Insight into the biochemical mechanism of DNA helicases provided by bulk-phase and single-molecule assays

Abstract

There are multiple assays available that can provide insight into the biochemical mechanism of DNA helicases. For the first 22 years since their discovery, bulk-phase assays were used. These include gel-based, spectrophotometric, and spectrofluorometric assays that revealed many facets of these enzymes. From 2001, single-molecule studies have contributed additional insight into these DNA nanomachines to reveal details on energy coupling, step size, processivity as well as unique aspects of individual enzyme behavior that were masked in the averaging inherent in ensemble studies. In this review, important aspects of the study of helicases are discussed including beginning with active, nuclease-free enzyme, followed by several bulk-phase approaches that have been developed and still find widespread use today. Finally, two single-molecule approaches are discussed, and the resulting findings are related to the results obtained in bulk-phase studies.

1. Introduction

DNA helicases are an essential group of motor proteins that unwind DNA duplexes into their component single strands in a process that is coupled to the hydrolysis of nucleoside 5’-triphosphates [1–6]. The purpose of DNA unwinding is to provide nascent, single-stranded DNA (ssDNA) for the processes of DNA repair, replication, and recombination [7]. All DNA helicases examined to date, share several common biochemical properties, including the binding of single- and double-stranded DNA (dsDNA), nucleoside 5’-triphosphate (NTP) binding, and hydrolysis, and NTP hydrolysis-coupled, polar unwinding of duplex DNA into its component single strands [8]. Although the outcome of the action of DNA helicases is the same (i.e., dsDNA unwinding), how they achieve this goal is quite different [9]. This is dictated by the in vivo role of the enzyme, its oligomeric structure, and the partner proteins with which it interacts [4, 10, 11].

Over the 41 years since their discovery, assays to provide mechanistic insight into DNA helicases have evolved from demonstrating the ability to strand separate and utilize energy to enable strand separation, to more complex single-molecule assays that provide both visual insight as well as knowledge of the forces associated with motor protein function [12–19]. In this review, I summarize these approaches and the insight they have provided into the biochemical mechanism, highlighting the advantages and drawbacks of each approach. This review deals exclusively with enzymes that strand separate, that is, bonafide DNA helicases.

2. The importance of nuclease-free, active protein

One of the most critical aspects of studying DNA (and RNA) helicases is to use protein preparations that are free of contaminating nucleases, and which contain high levels of active protein. Many approaches can be taken to achieve nuclease-free protein preparations that include at a minimum, column chromatography steps, and ammonium sulfate fractionation. Further details of these and other approaches can be found in Methods in Enzymology Volumes 182 and 463. In addition, protein tagging such as histidine or fusion to maltose-binding protein to name just a few can also be used [20]. Tagging approaches must be used with care as they can induce oligomerization and as such, care must be taken to demonstrate that the tag does not influence activity, or the plasmid construct must be designed with the tag removal in mind [21–25]. Here tag removal can be done by proteases or chemical means [24, 26].

Exo- and endonucleases can influence helicase activity by removing DNA strands, eliminating fluorophores, introducing nicks, or destroying substrates altogether. In addition, the removal of one or more strands by a nuclease could influence the interpretation of data, resulting in falsely assigning helicase activity to a protein. Therefore, once homogeneous preparations of the enzyme have been obtained, assays to ascertain the presence of nuclease activity are essential. This is important because many studies use protein expressed in E.coli, there are 17 exonucleases in this organism, and Exonuclease I, which is not readily visualized by Coomassie staining at low concentrations, may still be present in the final pool of purified protein and demonstrate activity [27, 28]. Assays to test for the presence of nucleases include incubating purified protein in magnesium ion-containing buffers with 5ènd-labeled duplex DNA (either oligonucleotide or plasmid length) and separately, 5ènd-labeled single-stranded DNA (typically oligonucleotides are used). The reactions are then mixed with a loading buffer containing SDS and Proteinase K and then subjected to electrophoresis in either agarose (plasmid DNA) or polyacrylamide (oligonucleotide) gels. Quantitative analysis to ascertain the loss of radioactive signal must then be done. If a significant signal loss is observed, additional steps in the purification of the helicase must be added or individual steps adjusted to eliminate nucleases. Following the re-purification of the protein, nuclease assays must be repeated to ensure that these contaminants have been removed.

