Simultaneous probing of transcription, G-quadruplex, and R-loop

Abstract

DNA and RNA can form various non-canonical secondary structures, including G-quadruplex (G4) and R-loops. These structures are considered transcriptional regulatory elements due to their enrichment at regulatory regions. During transcription, G-rich sequences in the non-template strand promote R-loop formation in the DNA template strand. These R-loops induce G4 structures in the non-template DNA strand, further stabilizing them. Additionally, the high rG: dC base-pairing within the R-loop contributes to the stability of DNA/RNA hybridization. Our previous study investigated the interplay between G4s and R-loops and its impact on transcription. We employed two techniques to demonstrate transcription-mediated G4 and R-loop formation. The single-molecule method allows us to detect intricate details of transcription initiation, elongation, and co-transcriptional R-loop and G4 formation. It provides a high-resolution view of the dynamic processes involved in transcriptional regulation. As an orthogonal approach, a gel-based assay enables the detection of the transcription-mediated R-loops and the RNA product. We can measure the progressive formation of R-loop and total RNA produced from transcription by analyzing gel electrophoresis patterns. In summary, these techniques provide valuable insights into the non-canonical nucleic acid structures and their impact on gene expression.

1. Introduction

Accurate control of gene expression is essential for the proper development of organisms. Transcription is the first step in gene expression, which involves copying a gene’s DNA sequence to make RNA (Britten and Davidson 1969, Lockhart and Winzeler 2000). Transcription is a unidirectional multistep process encompassing thousands of unique cycles within each cell (Heyduk and Heyduk 2018, Vervoort, Devlin et al. 2022). The transcription cycle consists of pre-initiation, initiation, pausing, elongation, termination, and recycling (Cossa, Parua et al. 2021). These stages are tightly regulated by transcriptional cyclin-dependent kinases and their cognate cyclins (Dousis, Ravichandran et al. 2023). Even though the transcription mechanism is shared in almost all organisms, the transcription details of each stage can vary due to the presence of different RNA polymerases. For example, RNA polymerase is structurally more complicated in eukaryotes than in prokaryotes, leading to a higher complexity of interactions with transcription factors, enhancers, and mediators that work together to activate or repress gene expression. (Lemon and Tjian 2000, Sutton and Walker 2001, Cramer, Armache et al. 2008).

1.1. T7 RNAP transcription cycle

In vitro transcription is typically performed using bacteriophage T7 RNA polymerase (T7 RNAP) because it produces full-length RNA transcripts as a holo-enzyme with high fidelity (Dousis, Ravichandran et al. 2023). T7 RNAP-mediated transcription encompasses three distinct stages: initiation, elongation, and termination (Huang and Sousa 2000). T7 RNAP binds the T7 promoter, forming an initiation complex, which undergoes several conformational changes to open the transcription bubble (Yin and Steitz 2002). Due to the instability of the initiation complex, short RNAs of 2–13 nucleotides are generated through an abortive transcription cycle (Jia and Patel 1997, Koh, Roy et al. 2018). When the RNA length exceeds ~13 nucleotides, the RNAP rearranges the DNA complex and escapes from the promoter region, forming a stable and processive enzyme elongation complex (Jia and Patel 1997). Termination occurs either in response to specific signal sequences or upon reaching the end of a DNA template, which typically yields full-length RNA (Proudfoot 2016). However, the transcription process is either greatly enhanced or suppressed due to the presence of special DNA sequences at transcription regulatory regions either on the template or the non-template strand (Wang and Vasquez 2017). In particular, GQ formation in the non-template DNA significantly promotes transcription (Lee, McNerney et al. 2020, Hwang, Palmer et al. 2024, Lee, Joshi et al. 2024).

1.2. G-quadruplex (G4) and R-loop

Guanine-rich nucleic acid sequences have a strong propensity to form four-stranded G-quadruplex (G4) structures under physiological conditions. The stability of G4 depends on both the loop length and the coordinated bonding with a monovalent cation (Tippana, Xiao et al. 2014, Paul, Opresko et al. 2022). G4s can fold into various conformations in vitro, and G4 structures are present in living cells (Spiegel, Adhikari et al. 2020, Johnson, Paul et al. 2024). Bioinformatics studies predict over 400,000 G4 clusters enriched at important genomic regions near the transcription start site (TSS) at oncogene promoters (Huppert and Balasubramanian 2007). Many small molecules target potential G-quadruplex forming sequence (PQS) at the promoter region of many oncogenes, suggesting its therapeutic potential (Müller and Rodriguez 2014).

