Using Modified Synthetic Oligonucleotides to Assay Nucleic Acid-Metabolizing Enzymes

Summary

Here, a protocol for assaying nucleic acid metabolizing enzymes is presented, using examples of ligase, nuclease, and polymerase enzymes. The assay utilizes fluorescently labeled and unlabeled oligonucleotides that can be combined to form duplexes mimicking RNA and/or DNA damages or pathway intermediates, allowing for the characterization of enzyme behavior.

Abstract

The availability of a range of modified synthetic oligonucleotides from commercial vendors has allowed the development of sophisticated assays to characterize diverse properties of nucleic acid metabolizing enzymes that can be run in any standard molecular biology lab. The use of fluorescent labels has made these methods accessible to researchers with standard PAGE electrophoresis equipment and a fluorescent-enabled imager, without using radioactive materials or requiring a lab designed for the storage and preparation of radioactive materials, i.e., a Hot Lab. The optional addition of standard modifications such as phosphorylation can simplify assay setup, while the specific incorporation of modified nucleotides that mimic DNA damages or intermediates can be used to probe specific aspects of enzyme behavior. Here, the design and execution of assays to interrogate several aspects of DNA processing by enzymes using commercially available synthetic oligonucleotides are demonstrated. These include the ability of ligases to join or nucleases to degrade different DNA and RNA hybrid structures, differential cofactor usage by the DNA ligase, and evaluation of the DNA-binding capacity of enzymes. Factors to consider when designing synthetic nucleotide substrates are discussed, and a basic set of oligonucleotides that can be used for a range of nucleic acid ligase, polymerase, and nuclease enzyme assays are provided.

Introduction

All life forms require nucleic acid processing enzymes to carry out fundamental biological processes, including replication, transcription, and DNA repair. Key enzymatic functionalities for these pathways are polymerases, which generate copies of RNA/DNA molecules, ligases which join polynucleotide substrates, nucleases that degrade them, and helicases and topoisomerases, which melt nucleic acid duplexes or change their topology^{1,2,3,4,5,6,7,8,9,10}. In addition, many of these enzymes provide essential molecular tools for applications such as cloning, diagnostics, and high-throughput sequencing^{11,12,13,14,15}.

The functional characteristics, kinetics, and substrate specificities of these enzymes can be determined using labeled DNA/RNA substrates produced by annealing oligonucleotides. Tracking substrates and products has traditionally been achieved by introducing a radioactive label (³²P) at either the 5' strand end, which can then be detected by photographic film or with a phosphor imaging system^16,17. While radiolabeled substrates offer the benefit of increased experimental sensitivity and do not alter the chemical properties of a nucleotide, the potential health hazards from working with radioisotopes have encouraged the development of non-radioactive nucleic acid labeling to provide a safer alternative for DNA and RNA detection^18,19,20. Among these, fluorescence detection, including direct fluorescence detection, time-resolved fluorescence, and energy transfer/fluorescence quenching assays stand out as the most versatile^21,22,23,24. The extensive array of fluorophores enables different designs of DNA/RNA substrates featuring unique reporters on each oligonucleotide²⁵. Additionally, the stability of fluorophores, when compared to radioisotopes, allows users to produce and preserve significant quantities of fluorescently labeled DNA substrates¹⁹. These fluorophore-labeled substrates can be incubated with the protein of interest, along with different combinations of metal and nucleotide cofactors, to analyze binding and or enzyme activity. Visualization of binding or activity can be observed using various fluorophore dye channels with a gel imaging system. As only the fluorescently labeled oligonucleotides will be visible using this technique, any increase or decrease in the size of the labeled oligonucleotide will be easy to follow. Gels can also be stained afterward, with nucleic acid staining dyes to visualize all DNA bands present on the gel.

Poly-nucleic acid ligases are enzymes that join fragments of DNA/RNA, catalyzing the sealing of breaks by the formation of a phosphodiester bond between 5' phosphorylated DNA termini and the 3' OH of DNA. They can be divided into two groups according to their nucleotide substrate requirement. The highly conserved NAD-dependent ligases are found in all bacteria²⁶ while the structurally diverse ATP-dependent enzymes can be identified through all domains of life^8,27. DNA ligases play an important role in Okazaki fragment processing during replication as well as being involved in various DNA repair pathways, such as nucleotide and base excision repair, through the sealing of spontaneous nicks and nicks that are left after repair^8,10. Different DNA ligases exhibit varying capacities to join different conformations of DNA breaks, including nicks in a duplex, double-stranded breaks, mismatches, and gaps, as well as RNA and DNA hybrids^28,29,30. A diverse range of ligatable substrates can be assembled by annealing oligonucleotides with a 5' phosphate to generate juxtaposed 5' and 3' termini in a nucleic acid duplex^31,32,33. The most common method of analysis is resolution by urea PAGE in an end-point assay format; however, recent innovations have included the use of capillary gel electrophoresis, which allows high throughput³⁴, mass-spectrometric profiling³⁵, as well as a homogenous molecular beacon assay, which allows time-resolved monitoring³⁶.

The first step in a ligation reaction is the adenylation of the ligase enzyme by adenosine triphosphate (ATP) or Nicotinamide adenine dinucleotide (NAD), resulting in a covalent enzyme intermediate. The second step in the reaction is adenylation of the nucleic acid substrate on the 5' end of the nick site, which is followed by ligation of the nucleic acid nick strands. Many ligase enzymes that are recombinantly expressed in E. coli are purified in the adenylated form and, therefore, are able to successfully ligate nucleic acids without the addition of a nucleotide cofactor. This makes it difficult to determine what particular type of nucleotide cofactor they require for the ligation of nucleic acids. In addition to describing assays to evaluate DNA ligase activity, a method to reliably determine the cofactor usage by de-adenylating the enzyme using unlabeled substrates is also presented.

Nucleases are a large and diverse group of DNA/RNA modifying enzymes and catalytic RNAs that cleave the phosphodiester bonds between nucleic acids³⁷. Nuclease enzyme functionalities are required in DNA replication, repair, and RNA processing and can be classified by their sugar specificity for DNA, RNA, or both. Endonucleases hydrolyze the phosphodiester bonds within a DNA/RNA strand, while exonucleases hydrolyze DNA/RNA strands one nucleotide at a time from the 3' or 5' end and may do so from either the 3' to 5' or the 5' to 3' end of the DNA³⁸.

