Biochemical Analysis of DNA Synthesis Blockage by G-Quadruplex Structure and Bypass Facilitated by a G4-Resolving Helicase

Abstract

G-quadruplex (G4) DNA poses a unique obstacle to DNA synthesis during replication or DNA repair due to its unusual structure which deviates significantly from the conventional DNA double helix. A mechanism to overcome the G4 roadblock is provided by the action of a G4-resolving helicase that collaborates with the DNA polymerase to smoothly catalyze polynucleotide synthesis past the unwound G4. In this technique-focused paper, we describe the experimental approaches of the primer extension assay using a G4 DNA template to measure the extent and fidelity of DNA synthesis by a DNA polymerase acting in concert with a G4-resolving DNA helicase. Important parameters pertaining to reaction conditions and controls are discussed to aid in the design of experiments and interpretation of the data obtained. This methodology can be applied in multiple capacities that may depend on the DNA substrate, DNA polymerase, or DNA helicase under investigation.

1. Introduction

G-quadruplex (G4) DNA structures can pose an obstacle to nuclear or mitochondrial genome replication, resulting in mutagenic events that compromise cellular homeostasis and bear relevance to human disease [1, 2]. Given the prevalence of G4 signature motifs and experimental evidence for the existence of G-quadruplex structures in both the nuclear [3–6] and mitochondrial (mt) [7–9] genomes, the consequences of G4 for DNA polymerization has attracted considerable interest. The first research advance to characterize the biochemical impact of G4 DNA on DNA synthesis catalyzed by a DNA polymerase using an in vitro system was from the Usdin lab [10]. Since then, many labs [11–19], including ours [20], have examined the ability of DNA polymerases to replicate past G4 structures in vitro using primer extension assays with defined radiolabeled DNA substrates harboring specific DNA sequences demonstrated to form G4 under physiological reaction mixture conditions. The inclusion of auxiliary factors which aid DNA polymerases in the biochemical reactions has been useful to tease out the importance of protein functional interactions in the mechanism of G4 bypass. In addition, by testing human nuclear or mitochondrial sequences with predicted G4 motifs, the biochemical studies have provided insights to the proposed mechanism of mutagenesis in vivo.

In this Methods paper, we describe experimental procedures and cautionary notes to perform primer extension assays with G4 DNA templates to investigate the effects of G-quadruplexes on DNA synthesis in a defined and manipulatable biochemical system. In our own lab, the application of this experimental approach has enabled us to elucidate the role of a G4-resolving helicase to allow smooth DNA extension past G4 obstacles; however, the outcome for mutagenesis is dependent on the fidelity of the DNA polymerase being tested [20].

2. Primer Extension on G4 DNA Template

2.1. G4 DNA Substrate Design and Preparation

Here, we desired to test the ability of purified recombinant mt DNA Polymerase γ or PrimPol to catalyze DNA synthesis past G4 derived from DNA sequences determined to be sites of variation from published SardiNIA mt genome sequence data [20]. Furthermore, we elected to test G4 DNA substrates that were derived from the mt genes NADH dehydrogenase subunit 1 and subunit 2, in which variants (point mutations) showed enriched frequency in the SardiNIA population. The linear oligonucleotides harboring a G4 site were 75-mer (mitoG4-29) and 64-mer (mitoG4-55), lengths that can be readily acquired from a commercial vendor specializing in oligonucleotide synthesis (in our case, Lofstrand Labs Ltd) at a suitable purity (95-99%). Matching oligonucleotides in which key guanine residues are replaced with one of the other three nucleotides (thymine, adenine, cytosine) serve as control (non-G4) templates. For partial duplex DNA substrate preparation, 10 pmol (500 nM in the reaction) complementary oligonucleotide (25-mer) (which can subsequently be annealed to the 3’ end of the template oligonucleotide (25 pmol)) is 5’ end-labeled with 30 μCi of [γ-³²P]ATP (3000 Ci/mmol, 10 mCi/ml) and T4 polynucleotide kinase (New England Biolabs) in a 20 μL reaction containing 70 mM Tris-HCl (pH 7.6), 10 mM MgCl₂, and 5 mM dithiothreitol. Unincorporated ATP is removed using a G25 spin column (Cytiva) centrifuged for 2 min at approximately 700 x g. Annealing of the radiolabeled oligonucleotide to the complementary oligonucleotide at a 1:2.5 ratio (200 nM and 500 nM, respectively) is achieved by incubating both together in a 50 μL reaction with 100 mM KCl (or 100 mM LiCl to disrupt the G4 structure) in a 95 °C heat-block for 5 min followed by slow cooling overnight to room temperature (RT). The annealing step can also be accomplished using a polymerase chain reaction thermocycler or a 37 °C isothermal method as described by Rogers et al. [21]. Annealed substrates are stored at 4 °C and remain stable for up to 4 weeks during which time the substrates are suitable for experiments, assuming a relatively high starting specific radioactivity of the labeled substrate.

