Fluorescent Molecular Beacons Mimicking RNA Secondary Structures to Study RNA Chaperone Activity

Molecular beacons (MBs) are oligonucleotide probes with a hairpin-like structure that are typically labelled at the 5 0 and 3 0 ends with a ﬂuorophore and a quencher dye, respectively. The conformation of the MB acts as a switch for ﬂuorescence emission. When the ﬂuorophore is in close proximity to the quencher, ﬂuorescence emission cannot be detected, meaning that the switch is in an OFF state. However, if the MB structure is modiﬁed, separating the ﬂuorophore from the quencher, the switch turns ON allowing ﬂuorescence emission. This property has been extensively used for a wide variety of applications including real-time PCR reactions, study of protein-DNA interactions, and identiﬁcation of conformational changes in RNA structures. Here, we describe a protocol based on the MB technology to measure the RNA unfolding capacities of the CspA RNA chaperone from Staphylococcus aureus. This method, with slight variations, may also be applied for testing the activity of other RNA chaperones, RNA helicases, or ribonucleases.


Introduction
Molecular beacons (MBs) are oligonucleotide probes commonly used to target DNA for real-time monitoring of polymerase chain reactions (RT-PCRs). The central nucleotides of the MB are complementary to a specific DNA (or RNA) target and do not base pair with one another, while the five to seven nucleotides at each terminus are complementary to each other, creating a hairpin-like conformation (Fig. 1a). Since the 5 0 and 3 0 ends are labelled with a fluorophore and a quencher, respectively, the MB acts as a switch. In their native conformation, the extremes are close enough for the quencher to prevent fluorescent emission from the fluorophore (switch OFF). When the MB hybridizes with its specific target, its native structure is disrupted, and both dye molecules fall apart from each other, allowing fluorescence emission (switch ON) (Fig. 1a). Since MBs tolerate very versatile designs, they have been used for various applications [1,2]. Molecular biologists have taken advantage of their potential for studying different mechanisms such as protein-DNA interactions [3], single-stranded DNA cleavage by specific nucleases (Fig. 1b) [4], and structural changes on ribozymes and riboswitches (Fig. 1c) [5,6]. In this last case, RNA conformational changes have been determined by the use of MBs that target specific RNA regions that become free for hybridization. This usually occurs after binding of the metabolite, which induces the subsequent RNA structural change on the ribozyme or riboswitch (Fig. 1c) [5,6]. Thus, only when the MB is bound to its RNA target, the probe structure unfolds and becomes fluorescent. On the other hand, in order to study RNA chaperone activity, a more direct approach by using a MB that mimics the regulatory RNA hairpin targeted by a specific RNA-binding protein (RBP) has been adopted [7,8]. This strategy assumes that binding of the RNA chaperone to the MB may cause a similar RNA conformational rearrangement to the one occurring on the native RNA. Therefore, the MB may act as a direct reporter of its own structural rearrangement (Fig. 1d). We choose this approach to demonstrate that the Fig. 1 Examples of different MB designs dedicated to (a) quantifying specific DNA or RNA molecules, (b) analyzing the single-stranded DNA cleavage by specific nucleases [4], (c) studying the structural changes on ribozymes and riboswitches [5,6], or (d) determining the RNA chaperone activity on hairpin-like structures by cold shock proteins (CSPs) [7][8][9]. In all cases, the MB switch turns on when the fluorophore (FAM) folds away of the quencher (BHQ_1) due to the base-pairing of the MB with its specific target (a, c) or to the MB cleavage or unfolding by the activity of an RNA-binding protein (b, d) RNA chaperone CspA of Staphylococcus aureus unfolds the RNA hairpin present in the 5 0 UTR of its own mRNA [9] (Fig. 2a). This hairpin (ΔG ¼ À24.60 kcal/mol) is cleaved by endoribonuclease III (RNase III) mainly at position G-53, generating a shorter cspA mRNA version that is more efficiently translated than the unprocessed mRNA [10]. CspA would repress its own expression by unfolding the hairpin and thus antagonize the function of RNase III [9].