An R-loop is a DNA:RNA hybrid consisting of three-stranded nucleic acid structures observed in vivo and in vitro. R-loops can form co-transcriptionally when a nascent RNA transcript anneals to the template strand of DNA (Tan-Wong, Dhir et al. 2019). Additionally, an R-loop can form in trans when a transcript hybridizes to a DNA strand in trans (Petermann, Lan et al. 2022). R-loops occur frequently in the genome, significantly impacting gene expression, DNA replication, and histone modifications (Niehrs and Luke 2020). Several techniques are available to identify and characterize R-loop structures (Duquette, Handa et al. 2004, Lee, McNerney et al. 2020). The monoclonal S9.6 antibodies are extensively used to detect and isolate R-loop (Bou-Nader, Bothra et al. 2022). RNase H enzymes are primary proteins that selectively digest the RNA strand of the R-loop, which allows the two complementary DNA strands to re-anneal (Lee, McNerney et al. 2020, Hwang, Palmer et al. 2024).

1.3. G4 and R-loop during the kinetic cycle of transcription

G4 and R-loop structures can synergistically regulate transcription and are enriched near the transcription start site (Kim 2019). R-loops can form co-transcriptionally at the DNA template strand with high GC content or GC skew (Tan-Wong, Dhir et al. 2019, Lee, McNerney et al. 2020). When the R-loop forms on the template strand, the displaced G-rich non-template strand can fold into G4. Previously, we demonstrated that such co-transcriptionally induced R-loops and G4s enhance transcription (Lee, McNerney et al. 2020). In contrast, the transcriptional yield is significantly reduced when PQS is present in the template strand, reflecting the importance of PQS orientation on transcription. Here, we showcase the DNA constructs that harbor PQS in the non-template strand and measure the transcription, R-loop and G4 formation using single-molecule FRET and gel-based assays.

2. Reagents

• T7 RNA Polymerase (NEB, M0251S)

• RNA Pol Reaction Buffer (NEB, B9012SVIAL)

• Potassium Chloride (Sigma-Aldrich, P3911)

• RNase Inhibitor, Murine (NEB, M0314S)

• Ribonucleotide Solution Mix (rNTP) (NEB, N0466S)

• RNase H (NEB, M0297S)

• Sodium bicarbonate (Sigma-Aldrich, #S5761)

• Sodium chloride (Sigma-Aldrich, #S3014)

• 1M Tris-HCl (pH8.5) buffer (Fisher scientific, #T1085)

• Biotin-PEG-SVA, MW 5,000 (Laysan Bio Inc., #Biotin-PEG-SVA5000–100 mg)

• Catalase from bovine liver (Sigma-Aldrich, #C3155)

• Glucose oxidase from Aspergillusniger (Sigma-Aldrich, #G2133)

• mPEG-Succinimidyl Valerate, MW 5,000 (Laysan Bio Inc., #MPEG-SVA-5000–1g)

• N-(2-Aminoethyl)-3-aminopropyl trimethoxy silane (United Chemical Technologies, #1760-24-3)

• NeutrAvidin (Thermo Scientific, #31000)

• Trolox(R), 97%, ACROS Organics^™ (Fisher Scientific, #AC218940010)

• ETT PTFE Tubing Size 26,100 ft, Natural Color (Weico Wire, #ETT-26)

• 0.5 M EDTA pH 8.0 (ThermoFisher Scientific, AM9260G)

• SDS 10% (VWR, 97062–964)

• RNase A, DNase and protease-free (ThermoFisher Scientific, EN0531)

• DNaseI I (NEB, M0303S)

• 40% Acrylamide/Bis Solution, 29:1 (Bio-rad, 1610156)

• 10x TBE Buffer (Bio-rad, 1610770)