Nuclease-free protein preparations are then assayed to determine levels of activity of the purified helicase. This knowledge is critical as inactive protein may influence binding and activity of the active fraction of protein resulting in misinterpretation of experimental results. The levels of activity can be determined using ATPase assays (which also provide insight into the requirements for DNA for activity, see below) and one or more helicase assays. The assays used are dictated by the enzyme being studied as some require tailed substrates, others demonstrate optimal activity on forked DNA molecules while some require duplex DNA with free ends. For example, the optimal substrate for RecBCD is linear duplex DNA and titration of total protein relative to DNA ends reveals the level of the active enzyme [29]. Initial studies produced protein that was 30% active and this was improved to 100% active by the inclusion of a monoQ chromatography step which removed the inactive fraction [30]. Similar effects were observed for RecG when the DEAE column was replaced with Q-sepharose and buffer conditions that had caused RecG to flow through the former were changed to permit binding and fractionation to the latter [19, 31].

To determine the level of active protein, ATPase or helicase assays are done with a fixed concentration of DNA, typically in excess of the K_d, and protein titrated. In these assays, as the concentration of protein increases, a corresponding increase in enzyme activity should be observed. This will reach a maximum when available DNA binding sites required for initiation of the helicase have been saturated. The addition of more protein beyond the saturation point cannot produce a further increase in activity as binding sites have been saturated. The point at which saturation occurs is the stoichiometry of helicase to either dsDNA ends, fork junctions, or ssDNA tails. If the stoichiometry is 1:1 (protein:DNA) then the enzyme preparation is 100% active. If the K_d for DNA is unknown, titrations can be done at different concentrations of DNA and the level of activity should be consistent in all experiments.

3. Bulk-phase assays

3.1. Gel-based assays

Before loading samples onto gels (either time point aliquots or reaction endpoints), they should be mixed with loading buffer containing EDTA, SDS, and Proteinase K. This cocktail eliminates binding of the helicase to DNA by chelating magnesium ions, denatures protein(s) present in the assay and then cleaves denatured proteins into small fragments. This eliminates residual binding that can influence the production of clear gel results. This cocktail should not be used for electrophoretic mobility shift assays (EMSA; [32]). For detailed methods using the various substrates below the reader is referred to the following references: M13 [33]; lambda DNA [34]; plasmids [35–37]; oligonucleotides [30, 38–40] and DNA substrate purification [41].

3.1.1. Large DNA substrates

Large DNA substrates are herein classified as DNA molecules that are longer than 200 base pairs. These can be partial duplexes consisting of an oligonucleotide annealed to a single-stranded circular DNA molecule such as M13 (Fig. 1A) or, they may be intact, linear duplex DNA (Fig. 1B). Typically, linear duplex is plasmid sized (3–10 kB) but larger molecules such as bacteriophage lambda DNA (48.5kB) can also be used, although lambda DNA ends are cohesive and require heating to 70°C to denature before being used in the assay and also once reactions are complete but before electrophoresis [42]. The formation of linear duplex substrates from covalently closed circular plasmids can be achieved with restriction endonucleases that produce ends that contain 5ˋ- or 3ˋ-overhangs or be blunt, as this may influence helicase binding and resulting activity.

Figure 1. DNA substrates for helicase assays.