While many nuclease proteins are non-specific and may be involved in multiple processes, others are highly specific for a particular sequence or DNA damage^6,39,40. Sequence-specific nucleases are used in a wide range of biotechnological applications, such as cloning, mutagenesis, and genome editing. Popular nucleases for these applications are restriction nucleases⁴¹, zinc-finger nucleases⁴², transcriptional activator-like effector nucleases, and most recently, the RNA-guided engineered CRISPR nucleases⁴³. Damage-specific nucleases have recently been identified, such as the EndoMS nuclease, which has specificity for mismatches in the DNA through its mismatch-specific RecB-like nuclease domain^5,44. Nuclease activity assays, historically, have been done as discontinuous assays with radiolabeled substrates; however, in addition to their other drawbacks, these do not allow the identification of the site that is cut by a nuclease protein, which is possible when using fluorescently labeled substrates^45,46. More recently, continuous nuclease assays have been developed which work by using different DNA dyes that interact with DNA in different states; for example, emitting a higher fluorescent signal when interacting with dsDNA than in its unbound state, or binding specifically to short RNAs⁴⁷. Other continuous nuclease assays use DNA hairpins with a fluorophore group on the 5' and a quencher on the 3' end so that fluorescence increases as the oligonucleotide is degraded due to a separation of the fluorophore and the quencher⁴⁸. While these assays allow one to characterize the kinetics of DNA-degrading proteins, they require previous knowledge of the enzyme's function and substrate and are also limited to enzymes that change the DNA conformation to cause a difference in dye binding. For this reason, endpoint assays that resolve individual nuclease products are still desirable to provide insight into DNA modifications caused by protein activity.

Here, a detailed procedure is presented for the design of fluorescently labeled DNA/RNA oligonucleotides that can be mixed and matched to generate substrates for testing the activity of novel nuclease, polymerase, and ligase enzymes. The validation of this basic set of oligonucleotide sequences simplifies experimental design and facilitates economical profiling of a wide range of enzymatic functionalities without needing to purchase a large number of bespoke substrates. A detailed procedure is provided for running a standard DNA-processing enzyme assay with these substrates, using the example of DNA ligase activity and modifications for assaying and analyzing nuclease and polymerase enzymes are described. In addition, a modified assay for determining the cofactor specificity of the DNA ligase enzyme with high accuracy is given, and dual-labeled probes are used to evaluate the assembly of multi-component ligations. Finally, modifications to the basic assay format are discussed to allow it to be used to determine protein-DNA interactions with the same substrates by the electrophoretic mobility shift assay (EMSA).

Protocol

1. Design and purchase of oligonucleotides

NOTE: Design single-stranded oligonucleotides to be assembled and annealed into the desired duplexes. One or more of the strands in a duplex must bear a fluorescent moiety for tracking oligonucleotide processing by the enzyme of interest. A basis set of single-stranded sequences that can be assembled for a range of activities is provided in Table 1.

1. Incorporate the specific modifications needed for the enzyme of interest as described below.
   1. For DNA ligase substrates (Figure 1): Assemble the simplest substrate from three oligonucleotides: a 5' phosphorylated donor strand (NL2), a 5' FAM-labeled acceptor strand (NL1), and a complement that bridges the two (NL3).
      1. Ensure strands providing the 5' terminus of the ligatable nick are phosphorylated before assembly of the substrate master mix in step 2. Order this as a modification on NL2 (as given in Table 1) or use enzymatic phosphorylation with T4 polynucleotide kinase after resuspending the oligonucleotides.
      2. Include 5'-terminal phosphorylation of NL6 and NL8, which comprise the complement of double-stranded breaks depicted in Figure 1A (NL6/NL7 and NL8/NL9) as this most closely resembles the natural substrate produced from a restriction endonuclease. Use a dual-labeled substrate to determine the relative extents of ligation for multi-part assemblies (see step 6).
      3. Alter the complement strand to produce mismatches (NL10) and gaps (NL11).
         NOTE: Variations on the simple nicked substrate are depicted in Figure 1A. It is possible to use other sequences to produce an even wider range of mismatches or longer gaps by varying the underlined position.
      4. Substitute DNA oligonucleotides for RNA oligonucleotides.
         NOTE: Variations on the simple nicked substrate are depicted in Figure 1B. A wider range of DNA/RNA duplexes can be generated by additional combinations of the basis set given here to generate, for example, double-stranded breaks containing both RNA and DNA. An example of this variation is given in step 6 below, where a dual-label strategy is used.
   2. For DNA polymerase substrates: Assemble the oligonucleotides NL1 and NL3 listed in Table 1 to give a simple primer-extension assay. Investigate additional aspects of polymerase activity by introducing modifications into either the NL1 (primer) or NL3 (template) strands.
      1. Incorporate damaged base analogs into the NL3 oligonucleotide prior to position 20 to determine the ability to bypass damaged lesions on the template strand.
      2. Incorporate damaged base analogs into the NL1 oligonucleotide at position 20 to determine the ability to extend a damaged primer.
      3. Use either RNL1 or RNL3 in the duplex to investigate the extension of an RNA primer or the use of an RNA template.
   3. For nuclease substrates (Figure 2): Assemble oligonucleotides to give a non-exhaustive range of double and single-stranded substrates (Figure 2Ai) as well as a range of flapped and splayed junctions (Figure 2Aii) and damaged substrates (Figure 2B).
      1. To probe ribonuclease activities, iteratively substitute NL1, NL2, and NL3 with RNL1, RNL2, and RNL3. Use additional RNA versions of HJ5 and HJ6 to expand this set further.
      2. Use oligonucleotides MD5, MD6, and MD9 that have a centrally placed modification that mimics oxidative damage, an abasic repair intermediate, or a deamination product (Figure 2B). The substrates will detect cleavage of the strand at this position. Label the complement NL3 strand with an orthogonal fluorophore such as TAMRA to detect double-strand cleavage (see step 6).
      3. Use orthogonal labeling of the complement to detect double-stranded cutting at the mismatched sites on both the probe (NL5 and ND9) and the complement (MD10 and NL10) strands.
2. Order synthetic oligonucleotides incorporating relevant fluorophores and other modifications from a commercial vendor.
   NOTE: A 100 nM synthesis scale and HPLC purification subsequent to synthesis are suitable for the assays described.