As an alternative to the use of radioactive isotope for end-labeling the primer oligonucleotide of a partial duplex DNA substrate, fluorescently labeled primer can be used. For example, to monitor dCMP insertion on a partial duplex DNA substrate in which the template strand formed G4 DNA structure, the Eoff laboratory employed a primer strand that had a covalently linked 6-carboxyfluorescein (FAM) label [12]. The FAM-labeled oligonucleotide (purified using high-performance liquid chromatography by the manufacturer) was synthesized by Integrated DNA Technologies.

A few additional notes should be made regarding the design of the radiolabeled oligonucleotide partial duplex substrates that will be used for the primer extension assays. Oligonucleotides should be scrutinized for non-desired secondary structure (e.g., hairpin) that would interfere with annealing to the complementary oligonucleotide or form in the template strand downstream of the duplex formed by the annealed primer. We typically use Integrated DNA technologies - Oligo Analyzer Tool for this purpose. Other tools include Cloud-Based Informatics Platform for Life Sciences R&D | Benchling and Primer Design with Oligo Primer Analysis Software v. 7. For our application, we designed the DNA substrates in such a manner that the first G-rich block of the G4 structure was positioned such that the DNA polymerase would encounter the G4 structure at the 6^th nucleotide post-initiation of DNA synthesis from the 3’ hydroxyl of the annealed primer. Subsequent mapping of DNA synthesis termination sites by resolution of radiolabeled products using denaturing gel analysis confirmed that the initial pause or block was localized to the first G-rich block. Examples of the G4 and sequence related non-G4 substrates used in our study [20] are shown in Figure 1.

Figure 1. G4 DNA substrates and the control (non-G4) DNA substrates used for primer extension assays.

5’-³²P-end label of primer oligonucleotide is indicated by star. G4-forming sequence characterized by four G-rich blocks is indicated in red font. See text for details. Note that the MitoG4-29 and MitoG4-55 oligonucleotide sequences were predicted to form intramolecular G4 substrates using the G4Hunter algorithm and verified to form stable G4 structures in vitro by a combination of biophysical methods; however, the specific molecular topography of the G4 structure formed by either oligonucleotide sequence was not reported [44].

2.2. Primer Extension Assay using G4 Templates

In vitro reconstitution of DNA synthesis by human replisome machinery using purified recombinant proteins represents a unique experimental approach to study the parameters of faithful G4 replication under controlled biochemical conditions. Specifically, primer extension assays make it possible to directly compare the ability of various polymerases with the aid of auxiliary factors to extend DNA synthesis past G4 structures of varying stability determined by biophysical measurements. A basic schematic diagram of the approach is shown in Figure 2. Our group has previously employed primer extension assays to assess the requirements for human mt G4 DNA replication in a reconstituted mt replisome system [20].

Figure 2. Cartoon schematic showing the in vitro primer extension reaction with mitochondrial DNA polymerases and associated factors and analysis of products.

1) Radiolabeled partial duplex DNA substrates containing stable G4 (along with non-G4 control) are synthesized. Purified recombinant human mt DNA polymerase γ (Polγ) is added to the reaction mixture. Auxiliary replisome factors (mtSSB, Twinkle helicase, TFAM) can be added individually or in combination to assess their ability to assist Polγ to catalyzed DNA synthesis past the G4 block. 2) Replisome components are incubated with labeled DNA substrates and dNTPs under appropriate reaction conditions. 3) In the absence of the G4-resolving helicase Pif1, primer extension by Polγ is blocked by the presence of a stable G4. Additional replisome factors fail to facilitate extension through the G4 arrest site. In the presence of Pif1, Polγ catalyzes DNA synthesis past the G4. 4) Following termination of the reactions, products are resolved on a denaturing gel, detected by autoradiography, and analyzed by ImageQuant technology. Full-length extension and G4 arrest products are visualized. Note that combinations of Twinkle, mitoSSB, and TFAM are insufficient to allow for extension past the G4, whereas Pif1 enables efficient G4 bypass.