• Ammonium persulfate (A3678, Millipore Sigma)

• TEMED (1610800, Bio-rad)

• Low molecular weight DNA ladder (NEB, N3233S)

• Gel loading dye, Purple (6X), no SDS (NEB, B7025S)

• SYBR Green II RNA Gel Stain, 10,000X concentrate in DMSO (Thermofisher, S7564)

3. Materials and equipment

3.1. DNA constructs

PIFE1-Top	TGGCGACGGCAGCGAGGCTAAATTAATACGACTCAC/Cy3/ATAGGGAGACCACAACGTTAGGGTGGGTAGGGTGGGTTATCAGCTCCAGGTCT
PIFE1-Bottom	AGACCTGGAGCTGATAACCCACCCTACCCACCCAACGTTGTGGTCTCCCTATAGTGAGTCGTATTAATTTA
PIFE2-Top	TGGCGACGGCAGCGAGGCTAAATTAATACGACTCAC/Cy3/ATAGGGAGACCACAACGTTAGGGTGGGTAGGGTGGGTTATCAGCTCCAGGTCT
PIFE2-Bottom	AGACCTGGAGCTGAT/Cy5/AACCCACCCTACCCACCCAACGTTGTGGTCTCCCTATAGTGAGTCGTATTAATTTA
FRET1-Top	TGGCGACGGCAGCGAGGCTAAATTAATACGACTCAC/Cy3/ATAGGGAGACCACAACGTTAGGGTGGGTAGGGTGGGTTATCAGCTCCAGGTCT
FRET1-Bottom	AGACCTGGAGCTGATAACCCACCCTACCCACCC/Cy5/AACGTTGTGGTCTCCCTATAGTGAGTCGTATTAATTTA
FRET3-Top	TGGCGACGGCAGCGAGGCTAAATTAATACGACTCACTATAGGGAGACCACAACG/iCy3/TAGGGTGGGTAGGGTGGGT/Cy5/ATCAGCTCCAGGTCT
FRET3-Bottom	AGACCTGGAGCTGATAACCCACCCTACCCACCCAACGTTGTGGTCTCCCTATAGTGAGTCGTATTAATTTA
Biotin-18 mer	GCCTCGCTGCCGTCGCCA-Biotin
T7 promoter-top	GAA ATT AAT ACG ACT CAC TAT A
T7 promoter-bottom	TAT AGT GAG TCG TAT TAA TTT C

3.2. Equipment

• Thermal cycler (Bio-Rad, C1000)

• Syringe pump (HARVARD APPARATUS, PHD2000)

• Spectrophotometer (Thermo Scientific, NANODROP 2000)

• Home-built Total Internal Reflection Fluorescence (TIRF) microscope set-up

• Mini-PROTEAN Tetra Vertical Electrophoresis System for 1.0 mm gels (Bio-rad, 1658001FC)

• PowerPac HC Power Supply (Bio-rad 1645052)

3.3. DNA sample preparation

All the synthesized DNA oligonucleotides containing either biotin, Cy3 and Cy5, or amine labels are purchased from Integrated DNA Technology (IDT, USA). Amine-modified oligonucleotides are labeled by ethanol precipitation. Briefly, 30 μL of 100 mM amine-modified ssDNA is mixed and incubated overnight at room temperature with 0.2 mg of NHS ester-conjugated Cy5 dye (Cytiva PA15100) in 100 mM NaHCO₃, pH 8.5. The excess dyes are removed via three times of ethanol precipitation (Lee, Sanford et al. 2020, Badiee, Kenet et al. 2023). Biotin-tagged 18-mer is used for surface immobilization for Total Internal Reflection Fluorescence (TIRF) measurements. DNA duplex constructs are prepared by mixing top, bottom and 18-mer oligomers in 1:1.2:1.5 in 10 mM Tris buffer (pH 8.0) with 5 mM MgCl₂ and annealed them in a thermocycler with the following protocol: 95°C for 2 min; gradual cooling to 40°C at the rate of 2°C/min; further cooling by 5°C/min until 4°C (Paul, Ha et al. 2021). The annealed DNA is stored in a −20 °C freezer until further use. We indicate that PIFE1 DNA constructs are a mixture of PIFE1-top, PIFE1-bottom, and biotin 18-mer. The same nomenclature applies for PIFE2, FRET1, and FRET3 constructs.