2. Assembling and annealing nucleic acid duplexes

1. Resuspension and dilution of oligonucleotides
   1. Before opening, centrifuge the lyophilized oligonucleotides in their 2 mL tubes at full speed in a benchtop centrifuge for 2-5 min to ensure the nucleic acid is on the bottom of the tube.
   2. Prepare a master stock of 100 µM by resuspending the oligonucleotides in TE buffer (10 mM tris(hydroxymethyl)aminomethane (Tris), 1 mM ethylenediaminetetraacetic acid (EDTA)). Ensure the oligonucleotides are thoroughly resuspended by repeated gentle vortexing and brief centrifugation at full speed.
   3. Prepare a 10 µM stock by diluting an aliquot of master stock with TE buffer. Dilute the 10 µM stock with ultrapure water (MQ water) to prepare working stocks with concentrations of 0.5 µM, 0.7 µM, or 2.5 µM as per Table 2.
2. Assembling and annealing the reaction master mixes
   1. Use working stocks to make up the reaction master mixes using the combinations provided in Table 2 and volumes given in Table 3. For the standard DNA ligase assay and most other assays described here, the final buffer composition is 50 mM Tris pH 8.0, 50 mM NaCl, 10 mM Dithiothreitol (DTT) with 10 mM Mg as the divalent cation.
   2. Anneal the oligonucleotides in a PCR or microcentrifuge tube by heating at 95 °C for 5 min using a heating block or thermocycler. Allow to cool to room temperature for 30 min (volumes <1 mL) to 1 h (volumes >1 mL). For longer oligonucleotides (>40 nt), carry out slower cooling by using a thermocycler with a down ramp of 95 °C to 25 °C over 45 min, or float the tube containing annealing mixture in a 1 L beaker of boiling water and allow to cool to room temperature overnight.
   3. Add nucleotide cofactors and other heat-sensitive buffer components to the master mix after cooling to room temperature. Use the final reaction mixture directly for the assay by the addition of enzyme (see step 3 below) or store at -20 °C for future use.

3. Standard assay setup

1. Assembly and initiation of the assay reaction
   1. Combine 22.5 µL of the substrate master mix of interest with 2.5 µL of the DNA ligase or other enzyme of interest in a PCR tube. Run reactions in duplicate or triplicate, especially if the results will be quantitated.
   2. Include a no-protein control (buffer only) in the assay samples. Include no cofactor controls at this point, if needed.
      NOTE: Enzymes can be stored at -20 °C in 50% v/v glycerol, allowing them to be pipetted directly from solution. Ensure enzyme solutions with glycerol are well mixed prior to addition, either by pipetting to mix or by gentle vortexing.
2. Immediately transfer the reactions to a PCR machine at 25 °C and incubate for 30 min. Vary the temperature and duration depending on the optimal conditions for enzyme activity.
3. Quench the reactions by adding 5 µL of loading dye (95% formamide, 0.5 M Ethylenediaminetetraacetic acid (EDTA), bromophenol blue) and incubate at 95 °C for 5 min.

4. Analysis of assay results

1. Prepare Tris-Borate-EDTA (TBE)-Urea PAGE gels as described below.
   1. Prepare a stock of 20% acrylamide, 7 M Urea, and 1x TBE solution. For the oligonucleotide set described here, use Acrylamide/Bis Solution in a 29:1 ratio for optimal resolution.
   2. For one gel, combine 10 mL of 20% acrylamide and 7 M urea solution with 100 µL of APS (10 %) and 3 µL of Tetramethylethylenediamine (TMED) and cast in a gel caster.
2. After the gel solidifies, run the samples on the TBE urea gel at 45 - 55 °C.
   1. Pre-run the gel in 1x TBE buffer for 30 min at 10 mA per gel with external heating.
   2. Remove excess urea in the wells of the gel by flushing with 1x TBE using a pasture pipette.
   3. Load 10 µL of each reaction and run at 10 mA for 1.0-1.5 h with external heating.
3. Visualize the gel on the imager with the correct settings for the chosen fluorophore. For FAM, use a filter set that gives excitation/emission at 495/519 nm, which is stored as a pre-set in most imagers.
4. Quantify the band intensity of the product and substrate using image processing software with the imager, or an external program such as ImageJ^49,50 and calculate the percentage of the product using the formula
   
   Where P is the integrated value of the product band, and S is the integrated area of the substrate band. In the case of the DNA ligase reaction example, the product band runs at 40 nucleotides (nt) and the substrate band at 20 nt.

5. De-adenylation of the DNA ligase to test cofactor specificity

1. Preparation of reaction master mixes
   1. Prepare one set of the master mix containing the FAM-labeled NL1 oligonucleotide as described in Table 4. Prepare a second set containing the NL1 oligonucleotide with no FAM label, as described in Table 4.
   2. Separately, heat both DNA duplexes to 95 °C for 5 min and cool for 30 min to 1 h at 25 °C. Do not add nucleotide cofactor to either master mix.
2. Assembly and initiation of de-adenylation reaction
   1. Prepare a single de-adenylation reaction for each cofactor type/ concentration to be tested by combining 10 µL of the unlabeled master mix with 2.5 µL of ligase enzyme.
   2. Prepare additional tubes as no-cofactor control and no protein control (2.5 µL of buffer added in place of enzyme).
   3. Incubate the reactions at a temperature specific to the enzyme's optimum activity, for 1-2 h. Incubation time can be increased if the enzyme is still adenylated.
3. Run the ligation reaction with the cofactor.
   1. Add 10 µL of the labeled master mix and 2.5 µL of the desired nucleotide cofactors (e.g., ATP, NAD, ADP, or GTP) directly to the de-adenylated reaction (0.1-1 mM final concentration).
   2. Add 2.5 µL of reaction buffer to the no nucleotide cofactor control.
   3. Incubate the reactions for the same time period and temperature as previously used. Quench and visualize as described in step 4.

