the lipid cakes at -20 °C for up to 1 month.
2. RNA encapsulation
   1. Prepare the following buffers and solutions by adjusting the K⁺ and Mg²⁺ concentrations according to the biological system of interest.
      1. Prepare the 5x standard buffer (5x SB): 2.5 M KCl, 400 mM MOPS; adjust the pH to 6.9 with KOH, sterile filter, and store at 4 °C in the dark (respectively 1x SB: 500 mM KCl, 80 mM MOPS, pH 6.9).
      2. Prepare the antiblinking buffer (AB): 100 mM MgCl₂, Trolox, 1x SB, sterile filter, and store at 4 °C in the dark for up to 1 week.
         NOTE: Use a spatula tip of Trolox for 10 mL final volume, vortex to mix, and readjust the pH.
   2. Assemble the extruder using 100 nm polycarbonate (PC) membrane and 10 mm polyester (PE) drain disc, equilibrate the syringe and the membranes with AB, and heat to 30 °C (or essentially above the DMPC glass transition temperature of 24 °C).
   3. Mix 5 µL of 1 µM dual-end labeled RNA and 45 µL of AB (to a total volume of 50 µL).
   4. Hydrate the lipid cake with this RNA solution. At 30 °C, mix for 5 min by shaking at 1,400 rpm, then for 15 min at 700 rpm. Centrifuge for 2 min at 13,000 g and carefully transfer the supernatant into a new tube.
   5. Dilute the sample with 250 µL of AB.
   6. Fill the syringe with the RNA-lipid suspension. Extrude 35x at 30 °C on a heating block to encapsulate the dual-end-labeled RNA in phospholipid vesicles of 100 nm diameter.
      NOTE: The size distribution of vesicles after using the extruder can be validated by dynamic light scattering (DLS), as shown in reference¹¹.

Figure 3: Lipid cake preparation. (A) The tube caps are poked to enable solvent evaporation. (B) Tubes containing the lipid mixture are placed in a Schlenk flask, and the chloroform is evaporated to obtain a lipid cake. Please click here to view a larger version of this figure.

4. smFRET-TIRF microscopy

1. Surface immobilization
   1. Prepare the following buffers and solutions.
      1. Prepare the sugar-antiblinking buffer (SAB): 1% D-glucose (w/v) in AB, sterile filter, and store at 4 °C in the dark for up to 1 week.
      2. Prepare the 100x OSS: 100x Oxygen scavenging system (OSS): 3 mg of glucose oxidase (corresponding to 1.7 U), 10 µL of catalase (corresponding to 22 U) in 90 µL of 1x SB, store at 4 °C in the dark for up to 1 week.
      3. Freshly prepare the imaging buffer (IB) by mixing 198 µL of SAB and 2 µL of 100x OSS solution.
   2. Allow the microfluidic chamber to reach room temperature. Flush the chamber 2x with 200 µL of 1x SB.
   3. Fill the chamber with 50 µL of 20 µL/mL streptavidin solution and incubate for 5 min.
   4. Flush the chamber with 200 µL of 1x SB and then with 100 µL of AB.
   5. Immobilize the encapsulated RNA on the surface by adding 75 µL of vesicle suspension and incubate for 10 min.
   6. Flush the chamber with 200 µL of freshly prepared IB and incubate for 5 min.
   7. The chamber is now ready for data acquisition (Figure 2B). To prevent evaporation during long measurements, seal the holes as shown in Figure 2A. If leaks from channel to channel are of concern, seal the unused channel before starting the surface immobilization protocol.
      NOTE: We recommend performing blank measurements as a quality check for the cleanliness of the microfluidic chambers and phospholipid vesicles against fluorescence contaminations.

Results

We present the site-specific single- and dual-fluorescent labeling of the 915-nt RNA of interest, the yeast mitochondrial Sc.ai5γ group II intron, flanked by exonic sequences. The FRET fluorophore pair is positioned at the RNA ends via EDC/NHS activation of the 5'-phosphate and periodate oxidation of the 3'-ribose, followed by respective dye attachments. We then verified the RNA-dye conjugates by fluorescent gel electrophoresis, as presented in Figure 4A. The co-migration of RNA and the fluorophores on the agarose gel confirms the successful labeling. Next, as shown in Figure 4B, ensemble fluorescence spectroscopy was used to characterize the dual-labeled group II intron. Energy transfer, that is, FRET, was observed upon excitation of the donor dye proving the dual labeling of the RNA. Notably, consistent with the distance-dependent nature of FRET, the folding of the group II intron RNA in the presence of metal ions led to an increase in FRET efficiency, as evidenced by the decrease in donor emission (green arrow) and a corresponding increase in acceptor (red arrow) emission. This indicates that this FRET labeling tracks the conformational changes of the ribozyme.