The 5ènd-labeled DNA substrates shown are grouped according to the type of gel used to analyze unwinding. For all substrates, the “*” indicates the 5’-end of each strand of DNA. Substrates boxed in orange require agarose gels (typically 1%) while those boxed in yellow require polyacrylamide gels with optimal separation of products from intermediates and substrates being achieved using bis:acrylamide ratios of 1:15 [38]. A, bacteriophage ssDNA is annealed to a 5ènd-labeled oligonucleotide. B, linear duplex DNA, either plasmid or bacteriophage lambda. C -F, oligonucleotide length DNA substrates are shown. C, substrates to determine the requirement for an ssDNA tail to initiate unwinding as well as the polarity of translocation and duplex unwinding for an enzyme that can initiate unwinding on ssDNA. The generic helicase is shown as a grey sphere. D, Substrates to determine the direction of translocation and unwinding for an enzyme that initiates the reaction from a duplex end. E, Substrates that mimic a nascent, stalled DNA replication fork: i., both arms are single-stranded; ii., a substrate with a gap in the nascent lagging strand and iii., the gap is in the nascent leading strand. F, Substrates that mimic a regressed fork (i) or Holliday junction (ii). G, Helicase assays using M13 ssDNA substrates to show DNA unwinding by the MCM4/6/7 complex. Here, two substrates were made with equal length duplex DNA regions: A.i, is blunt ended whereas A.ii has a 5ˋ-ssDNA tail. This figure was adapted from [33] and used by permission “Copyright (2001) National Academy of Sciences”. H, DNA unwinding using 5ˋ-end labeled linear plasmid DNA 4,314 bp in length [35]. This figure was adapted from [35] and used by permission “Copyright (1998) National Academy of Sciences”. I and J, helicase assays to determine the polarity of translocation for RecBC which initiates unwinding at a blunt end [13]. The substrate is shown at the top, gels in the middle and quantitation at the bottom of each panel.

To visualize unwinding, DNA must be radioactively labeled with either the 5ˋ-ends of the annealed oligonucleotides of the partial duplexes or those of the linear substrates being used. During the action of the helicase, the unwound strands may be trapped using either a nucleic acid [43] or a single-strand DNA binding protein (ssb; see the section on reporter proteins below). The effects of the trap on helicase activity must also be determined as some helicases are inhibited by excess oligonucleotide (bacteriophage T4 UvsW) or the ssb itself (SMARCAL1) [44, 45]. Once reactions are complete, samples are subjected to electrophoresis in agarose gels (commonly 1% in TAE buffer) at low voltage, usually <1 V/cm of gel length [46]. Gel boxes that are longer than they are wider are recommended as these minimize the effects of heating during electrophoresis that can cause duplex separation to give the appearance of helicase activity. Following electrophoresis, gels are dried onto DE81 and Whatman paper simultaneously for the partial duplexes or Whatman paper for plasmid-length DNA. Historically, the DE81 paper was used to bind to displaced oligonucleotides enabling their quantification. For assays involving linear duplexes, DE81 paper is used only when the loss of single strands is suspected as in the case of enzymes such as RecBCD where helicase and nuclease activity are coupled [47]. However, as DE81 paper is no longer manufactured, alternative papers may be used such as Hybond-XL [48]. Assays using long DNA substrates can be used to reveal the presence of helicase or nuclease activity (both intrinsic to the enzyme such as that seen for RecBCD as well as contaminating endonucleases) and can also be used to provide insight into processivity [34, 35]. Here, processivity is defined as the number of base pairs unwound per enzyme binding event.

For the partial duplexes, the substrate migrates at the top of the gel near the wells while the displaced oligonucleotide migrates closer to the dye front (Fig. 1G). For linear duplex DNA, the unwound single-stranded products migrate faster than the substrate (Fig. 1H). For both substrates, the conversion of substrate to product is readily visualized and quantitation of the radioactive signal provides insight into reaction rate – substrate disappearance, intermediate formation and disappearance, and product formation. Assays using these DNA substrates can provide helicase kinetic parameters for both wild-type and mutant enzymes and can also be used to determine the effects of chemo-therapeutics on helicase activity [49–51]. In the first example shown in Fig. 1G, M13 ssDNA substrates were annealed to make a blunt-ended duplex region (Fig. 1A.i) or a duplex region with a 5’-ssDNA tail (Fig. 1A.ii). These substrates were used to determine the requirement for an ssDNA tail for initiation (substrate A.i was not unwound) as well as the polarity of translocation (substrate A.ii is unwound). In the second example, linear plasmid DNA was used to provide insight into the role of the RecB nuclease domain in the helicase activity of the RecBC and RecBCD enzymes. Here, the RecB_1–929C (RecB subunit lacking the 30kDa nuclease domain) was able to unwind dsDNA at 10 mM MgCl₂ but not at 3.5mM (Fig 1H). In contrast, the RecB_1–929CD and RecBC enzymes are efficient helicases at low [Mg²⁺].