6. Using dual-labeled substrates for splinted ligation or multi-part assembly

1. Design and purchase an oligonucleotide with a florescent moiety that has a different excitation/ emission spectrum to the fluorophore already used.
   1. In the set-up described here, use NL2 (TAMRA) oligonucleotide having 5-Carboxytetramethylrhodamine (TAMRA) on the 3' end (Table 1).
2. Assemble the master mix as described below.
   1. Combine the components of the reaction described in step 2, including equimolar ratios of all oligonucleotides used in the assembly, as well as buffer and divalent cations.
   2. Anneal by heating at 95 °C for 5 min and cooling at 25 °C for 30 min - 1 h. Add the cofactor and the enzyme and incubate as described in step 3.
3. Run and image the samples as described in step 4 using the appropriate channels for the fluorophore pair in the substrate. In the case of FAM and TAMRA, these are the Fluorescein (FITC) and Tetramethyl rhodamine (TRITC) channels present on most imagers.

7. Evaluation of DNA binding by Electrophoretic Mobility Shift Assay (EMSA) on native gel

1. Prepare a 10% native TBE PAGE gel as described below.
   1. Combine 2.5 mL of 40% acrylamide, 1 mL of 10x TBE, 100 µL of 10% APS, 3 µL of TMED, and 6.5 mL of MQ water and cast in a gel caster.
2. Assemble the binding reaction as described below.
   1. Assemble the EMSA substrate according to Table 5 so that EDTA (10 mM) is included and metal ions are omitted.
   2. Combine 20 µL of the EMSA substrate master mix with 5 µL of the protein in a PCR tube. Include a no protein control sample. Incubate for 30 min at 25 °C.
3. Analyze by native electrophoresis as described below.
   1. Add 5 µL of 5x native loading dye (100 mM EDTA, 0.25 % bromophenol blue, 25% v/v glycerol, and MQ water up to 1 mL) to the samples.
   2. Load on the prepared gel and run at 60 V for 2-3 h with cooling by water circulation until the dye front is a few cm above the end of the gel.
   3. Visualize and analyze gels as described in step 4.

Results

Ligation by DNA ligase
DNA ligase enzymatic activity will result in an increase in the size of the fluorescently labeled oligonucleotide when visualized on a urea PAGE gel. In the case of the substrates for both DNA- and RNA-ligation listed in Table 2, this corresponds to a doubling in size from 20 nt to 40 nt (Figure 3A). Optimal enzyme activity can be determined by changing conditions such as temperature, protein concentration, or incubation time (Figure 3B) nucleotide cofactors, metal cofactors (Figure 3C). The relative activities of a ligase on different substrates can be compared in parallel (Figure 3D).

Quantification of product and substrate bands is expressed as the percentage of the total substrate ligated and, as shown in Figure 3, can be used to evaluate the specific activity of the enzyme or to determine activity optima such as divalent cation preference, temperature, or pH.

The bacterial DNA ligase DV-1-1-Lig (Figure 3) is capable of ligating nick and mismatch DNA substrates and can utilize both magnesium and manganese for ligation activity, with a preference for magnesium. Further details of this enzyme are provided in⁵¹.

Primer extension by DNA polymerase
DNA polymerase activity through the extension of the 20 nt fluorescently labeled primer will result in a size increase of up to 40 nt when using the oligonucleotide set presented here (Figure 4A). Partially extended products will appear as a ladder of products up to the size of the template (Figure 4B and Figure 4C). The primer extension assay depicted here (Figure 4C) led to the complete synthesis of the primer strand.

Degradation by nuclease
Nuclease activity results in a reduction in the size of the fluorescently labeled oligonucleotide when visualized on a urea PAGE gel. Nucleases with specific endonuclease activity may result in a single 20 nt product (Figure 5A-C). Nucleases with exonuclease activity will result in fluorescently labeled oligonucleotides of varying sizes (Figure 5D-F). Here, the results of a bacterial Antarctic metagenome nuclease protein are shown, which exhibits specific endonuclease activity on the double-stranded abasic site substrate and non-specific exonuclease activity on the single-stranded substrate (Figure 5C and Figure 5F, respectively).

Ligase de-adenylation and nucleotide cofactor usage assay
Many recombinantly expressed ligase enzymes are purified in an already-adenylated state from ATP or NAD scavenged from their host^{51,52,53,54,55}. Removal of this covalent AMP intermediate is necessary for accurate identification of the range of the nucleotide cofactors that the enzyme can use for the ligation of nucleic acid substrates⁵¹. Turnover of the adenylated ligase enzyme with unlabeled substrate in the absence of an exogenous nucleotide cofactor will deplete the adenylation intermediate. By using fluorescently labeled DNA substrates and adding nucleotide cofactor in the second step only, any ligation product that is visible when using a fluorescent channel (Fluorescein-FITC) on the gel imager is the result of adenylation with the nucleotide cofactor added at this step.

Activity assays set up with a bacterial DNA ligase expressed in an E. coli system show that it was purified in a pre-adenylated form and exhibited a background ability to ligate the nicked DNA with no additional cofactor (Figure 6Ai-ii). The addition of ATP to the adenylated ligase does increase the amount of ligated product formed in the reaction, but this is not always reliable as this basal activity could give an incorrect impression of broad-spectrum cofactor usage or further additions of nucleotide cofactors to a reaction could inhibit ligation of the substrate. By first incubating the ligase with an unlabeled version of the nicked DNA substrate, the AMP is turned over, and no ligation is observed on the labeled DNA nick substrate unless ATP is added to the reaction. The second activity assay (Figure 6B) shows that the de-adenylated DNA ligase has improved ligation activity on nick DNA substrate with the addition of ATP and ADP. There is also some improvement in ligation with the addition of GTP, which is greater than the background ligation seen in the no-cofactor control. The ligation observed in reactions with NAD is comparable to the ligation seen in the no cofactor control reactions, which rules out NAD as a cofactor for ligation.

Results of dual-labeled assay
The orthogonal fluorophores can be used to establish activity in different parts of a complex substrate. The example in Figure 7 shows the activity of a recently described DNA ligase R2D⁵⁶ compared to T4 DNA ligase on a splinted DNA/RNA duplex. Ligation of the DNA acceptor to the RNA donor section of the substrate was detected using 6-FAM, while ligation of the RNA acceptor to a DNA donor was detected with 5-TAMRA (Figure 7A). The relative extent of DNA ligation to both ends of the RNA substrate could be evaluated by the presence of a product band with fluorescence in both the 6-FAM and 5-TAMRA channels, which appears as yellow in the composite image, while the individual 6-FAM and 5-TAMRA are green and yellow, respectively (Figure 7B). As detailed in a previous publication⁵⁶, only R2D ligase was able to ligate DNA to the 5'end of RNA, while both ligases were able to ligate DNA to the 3'end of the RNA. When ligating both DNA molecules to the RNA in single reaction mixtures, R2D ligase can ligate the DNA oligos to both ends of the RNA, as a band shifts upwards and a change of color to yellow is observed (Figure 7C).