Specifically, we designed a MB that comprised a 49-mer ssDNA oligonucleotide, which included the central functional sequence of the cspA 5 0 UTR hairpin (ΔG ¼ À13.70 kcal/mol). A molecule of fluorescein (FAM) and a Black Hole Quencher (BHQ_1) were attached to the 5 0 and 3 0 ends, respectively (Fig. 2b). In the native MB conformation, BHQ_1 efficiently quenched the fluorescence from FAM, indicating that the designed MB accurately mimicked the cspA 5 0 UTR hairpin. In contrast, when the MB structure was disrupted (separating FAM from BHQ_1) either by the presence of the RNA chaperone CspA or by an increase in the temperature of incubation, fluorescence emission was registered. The folded Fig. 2 Molecular beacon (MB) design to study the CspA RNA chaperone activity on the RNA hairpin structure of the cspA mRNA [9]. (a) The proposed RNA structure for the 5 0 UTR of cspA is shown [10]. The U-rich motif required for CspA interaction is highlighted in red. (b) The MB consisted of a 49-mer ssDNA oligonucleotide labelled with the FAM fluorophore and the BHQ_1 quencher at its 5 0 and 3 0 ends, respectively [9] conformation of the MB could be efficiently restored (indicated by the ceasing of fluorescence emission) either by adding Proteinase K, which eliminated the chaperone activity by degrading CspA, or by decreasing the temperature of incubation. The specificity of CspA on the designed MB system was verified by the incubation of the MB with an unrelated protein.
This strategy allowed us to demonstrate that CspA unfolded the regulatory hairpin located at the cspA 5 0 UTR and, thus, interfered with cspA mRNA processing by RNase III. When CspA levels were low, the cspA 5 0 UTR RNA hairpin was targeted and cleaved by RNase III. The resulting processed mRNA suffered a conformational change that favored CspA translation [10]. When CspA levels rose, CspA decreased its own expression by unfolding the cspA 5 0 UTR RNA hairpin to avoid RNase III cleavage [9] (Fig. 3).  3 Schematic representation of the putative auto-regulatory mechanism modulating CspA expression as previously described by Caballero and colleagues [9]. The 5 0 UTR of the cspA mRNA forms a hairpin structure that is cleaved by RNase III to enhance CspA translation when CspA levels are low [10]. When the concentration of CspA inside the cell is high, the protein is able to interact with the hairpin structure through a U-rich motif and unfold it. As a consequence, the cspA mRNA is not processed by RNase III and CspA translation is decreased Here, we describe in detail the different steps that should be followed to determine the RNA folding rearrangements caused by the binding of any RBP by using a MB that mimics a natural target (whose synthesis could be ordered from a regular oligonucleotide supplier company). The protocol requires commonly available equipment at a molecular biology research center. It is noteworthy that, with slight modifications, this protocol may be adapted to test (1) any DNA or RNA folding structure that allows close proximity of BHQ_1 to FAM and that provides enough separation between them when disrupted; (2) the activity of RBPs such as RNA chaperones, RNA helicases and ribonucleases that target and/or process hairpin-like structures; and (3) the function of small regulatory RNAs that produce conformational changes on hairpin-like structures of their mRNA targets.

Materials
Prepare all solutions using ultrapure water (prepared by purifying deionized water to reach a sensitivity of at least 18 MΩ at 25 C) and analytical grade reagents for use in molecular biology. Store solutions at room temperature unless stated otherwise. Follow safety and waste disposal regulations when handling harmful products accordingly. 2. Sterile material for bacterial growth: 10-, 100-and 1000-μL pipette tips, test tubes, 2-L Erlenmeyer flasks, graduated cylinders, 250-mL centrifuge tubes, petri dishes, 1.5-mL Eppendorf tubes.
Store at À20 C.
4. 10 mg/mL RNase A stock solution. Store at À20 C.
8. Centrifuge with a rotor for 50-mL tubes, which allows centrifugation at 16,000 Â g.
10. 5 mg/mL DNase I stock solution prepared by dissolving DNase I powder in 0.15 M NaCl.
15. Protein molecular weight marker. Store at À20 C. 3. Inoculate 500 μL of the bacterial preculture (1/1000 dilution factor) into two sterile pre-warmed 2-L Erlenmeyer flasks containing 500 mL of LB medium supplemented with 100 μg/mL ampicillin and 1% glucose. Mix and incubate the cultures at 37 C and 200 r.p.m. until an optical density (OD 600nm ) of 0.5 is reached.