The minimal required components for reconstitution of canonical mt replication in vitro are the heterotrimeric mt DNA Polymerase γ, which consists of an enzymatic subunit (A) and two homodimer accessory subunits (B2), the hexameric TWINKLE helicase, which unwinds mt DNA to allow for progression of Polymerase γ [22], and the mitochondrial single-strand binding protein (mtSSB), which stimulates DNA Polymerase γ and stabilizes the replisome to prevent reannealing of the separated strands [23]. Beyond these core components, there are several additional auxiliary factors that are likely involved in the resolution of mt G4 DNA to allow smooth DNA synthesis. We previously included in our experiments the mt transcription factor A (TFAM), which is thought to play a role in mt replication of G4 structures [24], as well as the primase-polymerase PrimPol and the G4-resolving helicase Pif1, both of which are known to localize to mt [25, 26]. Below we describe considerations for carrying out primer extension assays on G4 templates in a reconstituted mt replisome system (Figure 3), which can be applied to other reconstituted systems (e.g., nuclear G4 replication), or to interrogate the ability of specific polymerases/helicases to unwind and allow progression through G4s in vitro.

Figure 3. Pif1 stimulates DNA Polymerase γ synthesis past the G4 structure in a specific manner.

(A) Reaction mixtures included 15 nm DNA Polymerase γ AB2, 100 μm dNTPs and 4 mm ATP in the presence of 26 nM Pif1 or varying concentrations of FANCJ. Primer extension products using the mitoG4-29 and non-G4-29 DNA substrates after a 10 min incubation at 30 °C were resolved on a denaturing 16% polyacrylamide gel. See text for details.

Before starting, we found it critical to optimize the reaction conditions of the primer extension assay such that 1) G4-containing substrates are stabilized by the presence of monovalent cations (e.g., K⁺, Na⁺) or destabilized by Li⁺, 2) the stoichiometries of enzyme components and replisome factors reflect those that are observed in vivo, and 3) reaction conditions favor the binding of all replisome factors to DNA templates with comparable affinities. These parameters and some salient points are discussed below.

To maintain G4 stability, all primer extension assays should be carried out under conditions in which the G4 structure is stabilized by the presence of either potassium or sodium ions (KCl or NaCl). Negative control reactions can be run in the presence of lithium ions (LiCl), which are known to destabilize G4s; however, an assessment of the effect of Li⁺ ion on the catalytic activity of the DNA polymerase (or helicase) included in the reaction mixture should be evaluated. In our case, we observed that Polymerase γ reaction mixtures containing Li⁺ displayed a marked increase in extension of both partial duplex DNA substrates harboring different G4 DNA templates, including a greater fraction of fully extended product for both substrates [20].

Secondly, replisome components should be added to each reaction in molar concentrations that reflect their in vivo stoichiometries; for example, our reactions included a 2-fold higher molar concentration of purified Polymerase γ accessory subunit B2 (30 nM in our experiments [20]) compared to the catalytic subunit A (15 nM). A third consideration is that the relative concentrations of additional replisome-associated factors to be tested in the reaction should be determined empirically based on the binding affinity of each factor to the DNA substrate under the given reaction conditions. For our assays, we found it was important to carry out electrophoretic mobility shift assays (EMSAs) with a broad titration of each auxiliary protein to determine the minimal protein concentration required for DNA-protein complex formation. EMSAs can be carried out in the same reaction conditions as the primer extension assays except that the chemical energy source (e.g., ATP) is omitted. The EMSA binding mixtures are prepared for loading on native polyacrylamide gels by the addition of 6X native dye containing 74% glycerol, 0.01% xylene cyanol and 0.01% bromophenol blue. The products of EMSA binding mixtures are resolved for 2 hr on non-denaturing 5% polyacrylamide (19:1 acryl/bis-acryl) gels (14 x 16 cm) electrophoresed at 200 V, 4 °C in 0.5X TBE running buffer using a Hoefer SE-400 electrophoresis unit (Hoefer Inc). Electrophoresis times can be modified according to DNA substrate length and protein size/molecular weight (200 V for 2 hr at 4 °C, for example). Importantly, the concentration of each protein included in the experimental system should be chosen such that it favors efficient binding to the partial duplex DNA template, irrespective of G4-forming sequence, but does not impede polymerase progression on the control (non-G4) DNA substrate. Additionally, EMSAs should be used to confirm each factor binds G4 and control non-G4 substrates under the conditions of the biochemical assays. As an example, we performed EMSAs to assess DNA substrate binding by increasing concentrations of TFAM, and then performed Polymerase γ primer extension assays using the G4 or non-G4 DNA substrates at a range of TFAM concentrations (14 – 865 nM) in which DNA substrate binding increased from hardly detectable to all substrate bound [20]. An alternative design would be to use more select protein (e.g., TFAM) concentration relative to the experimentally determined apparent dissociation constant (K_d).