Figure 4: Characterization of the RNA-dye conjugates. (A) Analytical gel-based analysis of single- and dual-fluorescently labeled RNA shows the co-migration of the dyes with the RNA on a 2% agarose gel. Co-localization of the fluorophores in the dual-labeled sample is indicated by the yellow band in the merged image (bottom) of the Cy3 (top, green) and Cy5 (middle, red) channels, visualized under 532 nm and 635 nm illumination, respectively. Lane 1: 5'-sCy3 only labeled RNA, lane 2: 3'-Cy5 only labeled RNA, and lane 3: dual-end (5'-sCy3 and 3'-Cy5) labeled RNA. (B) Ensemble fluorescence spectroscopy confirms dual-labeling. Energy transfer upon donor excitation (λex = 515 nm, λem = 670 nm) verifies that both dyes are successfully coupled to the RNA. The gray curve represents the emission profile of the pre-catalytic RNA, while the black curve demonstrates increased FRET efficiency in the folded group II intron RNA (incubated with 500 mM KCl at 70 °C for 3 min, cooled to 42 °C for 5 min, followed by the addition of 100 mM MgCl₂). This figure was adapted from Ahunbay et al.³ Please click here to view a larger version of this figure.

With the fluorescently dual-labeled wild-type group II intron in hand, we are now positioned to explore its dynamics at the single-molecule level. Once encapsulated in phospholipid vesicles, the labeled RNA is immobilized on a microscopy surface at very low surface densities to achieve single-molecule resolution for smFRET-TIRF. As seen in Figure 5A, several individual molecules can be tracked simultaneously. TIRF microscopy enables the real-time monitoring of FRET efficiency and its changes over time. Figure 5B exemplifies the static and dynamic single-molecule FRET traces of the labeled and encapsulated RNA. A typical dynamic trace exhibits anticorrelation between donor and acceptor signals that fluctuate the FRET efficiency. When the acceptor emission upon donor excitation increases, the donor emission correspondingly decreases, indicating a dynamic change in the inter-dye distance. This anticorrelation suggests conformational changes in the RNA molecule.

Figure 5: Highly dynamic behavior of group II intron RNA revealed by smFRET. The dual-end labeled group II intron RNA is encapsulated in a phospholipid vesicle and immobilized on the surface for imaging with an objective-based TIRF microscope. (A) Merged image of individual labeled RNA molecules exhibiting donor emission (sCy3, green) and acceptor emission (Cy5, red) upon 532 nm excitation. (B) The typical smFRET trajectories of (left) a static RNA molecule where the donor and acceptor intensities do not fluctuate over time and (right) a dynamic RNA molecule, where the donor and acceptor intensities anticorrelate, with FRET efficiencies in yellow. Single-molecules are localized and analyzed using MASH-FRET¹⁷. Direct excitation, bleed-through, and γ-factor corrections are applied. Abbreviations: smFRET = single-molecule Förster Resonance Energy Transfer. Please click here to view a larger version of this figure.

Discussion

FRET at the single-molecule level is unique because it allows the observation and analysis of individual molecules, revealing sample heterogeneity and capturing transient states that can be obscured in ensemble measurements^1,2. Observing individual RNA molecules using smFRET provides high-resolution insights into their folding pathways and dynamics. This protocol describes the chemical dual-end labeling of RNA and its surface immobilization via phospholipid vesicle encapsulation, which together enable following dynamic conformational changes via smFRET-TIRF microscopy.

Studying RNA dynamics is a constantly growing field with the need for new site-specific fluorescent labeling strategies. We label the RNA ends by targeting the 5'-phosphate with carbodiimides and the 3'-ribose sugar with periodate. These approaches have been described earlier (5'-end^18,19 and 3'-end^19,20,21,22) but have not previously been applied to an RNA of similar size to the wild-type Sc.ai5γ group II intron ribozyme, which required optimization. Carbodiimide (e.g., EDC) activation of the 5'-phosphate is reversible. Therefore, imidazole was used to irreversibly react with the O-acylisourea intermediate to form a highly reactive phosphorimidazolide^19,20. However, it is now known that at higher pH, carbodiimides can modify nucleobases, specifically guanines and uracils, which has recently led to their use as structural probing agents^23,24.

To avoid cross-reactivity, purifying the activated RNA from the EDC prior to raising the pH to 7.5 for the dye coupling step is crucial. However, when we introduced a purification step between the activation and the dye attachment²⁵, we obtained very low yields. Analogously, in protein labeling, surface-accessible lysine residues can be activated with carbodiimides. However, instead of imidazole, which prevents the activation reversal, NHS is routinely used. We adopted this strategy too, thus replacing the phosphorimidazolide intermediate with an NHS-phosphate intermediate. This way we achieved pH control as well as an increased labeling density at lower temperatures and shorter incubation times, i.e., 25 °C for 4 h compared to 37 °C for 16 h. Developed for RNA, this 5'-labeling strategy can be applied to any other single-stranded nucleic acid with a 5'-phosphate.