The second example in Figure 8 uses a dual labeling approach to demonstrate the ligation of multiple oligonucleotides at different pH. The substrate, assembled from a mixture of seven oligonucleotides, was used to show the ability of the R2D ligase to assemble short fragments. The 5' strand has a FAM fluorophore-labeled 14 nt strand, while the complement strand has a TAMRA fluorophore on the 3' end of the 16 nt oligonucleotide. Successful ligation can be seen by additional higher bands for the FAM (green) oligo and the TAMRA (red) oligo. The overlap of both fluorophores gives the nucleotide band in yellow as completely ligated strands of the same size.

Results of DNA binding from a DNA ligase enzyme using EMSA
The binding of the DNA ligase enzyme to a phosphorylated nicked DNA results in a complex that runs at a higher molecular weight than the oligonucleotide alone (Figure 9A). Here, the binding of a bacterial ATP-dependent DNA ligase to nick DNA substrate is shown. At high protein concentrations, the majority of the labeled DNA substrate is bound and seen in the higher molecular weight bands; at lower protein concentrations, free DNA substrate predominates (Figure 9B).

Figure 1: Schematics of the different nucleic acid substrates designed for testing ligase enzymatic activity. Stars represent labeling with the 6- carboxyfluorescein at the 5' terminus (5' FAM). Labeled strands are indicated by a black or dark blue/red (in the case of DNA/RNA duplexes) line, while unlabeled portions of substrate duplexes are indicated by gray or light blue/red (in the case of DNA/RNA duplexes) line. Labeled strands are 20 nt, and if strands are ligated, they form a 40 nt product. 5' phosphorylated sites are indicated by a P in a circle. Oligonucleotide strands are labeled, and the names are referenced in Table 1 and Table 2. (A) Design of double-stranded DNA substrates with different types of breaks, used to test enzymatic activity of DNA ligases. (B) Design of RNA and RNA/DNA double-stranded substrates that incorporate DNA and RNA oligonucleotides. The substrates contain a single nick site and are used to test the ability of ligases to seal nick breaks of RNA/DNA duplexes. The red line represents RNA oligonucleotides, and the blue lines represent DNA oligonucleotides. Please click here to view a larger version of this figure.

Figure 2: Schematics of the different nucleic acid substrates designed for testing nuclease enzymatic activity. Stars represent labeling with the 6- carboxyfluorescein at the 5' terminal (5' FAM). Labeled strands are indicated by a black line, while unlabeled portions of substrate duplexes are indicated by a gray line. 5' phosphorylated sites are indicated by a P in a circle. Oligonucleotide strands are labeled, and the names are referenced in Table 1 and Table 2. (A) Design of substrates with double and single-stranded portions. (B) The design of double-stranded DNA substrates is modified to generate single-stranded flaps at the 3' or 5' end (Flapped 3', flapped 5', or flapped both ends (splayed)). (C) Design of double-stranded substrates incorporating damaged bases or mismatches at a central position. Please click here to view a larger version of this figure.

Figure 3: Testing ligation activity of a bacterial ATP-dependent DNA ligase DV-1-1-Lig. (A) Schematic of enzyme assays for ligase activity on DNA substrates with results analyzed on a TBE urea PAGE gel. Stars represent labeling with the 6-carboxyfluorescein at the 5' terminal (5' FAM). Labeled strands are indicated by a black line, while unlabeled portions of substrate duplexes that are not visible during analysis are indicated by gray lines. (B) Quantification of ligation by DV-1-1-Lig on nicked DNA at different time points. Activity against each substrate was carried out in duplicate. Reactions were incubated for different time periods (0.5 h, 1 h, 2 h, 3 h, and 4 h) at 25 °C, with 4 µM final protein concentration, 1 mM final ATP concentration, and 10 mM final magnesium ion concentration. (C) Ligation of nicked DNA substrate with magnesium (Mg) or manganese (Mn). Reactions were carried out for 3 h, at 25 °C, with 1 mM final ATP concentration and 10 mM final metal ion concentrations. (D) Results of ligation on different DNA substrates. Reactions were carried out for 8 h, at 20 °C, with 2 µM final protein concentration, 1 mM final concentration of ATP, and 10 mM final concentration of magnesium. The addition of protein to the reaction is indicated by a plus symbol (+). Control reactions are indicated by C, and the absence of contained protein is indicated by a minus symbol (-). Product (40 nt) and substrate (20 nt) are indicated by red arrows. The results of activity assays were visualized using an imaging system. Bands were quantified by integration of intensity using the ImageJ software⁴⁹ and graphs were generated using a graphing software. Please click here to view a larger version of this figure.

Figure 4: DNA polymerase assay. (A) Schematic of primer extension assay on a DNA substrate. Stars represent labeling with the 6-carboxyfluorescein at the 5' terminal (5' FAM). Labeled strands are indicated by a black line, while unlabeled portions of substrate duplexes that are not visible during analysis are indicated by gray lines. (B) Schematic of an anticipated TBE Urea PAGE gel results showing an increase in the length of the labeled primer with the addition of nucleotides. (C) Example of primer extension with E. coli Klenow fragment DNA polymerase enzyme. The addition of protein to the reaction is indicated by a plus symbol (+). Lane 1, 12.5 U; Lane 2, 2.5 U; Lane 3, 1.25U. Control reactions are indicated by a C, and the absence of contained protein is indicated by a minus symbol (-). Reactions were carried out for 15 min at 25 °C and contained 5 mM MgCl₂, 50 mM Tris pH 8.0, 50 mM NaCl₂, 1 mM DTT, and 0.25 mM dNTPs. Product (40 nt) and substrate (20 nt) are indicated by red arrows. The results of activity assays were visualized using an imaging system. Please click here to view a larger version of this figure.