Once the optimal reaction conditions are determined, experiments to measure primer extension are performed as previously described [20]. Briefly, radiolabeled substrate (0.5 nM final concentration) is added to a master-mix containing 10 mM Tris-HCl (pH 8.0), 50 mM KCl, 8 mM MgCl₂, 2 mM dithiothreitol, 100 μg/ml bovine serum albumin, and 100-250 μM ultrapure deoxynucleoside triphosphates (dNTPs, Cytiva), prepared with enough volume to accommodate the total number of reactions. The master-mix reaction mixture is then aliquoted to the appropriate number of 1.5 mL Eppendorf tubes, each requiring a 10 μl final reaction volume. It should be noted that because the G4 substrate is typically prepared in the presence of KCl, the DNA substrate preparation itself will contribute some additional KCl (e.g., 10 mM) to the final reaction mixture. In addition to the minimally required replisome components, reactions should contain the appropriate combination of replisome-associated factors to be tested; however, it is important to note that auxiliary factors may inhibit DNA synthesis in a concentration-dependent manner, highlighting the importance of titrating each component for optimization. Indeed, we found that TFAM concentrations exceeding a final concentration of 55 nM monomer under the specified reaction conditions impeded progression by DNA Polymerase γ. After addition of the DNA polymerase to initiate the reaction, the tube is briefly mixed, spun down in a tabletop microfuge, and promptly placed at 37 °C. Reaction mixtures in individual epi-tubes are typically incubated for 1-30 min at 37 °C and terminated in the order that they were initiated by addition of 5 μl of formamide loading buffer (80% formamide, 0.1% xylene cyanol, 0.1% bromophenol blue) and heating at 95 °C for 5 min.

After reactions are heated, they can be resolved on a denaturing (8 M urea) polyacrylamide (19:1 acyl: bis-acryl) gel electrophoresed on a 17 × 33.5 cm gel using a Gibco BRL SA vertical electrophoresis unit or similar apparatus. We find that 16% acrylamide is optimal for resolving products from template strands between 75 and 82 nt in length, but a range of acrylamide percentages from 10-20% can be used, depending on the design of the DNA substrate and the size of the predicted reaction products. We typically load 6-7 μL of sample into each well generated with an 18-well comb (0.5 mm thickness) with each well measuring 4 mm wide, 7 mm high and 3 mm spacing between each well. Remaining sample left over after loading can be stored at 4°C for as long as the radioactivity can be detected (typically 2 half-lives for ³²P). Gels are electrophoresed for 1 hr, 45 min at a constant 40 W, but this can be modified to optimize resolution of different length substrates. Gels can be imaged on any autoradiography imaging system. We typically expose our gels overnight using a storage phosphor screen/cassette and then image the following day using a Typhoon PhosphorImaging System.