The 5'-phosphate activation and 3'-ribose oxidation were mutually exclusive, because the chemistries are not orthogonal. To overcome this challenge and avoid cross-labeling, we started with the 5'-end, followed by a blocking step to inhibit the activated but not labeled sites before proceeding with the 3'-end labeling. While oxidizing the 3'-diol, excess sodium meta-periodate (NaIO₄) could quench the already attached fluorophore at the 5'-end. Therefore, we reduced the concentration of NaIO₄ used for single labeling from 20 mM to 10 mM.

We recommend working with multiple aliquots in parallel instead of scaling up the reactions. This protocol requires multiple ethanol (EtOH) precipitation steps. When working with several aliquots in parallel, prepare a precipitation mixture (30 mL of 100% absolute EtOH and 1 mL of 3 M NaOAc, pH 5.2). NaCl is not used due to its low solubility in EtOH. Precipitate the RNA with 3.1 vol. of this mixture by overnight incubation at -20 °C, followed by centrifugation. Wash the RNA pellet twice with 500 µL ice-cold 70% EtOH, spin down at 4 °C after each time, and dry under vacuum. EtOH precipitation takes advantage of the RNA's insolubility and the free dyes' solubility in 70% ethanol. Centrifugal filtration effectively removes free dyes due to their significant size difference from RNA and facilitates buffer exchange, eliminating salts. In addition to EtOH precipitation and centrifugal filtration methods, free dyes can also be removed using gel extraction and/or chromatography techniques (e.g., HPLC); however, the scale should be adapted accordingly. Do not vortex long RNAs to resuspend, as this may cause mechanical shearing²⁶. The optimal time to pause the protocol is when the RNA is pelleted. We use DNA low-binding tubes to improve nucleic acid recovery. Although the final reaction volume is 100 µL, 1.5 mL (and not lower volume) tubes are preferred for better purification by EtOH precipitation.

Once the unreacted free dyes were removed by precipitation and centrifugal filtration, we confirmed the labeling by fluorescent gel electrophoresis (Figure 4A), UV-Vis spectroscopy, analytical HPLC³. However, it is important to note that these methods cannot distinguish between an RNA molecule carrying both fluorophores and a mixture of RNAs, each labeled with a single color. Similarly, they cannot be used to determine if an RNA molecule carries multiple fluorophores of the same color. Mass spectrometry cannot be used due to size limitations. Ensemble³ and single-molecule FRET spectroscopies corroborate dual-fluorescent labeling, as shown in Figures 4B and 5B. The stoichiometry of 0.5 (sCy3 to Cy5 ratio of 1:1) in smFRET experiments confirms the equal conjugation of the two fluorophores. One concern was the double labeling at the 3'-end by labeling both aldehydes instead of the proposed cyclization. The lack of species with a stoichiometry of 0.25 (sCy3 to Cy5 ratio of 1:2) in smFRET experiments suggests that the dye attachment sterically hinders and prevents the attachment of a second dye.