3.4. Transcription assay

For single-molecule assays, transcription is carried out in a total volume of 50 μL.

• Mix the desired concentration of T7 RNAP (10 to 1000 nM) and rNTP (0.01 to 1 mM) with oxygen-scavenging imaging buffer and transcription buffer containing 40 mM Tris–HCl (pH 8.0), 50 mM KCl, 6 mM MgCl₂, 2 mM spermidine, and 1 mM dithiothreitol (DTT).

• Add the reaction mixture to the immobilized DNA.

• Take transcription measurements at different time intervals by collecting short and long movies.

• Add RNase H (final concentration: 0.125 U/μL) at the end of each measurement to observe R-loop digestion.

For gel-based assays, transcription is carried out in a total volume of 20 μL.

• Prepare the transcription buffer (40 mM Tris–HCl (pH 8.0), 50 mM KCl, 6 mM MgCl₂, 2 mM spermidine, and 1 mM DTT), RNase Inhibitor (0.2 units/μL), 10 nM of labeled DNA, and T7 RNA Polymerase (1.25 units/uL). Aliquot 19.2 μL of the master mix to PCR tubes.

• Initiate transcription by adding 0.8 μL of ribonucleotide solution mix (final concentration is 1 mM, 25 mM of stock rNTP).

• Perform transcription at 25°C for up to 1 hour.

• Add 0.5 μL of 0.5 M EDTA to terminate transcription.

• To dissociate T7 polymerase from the DNA, prepare a solution that consists of 50% glycerol and 10% SDS in a ratio of 4 μL to 0.2 μL. Add 4μL of the solution to each sample before loading samples into PAGE gel. Alternatively, to observe T7 polymerase-bound complex, add 4 μL of 50% glycerol to each sample before loading into PAGE Gel.

4. Detection of G4 and R-loop during transcription via single-molecule assay

4.1. smFRET narration

The steps below describe a method for measuring the transcription activity and transcription mediated R-loop and G4 formation at the single-molecule level. The detailed step-by-step protocol for smFRET and data analysis has previously been published (Roy, Hohng et al. 2008, Paul, Liou et al. 2021, Paul and Myong 2022).

PEGylated slide preparation and slide assembly:

• Clean pre-drilled quartz slide and coverslips thoroughly with soap, Milli-Q water, methanol, acetone. Etch each slide with 1 M potassium hydroxide.

• Burn the slides for 1–2 minutes and pass the coverslips 4 to 5 times through the flames to eliminate any source of fluorescence.

• Treat both the slides and the coverslips with aminosilane for 30–45 minutes. Rinse the slides and the coverslips with Milli-Q water, and dry with nitrogen gas.

• Perform PEGylation by incubating the slides and coverslips overnight in a mixture of 2% biotinylated PEG (biotin-PEG-5000, Laysan Bio, Inc) and 98% methoxy-polyethylene glycol (mPEG-5000, Laysan Bio, Inc.) in 100 mM bicarbonate buffer.

• Wash slides and coverslips with Milli-Q water and dry with nitrogen gas. The passivated slides are ready for immediate use. For later use, pack each slide in a vacuum sealed 50 mL Corning tube and store at −20 °C.

• Create the microfluidic sample chamber with double-sided tape by sandwiching the plasma cleaned slide and the coverslip, both coated with PEG and biotinPEG.

• Seal the edges with quick-drying epoxy glue.

• For real-time detection of R-loop and G4 formation during transcription, a reservoir is built by attaching a cut-piece of a 200 μL pipette tip to one hole and the epoxy mold to the corresponding hole for syringe attachment (Paul and Myong 2022).

Single-molecule FRET assay:

• Use an imaging buffer containing 2 mM trolox (dissolved in tris-HCl, pH 8.0), 0.5% glucose, 1mg/mL glucose oxidase and 4 μg/mL catalase, along with transcription buffer of 40 mM Tris–HCl (pH 8.0), 50 mM KCl, 6 mM MgCl₂, 2 mM spermidine, and 1 mM dithiothreitol.