A few comments about the analysis of primer extension products resolved on the denaturing polyacrylamide gels should be noted. The denatured radiolabeled oligonucleotide of the partial duplex substrate in the no enzyme (NE) control should migrate to a position near the bottom of the denaturing gel, whereas fully extended products should migrate to a position not far below the loading wells of the polyacrylamide gel (Figure 3). This will allow for optimal resolution of the primer extension reaction products. Intermediate bands represent partial extension products of DNA synthesis. One method to quantify primer extension products is to calculate the percent of full-length product formed. Full length product at the top of the gel is quantitated as a percentage of the entire signal in each lane. Individual bands can also be quantitated by drawing individual boxes around each band and dividing the signal in the band from the total signal in the gel lane. Individual products can be compared between reactions this way. In addition, a profile can be created for each lane of reaction products by drawing a line through the center of the gel lane from the largest product to the shortest. A graph is created with product length on the x-axis and signal on the y-axis. This kind of graph can show the distribution of products under different reaction conditions. Any background from the NE control is subtracted out from that percentage. This can be used to compare the effects of a particular DNA sequence (e.g., predicted G4-forming polynucleotide tract) on primer extension by a given DNA polymerase. In addition, [γ-³²P]ATP radiolabeled oligonucleotides representing different lengths of primer extension products can be used to make markers to map where nucleotide primer extension is blocked.

Some technical comments about the preparation of the glass plates and preparation/application of the denaturing polyacrylamide gel solutions should be noted. Prior to preparing the gel solution, we usually silanize one glass plate (typically the longer of two glass plates). The silanization process is done in a fume hood by applying a small amount of Sigmacote (Sigma) onto the plate about the diameter of a quarter and spreading it around the entire glass plate using a paper towel. Any excess Sigmacote should be removed and then the glass is left to dry overnight followed by washing. The two glass plates are cleaned with detergent and water, rinsed, and then with cleaned again with 70% ethanol to remove any dust or debris that could cause bubbles to form. 0.5 mm thick spacers are placed at the edges between the plates and then the seams between the two glass plates are sealed with gel sealing tape (electrical, 1-1.5 in wide) on the sides and bottom. An extra layer of tape is added to the bottom of the glass plates. Large clamps are placed along the sides of the glass plates over the tape. An alternative to taping the glass plates together is to use specialized silicone clamps such as those purchased from LABRepCo (Gel Casting Clamp for V15-17, V16 and V16-2 (1/pkg) - LabRepCo, LLC). A gel solution is usually prepared in bulk with the appropriate amount of acrylamide, urea and TBE. The solution must be heated to fully dissolve the urea and then filtered through a 0.45 μm filter device and stored at 4 °C until use. We typically use 40 ml of gel solution and add 20 μL of TEMED and 200 pμL of 10% ammonium persulfate (APS) for polymerization. The gel solution is carefully poured down one side of the glass plate with the plates tilted at a 45° angle to limit the formation of air bubbles. After filling the glass plates, we place a comb in carefully to prevent bubble formation and then we use three large metal binding clamps to hold the comb firmly between the two glass plates. We set the glass plate down horizontally and top off with gel solution (1-2 ml) on either side of the comb. The gel solution is typically polymerized within a couple hr at RT and ready for sample application. We remove the comb from the gel, set it up in the electrophoresis apparatus and add running buffer. We use a syringe with needle to wash running buffer into each well of the gel to clean out each well. To ensure the single-stranded DNA remains fully denatured after loading and minimize secondary structure formation of the nucleic acid that would alter its migration through the polyacrylamide gel, we pre-warm the polyacrylamide gel by applying current to the gel at a constant 40 W for 30 min or until the gel-plate sandwich reaches a temperature of 50-55 °C. This pre-warming step can be accomplished during the period that the reaction mixtures are being set up and incubated.

The use of primer extension assays described above represents a useful technique to assess the biochemical requirements for G4 DNA replication, one that could easily be applied to other reconstituted replisome systems to assess how different replication machineries handle G4 substrates (e.g. requirements for nuclear G4 DNA replication, resolution of telomeric G4s by various helicases, progression of the break-induced telomere synthesis (BITS) replisome through G4s), or, alternatively, to interrogate the impact of RNA G4s on in vitro transcription.