With this dual-fluorescent labeling, FRET signal changes can be attributed to structural rearrangements throughout RNA folding and catalysis. To maintain fluorescently labeled RNA within the evanescent field for single-molecule TIRF imaging, encapsulation is preferred over direct surface tethering. This approach involves trapping individual RNA molecules within lipid bilayers of vesicles, creating a controlled environment conducive to observing their dynamic behavior. The described protocol enriches mono-encapsulation, as photobleaching in a single step demonstrates¹¹. To understand RNA folding and function, bridging the in vitro and in vivo gap is essential²⁷. Molecular crowding agents can mimic the conditions inside cells to enhance RNA catalysis by group II introns^7,28. Alternatively, encapsulation creates constrained microenvironments that promote RNA folding²⁹, bringing our understanding of RNA structure and dynamics closer to a realistic cellular context.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
RNA labeling
1.5 mL DNA LoBind tubes	Eppendorf	30108035
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC-HCl or EDC)	Thermo Scientific	11844071	Pierce EDC, No-Weigh Format. Store at  -20 °C
3-morpholinopropane-1-sulfonic acid (MOPS)	Sigma-Aldrich	69947	BioUltra, for molecular biology, ≥99.5% (titration)
Acetic acid (glacial)	Sigma-Aldrich	1.00063	100%, anhydrous for analysis EMSURE ACS,ISO,Reag. Ph Eur
Centrifugal filtration unit	Sartorius	VS0132	Vivaspin 500, MWCO 50.000, PES, 500 μL
Cyanine5 hydrazide (Cy5-hydrazide)	Lumiprobe	23070	5 mg, CAS number 1427705-31-4
Dimethyl sulfoxide (DMSO)	New England Biolabs	12611P	Molecular biology grade
Ethanol	VWR Chemicals	20821.296P	Absolute ≥99,8%
Glycerol	Sigma-Aldrich	G5516	for molecular biology, ≥99.0%
Hydrochloric acid (HCl)	Sigma-Aldrich	1.00317	fuming 37%, for analysis EMSURE ACS,ISO,Reag. Ph Eur
N-hydroxysuccinimide (NHS)	Sigma-Aldrich	130672	98%
Sodium acetate (NaOAc)	Sigma-Aldrich	S8750	anhydrous, ReagentPlus, ≥99.0%
Sodium meta-periodate (NaIO4)	Thermo Fisher Scientific, Life Technologies	20504	Pierce product line
Sulfo-Cyanine3 amine (sCy3-amine)	Lumiprobe	213C0	5 mg, CAS number 2183440-43-7
Tris(hydroxymethyl)aminomethane (Tris)	Biosolve	200923	Molecular biology grade
Chamber preparation
3-aminopropyl)triethoxysilane (APTES)	Sigma-Aldrich	440140	99%
Acetone	Sigma-Aldrich	1.00014	for analysis EMSURE ACS,ISO,Reag. Ph Eur
Biotin-Polyethylene glycol-Succinimidyl valerate (biotin-PEG-SVA, bPEG) and Methoxy polyethylene glycol-Succinimidyl valerate (mPEG-SVA)	Laysan Bio, Inc.		BIO-PEG-SVA, MW 5,000 and MPEG-SVA, MW 5,000 - Combo Kit
Coverslips	Carl Roth	H876.2	24 x 24 mm, 0.13-0.16 mm thickness
Deconex 11 universal	Borer Chemie AG	17416492	Laboratory glassware cleaning solution
Diamond coated core drill bit	Crystalite corporation		1 mm thickness
Diamond driller
Ethanol	VWR Chemicals	20821.296P	Absolute ≥99,8%
Glass Coplin staining jar with cover
Imaging spacer	Grace Bio-Labs	654008	SecureSeal, 8 wells,  9 mm × 0.12 mm
Oxygen plasma cleaner	Zepto One		Diener
Potassium hydroxide (KOH)	Sigma-Aldrich	P9541	for molecular biology, ≥99.0%
Quartz slides	G. Finkenbeiner, Inc.		7.5 x 2.5 cm, 1 mm thickness
Sodium bicarbonate (NaHCO3)	Sigma-Aldrich	S6297	BioXtra, 99.5-100.5%
Lipid cake preparation
16:0 Biotinyl Cap PE, bPE (1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl) (sodium salt), C53H98N4O11PNaS)	Avanti Polar Lipids	870277P	Stable for 1 year at -20 °C
14:0 PC, DMPC (1,2-ditetradecanoyl-sn-glycero-3-phosphocholine or 1,2-dimyristoyl-sn-glycero-3-phosphocholine, C36H72NO8P)	Avanti Polar Lipids	850345P
2.0 mL microcentrifuge tubes, Safe-Lock	Eppendorf	0030120094	Autoclave to sterilize
500 mL Schlenk flask
Chloroform	Merck	1.02445	for analysis EMSURE ACS, ISO, Reag. Ph Eur
Parafilm
Surface immobilization via encapsulation
(±)-6-Hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox®)	Thermo Fischer	53188-07-1	97%
3-morpholinopropane-1-sulfonic acid (MOPS)	Sigma-Aldrich	69947	BioUltra, for molecular biology, ≥99.5% (titration). Store at +4 °C in the dark.
Adhesive seal tabs	Grace Bio-Labs	629200
Catalase	Sigma-Aldrich	9001-05-2, C30	from bovine liver, aqueous suspension, 10,000-40,000 units/mg protein
D-glucose	Sigma-Aldrich	G7528	≥99.5% (GC), BioXtra
Extruder polycarbonate (PC) membrane	Avanti Polar Lipids	610005-1EA	0.1 μm, 19 mm
Extruder set with heating block	Avanti Polar Lipids	610000
Glucose oxidase	Sigma-Aldrich	9001-37-0	from Aspergillus niger, Type VII,  lyophilized powder, ≥100,000 units/g solid (without added oxygen)
Magnesium chloride (MgCl2)	Sigma-Aldrich	7786-30-3, M1028	for molecular biology, 1.00 M ± 0.01 M solution
Polyester (PE) drain disc, membrane	Whatman, Cytiva	230300	10 mm
Potassium chloride (KCl)	Sigma-Aldrich	60128	BioUltra, for molecular biology, ≥99.5% (AT)
Potassium hydroxide (KOH)	Sigma-Aldrich	P9541	for molecular biology, ≥99.0%
Streptavidin	Thermo Fischer	434301