• Add desired concentration of T7 RNAP and rNTP to the transcriptional imaging buffer to initiate the transcription reaction.

• Place the assembled slide on the home-built prism-type TIRF microscope and sample chamber charged with 50 μL of NeutrAvidin (50 μg/mL).

• Apply 50 μL of 25 pM annealed oligomers labeled with Cy3, Cy5, and biotin to the NeutrAvidin-coated slide surface for immobilization. Add imaging buffer for smFRET visualization of ~300–500 fluorescent spots in one field of view of 25 ×75 μm area.

• Use a solid-state 532 nm diode laser to excite Cy3 fluorophore while simultaneously collecting the fluorescence from Cy3 and Cy5 using a water immersion Olympus objective (60 X, NA 1.2). EMCCD (DU897, Andor), which detects Cy3 and Cy5 signals with 100 ms frame integration time.

• For real-time detection of transcriptional mediated R-loop and G4 formation, load RNAP and rNTP onto the reservoir and use a syringe to pull the solution through the silicone tubing at a rate of 20 μL/s at certain time points during the smFRET measurement.

• Generate FRET histograms from >4000 molecules (21 frames of 20 short movies) collected from different imaging surfaces. Record long movies to observe the molecular behavior.

• Process all smFRET data with the IDL script (http://www.exelisvis.co.uk/ProductsServices/IDL.aspx) and analyze the data with the MATLAB script (https://www.mathworks.com/).

4.2. T7 RNAP binding rate

We used PIFE1 DNA constructs to check T7 RNAP binding to the promoter (Fig. 1A). T7 promotor binding increases the Cy3 signal due to Protein Induced Fluorescence Enhancement (PIFE), which has high sensitivity within 0–4 nm distance (Hwang, Kim et al. 2011). Upon addition of T7 RNAP to the immobilized Cy3 labeled DNA, we observed an increase in the average Cy3 fluorescence intensity, indicating the binding of RNAP to the promoter. Single-molecule time traces show a series of brief pulses of Cy3 PIFE with the pulse frequency increasing as a function of RNAP concentration (10 to1000 nM) (Fig. 1B). Rapid short pulses of PIFE indicate repetitive binding of RNAP to the promoter. Since PIFE frequency increases linearly with RNAP concentration (Fig. 1C), we set 100 nM RNAP as the optimal concentration for a single-molecule transcription assay.

Fig. 1.

A schematic depicting the binding rate of T7 RNAP. (A) Representative PIFE1 DNA constructs where Cy3 (green circle) is labeled on the non-template strand at the promoter site. RNAP binding induces Cy3 PIFE. PQS is the potential G-quadruplex forming sequence present on the non-template strand. (B) Representative single-molecule Cy3 intensity time traces. The spikes indicate the Cy3 PIFE upon RNAP binding to the promoter, and the spikes increase with RNAP concentration. (C) The RNAP binding rate is calculated from the frequencies of Cy3 PIFE signal within a certain time frame.

4.3. Transcription initiation rate

To detect the transcription initiation at the single-molecule level, we designed FRET1 constructs where Cy5 and Cy3 dyes are labeled on the template and non-template strands, respectively (Fig. 2A). T7 RNAP (100 nM) and different concentrations of rNTP (0.01 to 1 mM) are added to the DNA constructs, and FRET measurement is taken. The FRET histogram shows a high FRET peak (~0.7) along with a DNA-only FRET peak (~0.4), indicating that transcription initiation induces a transcription bubble, which increases the FRET signal (Fig. 2B). Single-molecule time traces show a series of ~0.7 FRET spikes at varying rNTP concentrations (Fig. 2C). The frequency of FRET spikes increases as a function of rNTP concentration (Fig. 2D). Based on the transcription initiation rate, we determined 100 nM RNAP and 1 mM rNTP as the optimal condition for single-molecule transcription assay.

Fig. 2.