4. Induce the expression of CspA by addition of IPTG to a final concentration of 0.4 mM. Save 1 mL of culture of one of the flasks and centrifuge it for 3 min at 18,000 Â g. Store the bacterial pellet at À20 C. This aliquot sample corresponds to the pre-induction control (see Note 9). Resume bacterial growth for another 5 h at 37 C and 200 r.p.m.
5. Save 1 mL of culture of one of the flasks and centrifuge it for 3 min at 18,000 Â g. Store the bacterial pellet at À20 C (postinduction control) (see Note 9). Harvest the rest of the cultures in 250-mL tubes and centrifuge for 10 min at 5000 Â g (see Note 10). Discard the supernatant and resuspend the pellets in 1 volume of PBS pH 7.3. Repeat the centrifugation step, discard the supernatant and store the bacterial pellets at À80 C (see Note 11). 3. Centrifuge the samples at 16,000 Â g for 30 min at 4 C (see Note 10). Transfer the supernatant (soluble fraction) to new tubes and store the pellet at À20 C. Pellets (insoluble fraction) contain inclusion bodies (IB), and constitute the IB control (see Note 9).
4. Supplement the soluble fraction with DNase I and RNase A to a final concentration of 10 μg/mL and 5 μg/mL, respectively, and incubate on ice for 30 min. Store 50 μL of the sample at À20 C (pre-filtered soluble fraction control) (see Note 9).
5. Filter the soluble fraction using a 0.45 μm filter whilst on ice (see Note 12). Store 50 μL of the sample at À20 C (postfiltered soluble fraction control) and the rest of the soluble fraction at À20 C (see Note 9). 2. Clean the system with 20% ethanol and ultrapure water.
3. Connect the column to the AKTAprime plus system "drop to drop" to avoid introducing air into the column. Equilibrate the column with 25 mL of binding buffer at a flow rate of 5 mL/ min.

4.
Apply the sample at a flow rate of 0.2 mL/min (see Note 16). 14. Clean the column with 480 mL of ultrapure water and 480 mL of 20% ethanol at a flow rate of 1 mL/min. Remove the column from the system and clean the system with ultrapure water and 20% ethanol.
15. To select fractions containing CspA, mix an aliquot of each peak fraction with sample buffer 6Â and perform a 12% PAGE as described above.

Load the CspA selected fractions into a Slide-A-Lyzer Dialysis
Cassette and dialyze against CspA Storage buffer overnight at 4 C.
17. To assess protein purity, mix an aliquot of the recombinant CspA chaperone with sample buffer 6Â and perform a 12% SDS-polyacrylamide gel electrophoresis (PAGE) as described above.
18. Determine the recombinant protein concentration by the Bio-Rad protein assay.
3.2 Assessment of the RNA Chaperone Activity with a Molecular Beacon

Molecular Beacon Design
The success of this assay lies in an adequate MB design, which is based on two main principles: (1) the presence of an RNA structure targeted by the RNA chaperone under study and (2) fluorescence quenching exerted by a quencher dye (e.g., BHQ_1) on a fluorophore (e.g., FAM), which occurs when both molecules are in close proximity to one another. Additionally, the selected RNA structure must keep the quencher close enough to the fluorophore at the working temperature (switch OFF). MB mimicking hairpin-like structures have been shown to comply these criteria before [7][8][9] Figure 4 shows an example of the results obtained with the MB designed for the analysis of S. aureus CspA activity [9] (Fig. 4). In this example, when the MB was incubated at 55 C and 65 C, fluorescence emission was registered, indicating that the MB was in an ON state. These fluorescence levels were directly proportional to the MB concentration. In contrast, when the MB was incubated at 37 C and 45 C, the fluorescence values were close to those of the background confirming that the MB was in an OFF configuration. This experiment validated the functionality of the designed MB (see Note 24). 2. Program the AriaMx thermal cycler to incubate the MB samples as follows: 37 C, 5 min; PAUSE, 37 C, 15 min; PAUSE, 37 C, 30 min; 65 C, 10 min; STOP (Table 2). Register the fluorescence emission every minute (see Note 25).