3. Stimulation of DNA Synthesis Past G4 Obstacles by a G4-Resolving DNA Helicase

As mentioned above, it is not uncommon that reaction mixture components and incubation temperature must be optimized to achieve a reasonable level of DNA synthesis catalyzed by the DNA polymerase in addition to G4 resolution catalyzed by the helicase under investigation. This optimization is essential to test for the functional interaction of DNA polymerase with helicase (and other auxiliary factors) on a substrate consisting of the primed partial duplex G4 DNA template described above. For our experiments, to test for the stimulatory effect of Pif1 helicase on either DNA Polymerase γ or PrimPol [20], we first determined that in the presence of ATP (4 mM) Pif1 efficiently resolved the G4 DNA substrate under the same reaction conditions used for either DNA polymerase, despite there being a significant difference in some components of the standard reaction mixtures for each polymerase. For example, Tris-HCl buffered at pH 8.0 is conventional for DNA Polymerase γ [27], whereas Tris-HCl buffered at more physiological pH (pH 7.0-7.5) has been used for PrimPol [28, 29]. To our good fortune, we determined that Pif1 displays robust helicase activity in the specified reaction mixtures for DNA Polymerase γ and PrimPol used in our study [20].

In addition to the difference in optimum pH, DNA Polymerase γ is more active at a greater concentration of KCl compared to PrimPol, i.e., 60 mM versus 10 mM, respectively. In this case, the KCl concentration differential deserves special comment because K⁺ is the stabilizing monovalent cation for the G-quadruplex barrel. In our experiments, we determined by native polyacrylamide gel electrophoresis analysis of the radiolabeled DNA substrates that even in the 10 mM KCl condition the mitoG4-29 and mitoG4-55 substrates (Figure 1) retained their stable G4 structures [20], an absolute requirement for G4 polymerase arrest assays. Aside from the chemical components of the reaction mixtures, the compatibility between the polymerase and helicase for temperature in which the reaction mixtures are incubated must be taken into consideration. For primer extension with Pif1 present, reactions were performed at 30 °C to match the optimal temperature for unwinding activity by the helicase despite our observations that synthesis by DNA Polymerase γ or PrimPol was moderately reduced at 30 °C compared to 37 °C [20].

An additional comment should be made regarding the potential differences that might be observed when comparing two DNA polymerases to extend DNA synthesis past a G4. In our studies [20], we consistently observed by denaturing gel electrophoresis analysis that the products of reactions containing PrimPol were longer than products of reactions containing Polymerase γ. Whereas DNA Polymerase γ arrested DNA synthesis completely within 2-4 nt of the first G-run, PrimPol synthesis was characterized by a ladder of products, typically in a window of 4-12 nt corresponding to DNA synthesis through the first G-run, first interstitial loop, second G-run and into the second loop. Despite this apparent difference between DNA Polymerase γ and PrimPol, Pif1 was able to stimulate DNA synthesis past the G4 block for either polymerase, resulting in full-length products.

Careful consideration of controls that address the specificity and mechanism of stimulation of polymerase-catalyzed DNA synthesis past a G4 roadblock by a G4-resolving helicase should be undertaken. One control is to test an engineered site-directed mutant version of the helicase (e.g., Pif1-K264A Walker A box, motif I [30]) that inactivates its ATPase/helicase function. Failure of the recombinant mutant ATPase/helicase dead protein, (purified in a manner identical to that of the recombinant wild-type helicase protein) to enhance DNA synthesis by the polymerase past the G4 structure provides evidence that the intrinsic catalytic ATPase/unwinding function of the helicase under investigation is responsible for the stimulation, rather than a contaminating G4-melting activity in the helicase preparation. Another useful control that provides experimental evidence that the helicase is targeting the G4 structure to stimulate DNA synthesis is to include a G4-stabilizing ligand such as telomestatin in the reaction mixture [20]. Under conditions in which the G4 ligand specifically inhibits G4 unwinding by the helicase under study (and has negligible effect on its more conventional duplex unwinding activity), it is predicted that the G4-binding compound will deter helicase stimulation of DNA synthesis past the G4 block, assuming the G4 ligand does not affect DNA polymerase activity on the control (non-G4) substrate, which can be readily tested.