A schematic depicting transcription initiation. (A) Representative FRET1 DNA constructs where Cy3 (green) and Cy5 (red) are labeled on the non-template and template stand. RNAP binding and progression in the presence of rNTP induces transcription bubble formation. (B) FRET histogram of FRET1 DNA before and after RNAP addition in presence of rNTP. The additional high FRET peak in presence of RNAP indicates bubble formation. (C) Representative single-molecule FRET time traces. The FRET spikes represent the RNAP induced transcription bubble, and the spikes increase with rNTP concentration. (D) The transcription initiation rate is calculated from the frequencies of FRET spikes with a certain time frame.

4.4. T7 RNAP elongation

Genome-wide mapping for R-loop formation (R-ChIP) demonstrated that R-loops tend to form in the promoter region and elongate to the downstream gene body (Niehrs and Luke 2020). Hence, to track the RNAP movement from the promoter to the gene body, we employed a dual-PIFE assay (Lee, McNerney et al. 2020). We used PIFE2 DNA constructs where Cy3 and Cy5 dyes are positioned at the promoter and after PQS, respectively (Fig. 3A). Two dyes are separated by nearly 40 base pairs (~13 nm). This region is FRET insensitive but allows us to simultaneously measure PIFE from both dyes by exciting green and red laser. Upon transcription initiation, we observed Cy3 PIFE followed by Cy5 PIFE (Fig. 3B), signifying the RNAP movement from the promoter to the PQS site. This dual PIFE assay suggests successful elongation. A periodic signal spike of dual PIFE indicates repetitive RNAP movement during the kinetic cycle of transcription.

Fig. 3.

A schematic depicting transcription elongation via dual-PIFE assay. (A) Representative PIFE2 DNA constructs where Cy3 and Cy5 are labeled on the non-template and template strand. RNAP binding and moving forward in the presence of rNTP induces a transcription bubble. Simultaneous excitation of 532 nm (green) and 641 nm (red) lasers and RNAP movement induces Cy3 and Cy5 PIFE during transcription elongation. (B) Representative single-molecule dual-PIFE time traces where simultaneous Cy3 and Cy5 PIFE spikes represent successful transcription elongation. The top time trace is in the absence of RNAP.

4.5. Real-time R-loop and G4 formation

RNAP elongation leads to the formation of the R-loop (Lee, McNerney et al. 2020). R-loop formation is promoted at a guanine-rich sequence due to the higher thermal stability of rG/dC base pairing (Tan-Wong, Dhir et al. 2019). We designed a FRET3 construct with Cy3 and Cy5 positioned on either side of the PQS to visualize the R-loop and G4 formation (Fig. 4A). RNAP and rNTP are loaded into the reservoir, and real-time transcription measurement is carried out as described in section 4.1 (Paul and Myong 2022). The collected FRET values are used to generate a histogram, which shows a mid and high FRET compared to the DNA-only low FRET (Fig. 4B). The mid and high FRET peaks correspond to R-loop without G4 and R-loop with G4, respectively. Accordingly, the mid FRET peak completely disappears upon RNase H digestion. In contrast, the high FRET peak persists (Fig. 4B). Real-time traces show an either dynamic exchange between low and mid FRET or static mid FRET, indicating the transition between the duplex DNA and the R-loop. Static high FRET corresponds to G4 (Fig. 4C). The transition to mid-FRET always preceded the high FRET, suggesting that the R-loop forms before G4. Hence, DNA that contains the PQS on the non-template strand forms G4 and R-loop during transcription.

Fig. 4.

A schematic depicting transcription mediated R-loop and G4 formation. (A) Representative FRET3 DNA constructs where Cy3 and Cy5 are labeled across the PQS in the non-template strand. R-loop and G4 are on the template and non-template strand, respectively. RNase H digestion removes the R-loop but does not affect the G4 structure. (B) FRET histogram of FRET3 DNA before (top) and after RNAP addition (middle). The mid- and high-FRET peak correspond to R-loop and R-loop/G4. The mid-FRET peak corresponds to R-loop removed by RNase H digestion (bottom). (C) Representative single-molecule FRET time traces. Time traces from the top represents DNA only, DNA and R-loop transition, R-loop followed by G4, steady R-loop and steady G4. The dashed line represents the moment of RNAP with rNTP flow.