3. Prepare an optical 96-well plate including the reaction mixes as indicated in Table 2 (see Note 26). Note that the CspA and BSA proteins should be added later.
4. Seal the plate with adhesive film (see Note 5) and load it into the thermal cycler. Start the incubation program.
5. At the first pause of the incubation program, pull out the 96-well plate from the thermal cycler, remove the adhesive film and add the appropriate quantity of CspA and BSA. Re-seal the plate with a new adhesive film. This step must be performed swiftly.
6. Reintroduce the plate into the thermal cycler and continue the incubation at 37 C during 15 min. Register the fluorescence emission every minute.
7. During the second incubation pause, pull out the plate, remove the adhesive film and add 10 μL of proteinase K (20 mg/mL). Re-seal the plate with a new adhesive film. This step must be performed swiftly.
8. Reintroduce the plate into the thermal cycler and continue the incubation for 30 min at 37 C and then increase the temperature up to 65 C during 10 min. Register fluorescence emission Fig. 4 Test of the molecular beacon functionality. Different concentrations of the MB mimicking the hairpin structure located at the 5 0 UTR of the cspA mRNA were incubated at different temperatures and fluorescence emission was registered.
The experiment was carried out using the AriaMx thermal cycler  (Fig. 2).
4. We used bovine serum albumin (BSA), a protein without capacity to bind nucleic acids, as a negative control. Any alternative protein lacking DNA/RNA binding domains can also be used. 5. We preferred to seal the 96-well plates with adhesive film because removing it and re-sealing the plates is faster than using flat caps. A quick sealing helps registering fluorescence emission sooner, after the RNA chaperone is added to the MB solution.
6. Storage of the 10Â reaction buffer will require it to be prepared without DTT. DTT should be added just before use.  10. Centrifuge should be pre-cooled before use.
11. Bacterial pellets can be stored at À80 C for several days.
12. We recommend the use of filters with a pore size of 0.45 μm instead of 0.2 μm to avoid filter saturation.
13. Pre-cast or custom-made gels may be used with the appropriate percentage of acrylamide according to the protein of interest (we used 12% PAGE).
14. Adjust voltage of the electrophoresis system accordingly.
15. If the protein of interest is not in the soluble fraction, bacterial growing conditions should be modified to force its solubilization. Alternatively, protein purification methods from inclusion bodies may be applied.
16. Due to the slow binding kinetics between GST and glutathione, it is very important to keep the flow rate as low as possible during sample application for maximum binding capacity.
17. The column used in this protocol is specific for separating proteins with a small size. If the RNA chaperone of interest has a bigger size, the column should be changed accordingly.
18. Some oligonucleotide supplier companies limit the synthesis of labelled oligonucleotide probes to 50 nucleotides (nt). In our design, the functional RNA hairpin region could be included in an oligonucleotide probe smaller than 50 nt. For larger regulatory structures, the synthesis of a MB may prove more challenging. This problem could be solved by dividing the MB synthesis into two shorter oligomers that can afterwards be ligated as previously described [7].
19. The reason for using a labelled DNA oligonucleotide as a MB is that it has been proven that CSPs can bind ssDNA as efficiently as RNA molecules [11]. Nevertheless, testing other RNA-binding proteins may require synthesis of RNA-based MBs.
20. If the region of the RNA structure under study is not strong enough to maintain the MB beacon in an OFF state, the basal level of fluorescence might be too high to obtain reliable results once the RNA chaperone is added.
21. Sometimes the quantity or the concentration of the chaperone under study can be limited. We recommend using the lowest concentration of the MB that gives good fluorescent levels in an ON state. This will help saving RNA chaperone sample.
22. The fluorophores of the MB are sensitive to the light; therefore, keep the stock and any other dilutions wrapped in aluminum foil and protect them from exposure to light to maintain their integrity.
23. The selected temperature might vary depending on the melting temperature of the MB structure.
24. If the control of the MB functionality does not show clear differences on the fluorescence signals between ON and OFF states, and/or the fluorescence background is too high, the MB should be redesigned.
25. The entire incubation protocol to be carried out with the AriaMx thermal cycler (or any equivalent equipment) can be programed from the beginning, including the corresponding pause times required to add the different components of the reactions.
26. Volumes of each reactive should be adjusted according to the concentration of the RNA chaperone.