3.1.2. Oligonucleotide DNA substrates

DNA helicases assays using small molecules less than 100 bp in length have become commonplace in the field. They are generally not suitable to determine the processivity of the enzyme due to insufficient lattice length but for some helicases, with limited processivity, they can be used. Further, and while they are easy to design and assemble, they are not without their problems as each strand requires purification, and complexes must be created by annealing, followed by electrophoresis to evaluate annealing efficiency. Here, one or more strands are 5ènd-labeled, and the use of acrylamide gels with a bis:acrylamide ratio of 1:15 is recommended as this provides optimal separation of the substrate from various intermediates that may not be achieved using the more common 1:19 ratio [38]. The ratio of 1:15 is also recommended for the helicase assays to follow [38]. In addition, if inhibition by excess single-stranded molecules used to ensure annealing is suspected, annealed complexes can be purified by excising bands following gel electrophoresis (a tedious and potentially substrate-damaging process, particularly if ethidium bromide staining is used) or, purification can be achieved using column chromatography employing non-porous ion exchange resins [22, 41].

The types of DNA substrates employed, vary depending on the enzyme being studied or the type of information required. First, a blunt-end duplex should be used to determine a requirement for a single-stranded tail for the initiation of translocation and DNA unwinding (Fig. 1C, top). Then, as helicases often track on only one strand of the DNA duplex and each of those strands has a polarity defined by the phosphodiester backbone, tailed DNA duplexes provide insight into which strand is the preferred strand for translocation (Fig. 1C). If the enzyme translocates in a 5ˋ to 3ˋ direction, it will bind to the middle substrate in panel 1C and translocate up to the duplex and proceed to unwind. In contrast, this same enzyme will bind to the lower substrate and translocate away from the duplex region and dissociate, leaving the dsDNA intact, i.e., it will not be unwound. The reverse scenario applies to an enzyme that translocates on the opposite strand of the duplex. These DNA substrates can only be used for DNA helicases that initiate unwinding on ssDNA such as the E. coli proteins DnaB, Rep, and UvrD [39, 52, 53].

In contrast, if the enzyme initiates reactions from a dsDNA end (e.g., RecBCD or RecBC), then gapped DNA substrates are required to determine translocation polarity (Fig. 1D; [29]). In the example shown, a 25 base oligonucleotide is annealed to one end of a 100-base oligomer. At the same time, a 20 nt length “tester molecule” is annealed downstream on the same strand so that a gap, longer than the footprint of the enzyme, is positioned between the two duplex regions. Using these substrates, the enzyme will bind to the dsDNA end and initiate translocation and unwinding up to the gap. If the enzyme translocates in the 3ˋ to 5ˋ direction, when the protein encounters gaps, they will be bypassed regardless of size since they are in the opposite strand and then proceed to unwind the tester oligonucleotide downstream. As a result, the three-strand substrate is unwound into its component strands (Fig. 1I). In contrast, this same enzyme will dissociate when it encounters a gap larger than its footprint in the translocation or top strand. Consequently, the distal oligonucleotide cannot be unwound and the three-strand substrate is converted into an intermediate with the 20mer still bound (Fig. 1J) [38].

In addition to enzymes that initiate at ends or on exposed regions of ssDNA, there is a growing group of atypical DNA helicases that demonstrate preferential binding to forked DNA molecules [54, 55]. Many of these are involved in DNA transactions at replication forks either during the process of replication (DnaB; MCM complex; phage SPP1 G40P), during the rescue of stalled forks (PriA, RecG, UvrD, DinG, Rep, and SMARCAL1), or during fork processing and/or recombination intermediates (RuvAB) [56–63]. To provide insight into the biochemical mechanism of these enzymes, forked DNA substrates are usually used (Fig. 1E and F). Assays employing these DNA molecules use two different forms – those that have fixed junctions containing heterologous sequences in the arms and thus cannot branch migrate and those that are fully homologous and can be branch migrated. In addition, DNA strands can be selectively labeled or, two or more strands labeled at the same time to enable tracking of various intermediates and formation of products. Finally, the length of each fork arm and length of duplex regions within each arm are also important considerations as this may influence enzyme binding (which would be poor if the loading arm is significantly less than a footprint) and the initiation of unwinding (which may be influenced by gaps in one strand or the other).