An important issue to address is the specificity of a G4-resolving helicase to stimulate a DNA polymerase to catalyze polynucleotide synthesis past the G4 obstacle. This can be addressed by substituting another G4-resolving helicase in the reaction mixture. In our experiments [20], we substituted the purified recombinant FANCJ DNA helicase for Pif1 because both are characterized as 5’ to 3’ DNA helicases that require a 5’ single-stranded DNA tail adjacent to the G4 structure to resolve it [31, 32]. We found that FANCJ failed to stimulate synthesis by Polymerase γ despite its ability to robustly unwind the G4 structure under the reaction conditions being used. Although not further explored experimentally in our published work, we surmise that the biophysical and biochemical properties of a given G4-resolving helicase and its unique mechanism of G4 unwinding is responsible for the differential effect of the helicase to stimulate Polymerase γ G4 DNA synthesis, as previously discussed for Pif1 versus FANCJ [20]. It does not go unnoticed that in some cases the physical interaction between a helicase and DNA polymerase may provide the basis for the functional interaction in G4 DNA synthesis; however, this is an understudied area.

4. Primer Extension Fidelity Assay with G4 DNA Templates

The standard protocol to measure DNA polymerase fidelity was first developed by Dr. Thomas Kunkel (NIEHS) to measure error rates of DNA polymerase β [33]. Briefly, purified DNA polymerase is incubated with a gapped M13 DNA substrate, where the polymerase synthesizes the LacZ gene. Errors introduced during gap filling will inactivate LacZ and upon transformation into E. coli, can be scored for errors through blue/white selection and sequenced for error identification. However, this technique is limited to scoring only synthesis events which fill the entire gap. Due to the blockage of DNA synthesis by G4 structures, we could not use the Kunkel technique in our study. To overcome this deficiency, we developed a protocol to sequence in vitro DNA replication by directly sequencing the synthesized strand independent of where synthesis was terminated.

To develop this assay, we linked a DNA oligonucleotide to the free 3’-OH of the extended primer and the template oligonucleotide to introduce a known DNA sequence for subsequent PCR amplification and to introduce a barcode sequence to identify individually extended products. This ligation was achieved using Circligase II single-strand DNA ligase (Lucigen) along with specialized ssDNA linker sequence, 5’ Phos-AGANNNNNNAGATCGGAAGAGCACACGTCTGAACTCCAGTCAC-Biotin 3’ (Figure 4A). We developed the linker with a 5’ phosphate to initiate the ligation with the 3’-OH group, a 6x N unique molecular identifier (UMI), and a 3’ biotin to inhibit inter-linker ligation. Ligations were performed in 20 μL volume containing 10 μL extension reaction, 1x Circligase II buffer, 2.5 mM MnCl₂, 1 M Betaine, 0.3125 nM linker oligo, and 2.5 units CircLigase II. Reactions were heated to 60 °C for 1 hr to support ligation followed by inactivation by heating to 80 °C for 10 min. During the ligation reaction both the extension template strand and the newly synthesized strand are substrates for ligation. To differentiate the two strands, the newly synthesized strand was amplified using high-fidelity PCR with primers annealing to the linker sequences; GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT and the extension primer; CGCCAGGGTTTTCCCAGTCACGACC. PCR was performed in 50 μL reaction containing; 2 μL ligation reaction, 1x Herculase II DNA polymerase buffer, 0.25 mM dNTP, 0.25 μM primers, and 0.5 μL Herculase II DNA polymerase (Agilent). DNA was amplified over 30 cycles (95 °C – 20 sec, 55 °C – 20 sec, and 72°C – 30 sec) followed by dA-tailing by adding 1 unit Taq DNA polymerase (Takara) and incubating at 65 °C for 5 min.

Figure 4. Measurement of DNA polymerase fidelity during primer extension in vitro.

(A) Experimental design for adding linker sequences (blue) onto the newly synthesized DNA strand (red) following primer extension of DNA templates (black). After linker ligation, newly synthesized strand is preferentially amplified by PCR using a linker specific primer and the extension primer (green) followed by cloning and sequencing. (B) Representative polymerase errors made during primer extension. Newly synthesized sequence (cyan) is shown aligned to template strand (black) with G-quadruplex underlined and DNA polymerase errors (red) shown.