5. Detection of G4, R-loop, and RNA during transcription via gel-based assay

5.1. Electrophoretic Mobility Shift Assay (EMSA)

EMSA is a standard method for probing dsDNA containing R-loop and G4 species (Lee, McNerney et al. 2020). We use polyacrylamide gel electrophoresis (PAGE) instead of agarose gel electrophoresis. The polymerization of bis-acrylamide monomers results in a gel matrix that is more uniform and has a smaller pore size compared to agarose, offering higher resolution for small nucleic acid fragments (< 200 base pairs) (Green and Sambrook 2020). We use fluorophore-labeled DNA to measure transcription-mediated R-loop and G4 formation through EMSA gel electrophoresis (Fig. 5A). Since our labeled DNA contains both Cy3 and Cy5 dyes, DNA can be visualized via fluorescent gel imaging either at 628 nm excitation (for Cy5) or 520 nm excitation (for Cy3). The transcribed single-stranded RNA (ssRNA) has faster mobility than dsDNA and can be observed on a fluorescent gel imaging at 472 nm after SYBER Green RNA staining (Fig. 6A). DNA that contains a G4 and/or R-loop migrates slower through the gel than linear dsDNA.

Fig. 5.

A schematic depicting transcription observed via EMSA assay. (A) FRET3 DNA constructs that contain PQS on the non-template strand are used for transcription. Transcription initiation with rNTP results in the formation of R-loop and G4 structures. Transcription is terminated with EDTA and SDS is added to dissociate T7 RNAP from the DNA. Transcription samples are then loaded onto a gel for electrophoresis. (B) The red band represents linear dsDNA. ssRNA appears below linear dsDNA and is visualized by SYBER Green staining. DNA that contains R-loop or G4 appears above linear dsDNA bands. RNase H (RH) removes bands that contains an R-loop. RNase A (RA) digests ssRNA, resulting in a complex that moves faster through the gel.

Fig. 6.

A schematic depicting G4, R-loop, and RNA formation over time and gel quantification steps. (A) Products of transcription terminated at different times run through an acrylamide gel. The blue, red, and green bands represent the R-loop and G4/R-loop, DNA, and RNA, respectively. (B) Analysis of the gel with imageJ produces a peak intensity profile which can be quantified. (C) Quantification of the transcription reaction product.

Based on the position of the bands and the fluorescence signal on the EMSA gel, transcription-induced R-loop and G4 formation can be identified and quantified. First, run the transcription reaction as described in section 3.4 (Fig. 5A), then cast the acrylamide gel as described below. The methods described below use the Bio-Rad gel casting system but can be adopted for other gel-casting systems.

• Thoroughly clean the short and spacer glass plates with soap and water. Spray the glass plates with 70% ethanol and let it air dry completely. Assemble the glass plates.

• Prepare the solution for 10% polyacrylamide gel that contains 1X TBE buffer by combining acrylamide/bis- acrylamide solution, deionized water, 10X TBE, APS, and TEMED. The percentage of APS and TEMED to use depends on the condition of the acrylamide solution. Thoroughly mix the solution via tube inversion.

• Using a 1000 μL pipette, carefully transfer the 10% polyacrylamide solution in between the glass plates to avoid bubble formation.

• Insert the appropriate comb to cast the wells. Allow the gel to polymerize for about 10 minutes at room temperature.

• Assemble the gel into the Mini-PROTEAN Tetra Vertical Electrophoresis system. Fill the electrode assembly entirely with 1X TBE buffer. Fill the tank with 1X TBE quarter way.

• Load the gel with low molecular weight ladder, purple loading dye, and the prepared transcription reaction sample as described section 3.4.

• Run the gel at 4°C for 65 minutes at 12 mA for one gel or until the purple loading dye reaches the bottom green border of the Mini-PROTEAN Tetra Companion Running Module. If two gels are run simultaneously in the same apparatus, run for 65 minutes at 24 mA.

• Prepare RNA staining solution by adding 2.5 μL of SYBR Green II RNA gel stain to 50 mL deionized water. Add the RNA staining solution to a container.