Figure 5: Testing activity of a bacterial Antarctic metagenome nuclease using different substrates, metals, and incubation temperatures. (A) Schematic-specific endonuclease activity of a nuclease protein on the abasic site (indicated by an x) substrate. Stars represent labeling with the 6-carboxyfluorescein at the 5' terminal (5' FAM). Black lines represent the labeled strands, and gray lines represent the unlabeled DNA strands. Specific cutting by the nuclease is shown on the right side of the abasic site but may also occur on the left side. (B) Schematic of a denaturing urea PAGE gel showing uncut substrate (40 nt) in the absence of protein indicated by a minus symbol (-) and the cut product (20 nt), in the presence of a nuclease indicated by a plus symbol (+). (C) An example of a TBE urea PAGE gel showing results of specific endonuclease nuclease activity on a double-stranded DNA substrate with an abasic site. The assay was carried out with a protein concentration of 1.0 µM and 1 mM MgCl₂. Two control reactions were carried out at 20 °C, one without protein (well 1) and one without metal (well 2). The reactions were incubated for 5 h at increasing temperatures from 1 °C to 50 °C, as indicated above the gel image. The addition of the protein is indicated by a plus symbol (+), and the absence of the protein is indicated by a minus symbol (-). The substrate and product are indicated by red arrows. (D) Schematic showing non-specific exonuclease activity of a nuclease protein on a single-stranded substrate. Stars represent labeling with the 6-carboxyfluorescein at the 5' terminus (5'FAM). Black lines represent the labeled strands. Non-specific cutting of the nuclease results in labeled strands of varying lengths (<39 nt). (E) Schematic representing a denaturing urea PAGE gel showing uncut substrate (40 nt), where no protein was added (-), and showing cut substrate of varying lengths, where the protein was added (+) that may result from the activity of a non-specific exonuclease. (F) TBE urea PAGE gels show an example of non-specific exonuclease activity of the same Antarctic metagenome nuclease with a single-stranded substrate. The assay was carried out at 20 °C for 4 h with protein at a final concentration of 1.5 µM and a final concentration of the metal ion at 10 mM. The metal ion used is indicated above the image. The addition of the protein is indicated by a plus symbol (+), and the absence of the protein is indicated by a minus symbol (-). The substrate (40 nt) is indicated by the red arrow, and the products of different lengths are indicated by the red clamp. Results of all nuclease reactions were visualized using an imaging system. Please click here to view a larger version of this figure.

Figure 6: Ligation results of nicked DNA substrate by adenylated and de-adenylated versions of a bacterial DNA ligase. (A) i) Quantification of ligation by adenylated and de-adenylated ligase enzyme, with and without ATP. Points on the graph represent averages of each reaction replicate. Standard error bars are included. ii) TBE urea PAGE gel showing the results of ligation on nick DNA substrate by adenylated and de-adenylated ligase enzyme, with and without ATP. Reactions were performed in triplicate. The de-adenylated enzyme was pre-incubated for 2 h at 25 °C with unlabeled nick DNA substrate and 0.1 µM enzyme, followed by a 1 h incubation with labeled nick DNA substrate. The adenylated enzyme was not pre-incubated with the unlabeled substrate. ATP (1 mM) was added to reactions alongside the addition of labeled DNA substrate. (B) i) Quantification of ligation by the de-adenylated ligase on nick DNA with different cofactors (ATP, NAD, ADP, and GTP). Points on the graph represent averages of each concentration. Standard deviation error bars are included. ii) TBE urea PAGE showing results of ligation by the ligase, with and without the addition of different cofactors. Activity against each substrate was carried out in duplicate. Reactions were pre-incubated for 2 h at 25 °C with unlabeled nicked DNA substrate, 4 µM protein, and 5 mM magnesium ion, followed by a 4 h incubation with the addition of labeled nicked DNA substrate, 5 mM magnesium, and different cofactors at 1 mM final concentration. The addition of protein to the reaction is indicated by a plus symbol (+). Control reactions were used that did not contain protein (Controls, C) or cofactor (No cofactor). Product (40 nt) and substrate (20 nt) are indicated by red arrows. The results of activity assays were visualized using an imaging system. Bands were quantified by integration of intensity using the ImageJ software⁴⁹. Graphs were generated using a graphing software. Please click here to view a larger version of this figure.

Figure 7: Ligation of short fluorescent-labeled DNA oligonucleotides to either end of a 5' phosphorylated RNA oligo, positioned by DNA templates. (A) Schematic of DNA-splinted DNA-RNA ligation showing how possible ligation products can be visualized using a dual-labelling strategy. (B) Schematic of possible products expected to be seen on a Urea-PAGE gel imaged in two channels. (C) Experimental results from dual fluorophore ligation. The ligation reactions were tested using T4 DNA Ligase and R2D ligase, both in isolated reactions where DNA was ligated to either the 5' end (lane 2-4) or the 3'end (lane 5-7) of the RNA or in a single reaction where both DNA oligonucleotides were ligated in the same reaction mixture (lane 8-10). No protein control (NPC) had no enzyme added. The reaction conditions and the specific oligonucleotides used in this example are the same as given in⁵⁶. Please click here to view a larger version of this figure.

Figure 8: Ligation of multiple short oligonucleotide segments with three R2D variants at different pH. (A) Schematic of multi-part ligation strategy. (B) Representative gel showing pH dependence of a multi-part assembly, imaged with both FAM and TAMRA channels. Lane C no enzyme control, Lanes 1-3 pH 6.0, Lane 4-6 pH 6.5, Lane 7-9 pH 7, and Lane 10-12 pH 7.5. Please click here to view a larger version of this figure.

Figure 9: EMSA results. Schematic and native TBE PAGE gel of an Electrophoretic mobility shift assay (EMSA) visualizing the binding activity of an ATP-dependent DNA ligase (sourced from metagenome data from the M^cMurdo Dry Valleys, Antarctica) on a DNA substrate. (A) Schematic of EMSAs/DNA binding activity by a DNA ligase on nick DNA substrate. Binding results are visualized on a non-denaturing TBE PAGE gel. DNA substrates bound to ligase enzyme migrate slower through the gel, while enzyme-free substrates migrate faster, resulting in a shift of band size (yellow boxes) visualized on the gel. As the substrates are visualized on a non-denaturing gel, the oligonucleotides will remain annealed, allowing the FAM-labeled strand (yellow star) to be present on both free and bound substrates. A protein concentration gradient can be used to determine protein concentration for optimum binding. (B) Visualization of an EMSA on a native TBE PAGE gel, showing the binding ability of an ATP-dependent DNA ligase to a nick DNA substrate at different protein concentrations. The enzyme was incubated with nicked DNA substrate for 30 min at 25 °C, with 1 mM final ATP concentration and 40 mM EDTA. Three different protein concentrations were used (5.5 µM, 2.3 µM, and 1.5 µM). Reactions were run out on a 10 % native TBE gel, with native loading dye in the reactions. Samples containing protein were run out in triplicates for different protein concentrations. The control lanes contain nicked DNA substrate with 1 mM, final ATP, and EDTA but no protein. Please click here to view a larger version of this figure.