For screening and sequencing, products were then cloned into pSC-A-amp/kan vectors (Agilent) and transformed into E. coli. Individual E. coli colonies were submitted for sequencing using the T7 sequences on the cloning vector. Sequences were aligned against the template DNA oligo and individual extension events were identified using the UMI. Under different synthesis conditions, four unique mutational events were identified: 1) substitutions, 2) insertions, 3) deletions, and 4) non-templated additions at the 3’ end (Figure 4B). For mutational frequency analysis non-templated additions were ignored and complex mutations (>1 base pair (bp) insertion or deletion, or combination of > 1 bp insertion, deletion and/or substitution) were calculated as a single mutational event. Control experiments ligating onto the template oligo without DNA synthesis showed an error rate of 6.4 × 10⁻⁴ mutations/bp. Thus, the assay has the unique ability to identify terminated or stalled extension events, and shows the same background error-rate as the conventional gap filling assays with spontaneous replicative error at 6.4 × 10⁻⁴ mutations/bp [33]. To achieve statistical power, a minimum of 50 unique extension events were analyzed per replication condition. To capture rare mutational events, this technique can be easily adapted to high-throughput sequencing applications.

Using this approach to study DNA synthesis fidelity of PrimPol acting upon the G4 DNA templates, we determined a five-fold increase in error rate, compared to non-G4 DNA substrate, that persisted even in the presence of Pif1 helicase [20]. Thus, the biochemical data suggested that PrimPol catalyzed DNA synthesis past the G4 block enabled by Pif1 helicase was still error-prone despite the significant enhancement of G4 bypass. Conversely, DNA Polymerase γ bypass efficiency which was also greatly stimulated by Pif1 helicase was accomplished with the same high fidelity of DNA synthesis [20].

5. Discussion

This Methods article is one of a series in a Special Issue devoted to experimental techniques and strategies to investigate the molecular and cellular functions of DNA helicases and RNA helicases. We have focused on biochemical approaches to characterize the functional interaction of a helicase with a polymerase to stimulate DNA synthesis past a G4 roadblock. Molecular cooperation of nucleic acid binding proteins or catalytic enzymes with proteins stably or transiently associated with the replisome is an area of tremendous interest. Careful biochemical studies are required to characterize the mechanisms whereby efficient DNA synthesis past endogenous genome obstacles (e.g., alternate DNA structures, protein-DNA complexes, etc) or exogenously induced DNA damage is achieved.

In terms of G4 obstacles, an emerging area of interest is the consequence of G-quadruplex architecture and topology for nucleic acid metabolism in vivo, and the G4-resolving mechanisms and pathways that operate to enable smooth transactions in DNA replication and repair, as well as transcription. For example, our laboratory determined that a specialization among Fe-S cluster helicases exists to resolve G4 DNA structures in which the human FANCJ helicase mutated in Fanconi Anemia, as well as breast and ovarian cancer, efficiently resolves intramolecular G4 DNA structures [34]. Combining results from biochemical and cell-based genetic complementation assays, it was determined that a minimal threshold of FANCJ catalytic activity rescues phenotypes associated with a poor response to replication stress induced by a G4-binding ligand [35]. Notably, very recently published work using a reconstituted system suggests a model in which a multistep mechanism involving more than one G4-resolving helicase is at play to enable smooth and high fidelity eukaryotic DNA replication [36]. Thus, an exciting era of mechanistic and live cell studies to investigate the roles of DNA helicases (e.g., FANCJ [37]) and interacting factors in G4 DNA replication and other processes is well underway. For a new perspective, see [38].

With the now strong evidence that DNA G4 and RNA G4 structures exist in vivo and have consequences for normal genome homeostasis [39], the need for further studies of G4 metabolism has moved to the forefront of molecular biology with translational implications. For a perspective of G4-interacting DNA helicases and polymerases as therapeutic targets, see Estep et al. [2]. G4 binding ligands are being explored as potential chemotherapy drugs that exploit cancer-relevant targets including proto-oncogene promoters and telomeres where the action of both helicases and polymerases are relevant [40–43]. Basic research approaches such as the ones described here can be applied in various contexts of replicative synthesis or DNA repair synthesis to elucidate mechanisms important for genomic stability and cellular homeostasis.

Highlights.

• Experimental approaches to assess primer extension assay using a G4 DNA template are described.

• Experimental design of primer extension assays enables one to measure the extent and fidelity of DNA synthesis when a G4 obstacle is encountered.

• Coordination of a G4-resolving helicase with a DNA polymerase can be assayed biochemically.

• Biochemical studies to assess G4 DNA synthesis bypass aided by a G4-resolving helicase provides molecular insight that is relevant to genome metabolism and translational studies.