• Remove the gel with Bio-Rad gel releasers. Immerse the gel in the RNA staining solution. Stain the gel on a rocker for ~7 minutes. Transfer the gel to a container filled with deionized water. De-stain the gel on a rocker for ~3 minutes.

• Image the gel at 628 nm fluorescence for 3 minutes and 520 nm fluorescence for 3 seconds.

5.2. RNase H and RNase A digestion

Ribonuclease H (RNase H) is an enzyme that cleaves the RNA in DNA: RNA hybrids through hydrolysis. By RNase H digest, we can determine which bands in the EMSA gel correspond to the R-loop (Lee, McNerney et al. 2020) since the band intensity will decrease or disappear after RNase H treatment. RNase A selectively degrades single-stranded RNA. After transcription, samples treated with RNase A would not display RNA bands in the EMSA gel. Comparing the bands of RNase A treated samples to those without RNase A treatment provides additional confirmation for bands representing RNA (Fig. 5B).

• Prepare competitive T7 promotor sequence by annealing 10 μM of the non-template and 10 μM of the template strand with the conditions: 95°C for 2 min; gradual cooling at the rate of −2°C/min until 40°C; further cooling at the rate of −5°C/min until 4°C.

• Perform transcription as described in section 3.4.

• Add 2 μL of 10 μM T7 promoter sequence to inhibit T7 transcription on the sample DNA competitively.

• Add 0.5 μL of RNase H (2.5 units) for RNase H treatment and incubate at 25°C for 20 minutes.

• For RNase A treatment, add 2 μL of 5 M NaCl and 0.5 μL of RNase A. Incubate for 20 minutes at 25°C.

• Add 0.5 μL of 0.5 M EDTA, 0.1 μL of 1% SDS, and 4 μL of 50% glycerol to stop the enzyme reaction.

• Run the reaction mixture as described in section 5.1.

5.3. Gel analysis and quantification

The band position from the EMSA gel sharply distinguishes dsDNA, G4 and R-loop containing dsDNA, and ssRNA based on their different mobility (Fig. 5B and Fig. 6A). The gel bands can be analyzed and quantified using various software such as ImageJ Fiji, which is available online for downloading. Below are the steps for analyzing the gels with ImageJ software (Fig. 6B).

• Load the tiff file of the gel into ImageJ.

• Use the rectangle selection tool to select the entire lane as the region of interest to be analyzed.

• Go to Analyze > Gels > Select First Lane. “1” will be shown at the center of the rectangle to indicate successful selection.

• Click the rectangle selection and drag it to the next lane. Go to Analyze > Gels > Select Next Lane. The lane number will be shown at the center of the rectangle to indicate successful selection. Repeat until all lanes are selected.

• Go to Analyze > Gels > Plot Lanes to generate the intensity profile for each lane.

• Use the straight-line tool to demarcate the region of the peaks for each band to be analyzed. Use the wand tracing tool to obtain the area of the demarcated region.

• Band intensities are normalized by calculating the fraction of each DNA, R-loop, and G4 species.

The fraction of different band intensities is plotted against transcription time (Fig. 6C). As expected, the dsDNA decreases while the R-loop and G4-containing dsDNA increase with time. The intensity of ssRNA bands increases with time due to repetitive cycles of transcription (Fig. 6C).

6. Conclusion

Transcriptional regulation is a critical biological process involved in gene expression. The PQS within the genomic DNA functions as a transcriptional regulator in a position- and orientation-dependent manner. PQS in the non-template strand leads to R-loop formation, promoting G4 formation, thereby enhancing transcriptional yield. We used single-molecule techniques and EMSA to visualize transcription and its underlying mechanisms. We observed sequential steps of transcription where PQS in the non-template strand resulted in the formation of several distinct higher order DNA and RNA structures. Through single-molecule assays, we directly detect RNAP promoter binding, transcription initiation, and elongation. Our data indicates that R-loop forms first, followed by G4 formation. Through EMSA analysis we observed two different R-loop states, R-loop alone and R-loop with G4, that appeared at different times during transcription. The single-molecule assay can not detect the transcription yield whereas EMSA can follow the production of RNA. Our findings suggest that co-transcriptionally induced R-loops promote G4 formation and enhance the transcriptional yield.