Table 1: DNA substrate oligonucleotide sequences used to construct assay substrates. Damages are indicated in bold and underlined. Positions of mismatches and gaps in the final duplex are underlined. Abbreviations: 5' 6-carboxyfluorescien (5' FAM), 8-Oxo-deoxyguanosine (8OxodG), abasic tetrohydrofuran (dSpacer). Please click here to download this Table.

Table 2: Combinations of oligonucleotides used to assemble master mixes for different enzyme assays. Please click here to download this Table.

Table 3: Set up of assay master mixes with up to four oligonucleotides. See Table 2 for oligonucleotide combinations to use for different assay substrates and text for details of reaction conditions. At a minimum, a labeled oligonucleotide is required- additional portions of the duplex are indicated by brackets. Please click here to download this Table.

Table 4: Components for a labeled and unlabeled master mix for making nick DNA duplex. Please click here to download this Table.

Table 5: Set up of EMSA master mixes for different combinations of duplexes. See Table 3 for oligonucleotide combinations to use for different EMSA substrates and text for details of reaction conditions. The EMSA substrates are designed using the same oligonucleotides as those for the activity assays. Please click here to download this Table.

Discussion

Critical steps in the protocol
Oligonucleotide design and purchase: When purchasing the oligonucleotides for duplex formation, it is essential to consider sequence design. It is recommended to use an oligo analyzer tool to predict properties of the nucleotide sequence, such as GC content, melting temperature, secondary structure, and dimerization potential, before ordering⁵⁷.

Assembly and annealing of nucleic acid duplexes: When preparing RNA/RNA-DNA duplexes, care should be taken to prevent RNase contamination by using reagents and water that are RNase-free or adding an RNAse inhibitor (so long as the enzyme under investigation is not itself an RNAse). Equipment, gloves, and bench space should be treated with an RNase decontamination solution before assembling the master mixes to avoid degradation of the oligonucleotides⁵⁸. Master stocks of the RNA oligonucleotides can be made up in smaller batches in case of contamination.

DNA ligases always require a nucleotide cofactor, either ATP or NAD, at a final concentration between 0.1 and 1 mM. DNA polymerases require dNTPs for strand synthesis, and these are usually added in equimolar ratios at 50 µM of each. When preparing master mixes for ligase and polymerase duplexes, nucleotide cofactors, such as ATP, NAD, and dNTPs, must be added to the master mix after the mixture has been heated and cooled to room temperature, as they are heat sensitive.

The 10x buffer for annealing the oligonucleotides requires the addition of DTT, which has a short half-life and ideally needs to be prepared fresh or frozen before use. Smaller aliquots of the 10x buffer can be prepared in advance and stored at 4 °C, with the addition of DTT before use.

Oligonucleotide stocks and master mixes should be stored at -20 °C for long-term use, and fluorescently labeled oligonucleotides or mixes wrapped in foil to prevent photobleaching of the fluorophore. To avoid repeated freeze-thaw cycles which can compromise the oligonucleotide integrity, it is advisable to prepare and use smaller volumes (>1 mL).

Assay setup and analysis of results: The inclusion of controls is essential for the correct interpretation of assay results. A no-protein control (DNA substrate with buffer solution) should be used to ensure that apparent activity on the substrate is the result of the enzyme. In the case of nuclease, exogenous substrate degradation will cause low molecular weight bands or laddering in control, while in the case of ligase or polymerase, incomplete duplex denaturation will result in a high-molecular-weight band in control. A positive control of the size of the expected product is provided in Table 2 and should be included to verify the correct product size.

Important parameters in successful electrophoresis of the assay product have been described in a previous protocol⁵⁹. These include ensuring the percentage of acrylamide is appropriate for the substrate size and the use of an appropriate loading dye depending on the size of the substrate and the type of gel required for the visualization of results. The tracking dye bromophenol blue runs at the approximate size of a 10-base oligonucleotide and is therefore suitable for the oligonucleotide set described here; however, if there is a band of dye overlapping with the fluorescent product or substrate bands when using shorter nucleotides, the tracking dye may be exchanged to xylene cyanol, which runs more slowly⁶⁰.

To ensure good resolution of substrate/ product bands, the gel tank should be either externally heated using a water bath or by running a high voltage (e.g., 180 V). Pre-running urea PAGE gels for at least 30 min and flushing the wells with a pipette to remove excess urea is essential to obtain sharp, straight bands suitable for quantification.

For quantification of substrates and products, a range of 20%-80% of the total material should be converted to product to allow accurate integration of the fluorescent gel bands. If assaying a new enzyme, it is recommended to run a gradient of protein concentrations to allow subsequent measurements to be made in this range. Many fluorescence imagers have in-built band integration features that may be used for this purpose. If using an external program like ImageJ⁴⁹, ensure that the gel image is exported at high quality in tiff format.

Modifications and troubleshooting of the technique
Oligonucleotide design and purchase: The oligonucleotides given in this protocol are useful starting points and can be expanded, as suggested in the methods section, by altering the oligonucleotide sequences and incorporating modifications.

Standard assay setup and analysis of results: Pilot experiments are essential when testing the activity of a new enzyme as the optimal reaction conditions will vary between different proteins; the parameters given in this protocol are intended as a starting point only. Factors such as enzyme concentration, reaction temperature, duration of incubation, cofactor, and buffer composition/pH may be varied during the master mix and assay reaction assembly. Cofactor concentrations are especially important as an excess of either nucleotide cofactors or metal ions can be inhibitory, and it is advisable to test a range of concentrations of each. For DNA polymerases, recommended dNTP concentrations range from 20 to 200 µM total however, high dNTP concentrations decrease the fidelity⁶¹. Likewise, polynucleotide ligases are typically assayed with 0.1-1.0 mM nucleotide cofactor, but excessive concentrations cause accumulated DNA to adenylate and reaction inhibition⁶².

The standard protocol here uses Mg as a divalent cation; however, some enzymes are more active with manganese (Mn) or zinc (Zn). Master mixes containing Mn or Zn should be prepared with a lower pH buffer (<7.5), and metals should be added after heating. Master mixes should be used on the same day to avoid precipitation and inhibition of the enzyme activity.

Additives, including crowding agents such as polyethylene glycol (PEG) or stabilizers such as bovine serum albumin, can be used to increase the activity of some enzymes^63,64. These should be added after the annealing step due to their thermolability. All biologically derived reagents should be molecular biology grade to avoid introducing contaminating nucleases.

The volume of formamide quench buffer used in this protocol is suitable for the basis set of oligonucleotides. However, for longer oligonucleotides, the volume of formamide should be increased to between 10 µL and 20 µL to ensure complete strand separation, and quenched reactions can be snap-cooled on ice and stored at 4 °C until electrophoresis.

With some enzymes, tight binding of the protein to substrate/ product may preclude migration of the oligonucleotides into the gel during electrophoresis, especially when the enzyme is used at high concentrations. This will appear as a lack of product/ substrate bands on the gel and often an accumulation of fluorescent material (product and/or substrate complexed with protein) in the wells. To disrupt these complexes, the assay mixture can be treated with Proteinase K before the addition of quench buffer, or Sodium Dodecyl Sulfate (SDS) can be added during quenching.

DNA ligase activity produces two discrete bands, which are straightforward to quantify. It is possible to use the same method to quantify endonuclease activity on DNA damage substrate where a single 20 nucleotide product is anticipated or EMSA assays where the bound/ unbound states are clearly distinguished. Accurate quantification of polymerase and processive nuclease activity is difficult by band integration, and in general, it is recommended to use real-time assay with probes or beacons, as described in the introduction.

De-adenylation and cofactor specificity of DNA ligase: Some optimization of this protocol may be required to remove all background activity, which will then be restored by the addition of the nucleotide cofactor(s). If background ligase activity remains after de-adenylation (seen as ligation in the no-cofactor control), the incubation time of the de-adenylation step can be increased, or the concentration of ligase decreased. Ensure the DNA nick duplex with no label is more concentrated than the protein (> 4:1::substrate: protein). If the addition of a cofactor in the presence of labeled DNA does not restore ligation for any of the cofactors tested, it is possible the enzyme has denatured, and the de-adenylation step should be shortened. For a new ligase or a new batch of recombinant enzyme, it is useful to determine the extent of background ligation with the adenylated enzyme in the absence of additional nucleotide cofactors. This involves setting up another reaction where 10 µL of un-labeled master mix, 10 µL of labeled master mix, and 2.5 µL of a fresh enzyme that has not been de-adenylated are incubated simultaneously.

Use dual-labeled substrates: This protocol requires the design and purchase of an oligonucleotide with a florescent moiety that has a different excitation/ emission spectrum to the fluorophore originally used and then assembly of this into a duplex so that the two orthogonal fluorophores label different portions of the substrate. If in doubt, a number of online spectral viewers are available for checking excitation and emission spectra, for example, FPbase Spectra Viewer and BD Spectrum Viewer, which can be used to check both the compatibility of the fluorophore pair and the most appropriate imager settings. The specific settings used for visualizing the gel will depend on the model of the imager and the fluorophore(s) used. The composite image will record the signal from both channels on the single gel, and the band where both fluorophores are attached to the same product will show as an intermediate color. Separate images from individual channels can be analyzed to quantify the extent of ligation or degradation on differentially labeled strands.

Limitations of the technique
While working with fluorescently labeled substrates has many advantages, such as low health hazards, there are some limitations one must consider when working with these types of labeled substrates. One of them is the relatively high cost of ordering these oligonucleotides with fluorophore labels or modifications, such as the addition of damaged bases, compared to in-house radiolabeling. Another is that specialized imagining equipment and software are required for the detection of labeled strands. However, these features are now included with most new gel imagining systems. One should also consider the potential for the fluorophores to alter the binding properties of the target interaction proteins. Fluorescein and its derivative FAM are commonly used to label oligonucleotides and are sensitive to changes in pH and nucleobase-quenching⁶⁵. These limitations should be carefully considered when deciding the type and placement of fluorophores on the oligonucleotide for duplex design.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
30% Acrylamide/Bis Solution (29:1)	BioRad	1610156
Adenosine triphosphate (ATP)	Many suppliers
Ammonium persulfate (APS)	Many suppliers
Benchtop centrifuge	Many suppliers
Borate	Many suppliers
Bromophenol blue	Many suppliers
Dithiothreitol (DTT)	Many suppliers
Electrophoresis system with circulating water bath	Many suppliers
Ethylenediaminetetraacetic acid (EDTA)	Many suppliers
Fluoresnence imager, e.g. iBright FL1000	Thermo Fisher Scientific	A32752
Formamide	Many suppliers
Gel casting system	Many suppliers
Heating block	Many suppliers
Magnesium Chloride	Many suppliers		Other metal ions may be preferred depending on the protein studied
Microcentrifuge tubes (1.5 mL)	Many suppliers
Micropipettes and tips	Many suppliers		1 mL, 0.2 mL, 0.02 mL, 0.002 mL
Nicotinamide adenine dinucleotide (NAD+)	Many suppliers
Oligonucleotides	Integrated DNA Technologies	NA	Thermo Fisher, Sigma-Aldrich,  Genscript and others also supply these
pasture pipette	Many suppliers
PCR thermocycler	Many suppliers
PCR tubes	Many suppliers
RNAse away	ThermoFisher	7002PK	Only needed when working with RNA oligos
RNase AWAY	Merck	83931-250ML	Surfactant for removal of RNAse contamination on surfaces
RNAse-free water	New England Biolabs	B1500L	Only needed when working with RNA oligos
Sodium Chloride	Many suppliers
SUPERase IN RNase inhibitor	Thermo Fisher Scientific	AM2694	Broad spectrum RNAse inhibitir (protein-based)
SYBR Gold	Thermo Fisher Scientific	S11494	This may be used to post-stain gels and visualise unlabelled oligonucleotides
Tetramethylethylenediamine (TMED)	Many suppliers
Tris, or tris(hydroxymethyl)aminomethane	Many suppliers
Ultrapure water (Milli-Q)	Merck
urea	Many suppliers
Vortex	Many suppliers