The use of forked or branched DNA molecules must take into consideration that these DNA molecules are dynamic structures and may not resemble the planar schematics in Figure 1. This follows because they can adopt different configurations dictated by the solution conditions and this can influence protein binding and subsequent helicase activity. This is true for both forked DNA structures as well as the canonical Holliday junction [64–69]. Furthermore, recent work has shown that there is local backbone breathing at fork junctions and this may influence helicase binding and subsequent activity [70–72]. Collectively these data indicate that the design of fork substrates and the interpretation of data from helicase assays must be done with the utmost care.

The substrates in panel 1E resemble a stalled fork with gaps in the fork arms. Consequently, they are typically used to infer whether the enzyme prefers substrates with a gap in the nascent lagging (substrate ii) or leading strand (substrate iii). In addition, they have also been used to provide insight into the specificity of the strands at the fork as shown for PriA which requires the presence of the 3ˋ-OH group of the nascent leading strand exactly at the fork to initiate ATPase activity (Fig. 1E, fork ii) [60, 73, 74].

The DNA molecules in panel 1F resemble regressed fork structures (i and ii) or recombination intermediates (ii; the Holliday Junction). They have been used extensively to study fork processing by a variety of enzymes and in the analysis of branch migration by enzymes such as RuvAB and the fork helicase RecG [40, 75, 76]. Due to the ability to adopt multiple configurations and the increased number of dsDNA ends, care must be taken when interpreting results from studies using the DNA substrates as shown for RuvAB [76].

A critical aspect of the study of helicases is knowledge of the amount of active protein and stoichiometry of binding. For the fork rescue helicase RecG, a stoichiometry of 1 enzyme per fork is obtained from enzyme titrations relative to the forks in Figure 1E and F, whereas for UvrD, no stoichiometry is obtained suggesting multiple binding sites (either at ends, the junction itself or to internal sites) or possible oligomerization as well as significant ATPase activity on nascent ssDNA [22, 76–78]. In the absence of this knowledge, assays done with a vast excess of protein over DNA can result in misleading data. In contrast, when carefully done, the role of an enzyme in vivo may be clarified as shown for RuvAB where its ability to process three-arm forks has been disproved (Fig. 1F, fork i) and additional potential roles defined as shown for UvrD [76, 77, 79].

Finally and in addition to radioactive labeling of oligonucleotide ends, these small substrates can also be obtained with various fluorophores or fluorophore quenchers attached to oligomer ends and/or at internal sites. These ssDNA molecules can be annealed and used in bulk-phase helicase assays using polyacrylamide gels, in a spectrofluorometer, or single-molecule assays [18, 63, 80–87].

3.2. Time courses versus protein titrations

The titration of protein relative to DNA substrate reveals the level of active enzyme in protein preparations. This makes sense when enzymes bind uniquely to junctions, dsDNA ends or ssDNA tails, as observed for RecG, RecBCD and Rep respectively. In some cases, enzymes bind to the duplex regions of substrates as well and as such obtaining precise values for the stoichiometry of binding is next to impossible. This was observed for UvrD where a stoichiometry relative to fork DNA could not be achieved whereas a 1:1 ratio was seen for RecG [22, 76]. Regardless, there are many publications in the literature where a vast excess of protein relative to DNA is used and substrate specificity inferred. Due to the potential complications arising from these reaction conditions, interpretation of these data should be done with great care. This follows because the excess protein can bind to both multiple sites on substrates concurrently as well as to intermediates and/or products and influence reaction outcome. Instead, knowledge of the affinity of an enzyme for each fork structure can be obtained from EMSA, filter binding, or surface plasmon resonance studies [60, 88–90]. In addition, footprinting of the helicase to each component of the substrate can be done as well as to the fork structure to define the binding sites [91]. In some cases, this produces a single binding site such as that observed for RecG, while in other cases multiple binding sites were observed for SMARCAL1 and PriA [45, 92]. Once this information is obtained, then careful kinetic studies can be done as described below.

In these assays, and because helicases are enzymes, an excess of DNA over protein (that is catalytic assay conditions) should be used and time courses where the conversion of substrate into intermediates and products can be observed should be used. For oligomeric helicases, the protein concentration used should be high enough to form the oligomer [93, 94].

3.3. Assays to determine NTP utilization

Helicases translocate along nucleic acid lattices using the energy released by the hydrolysis of nucleoside triphosphates. They can be defined as molecular motors because they use chemical energy to perform mechanical work – translocation and strand separation. During the NTP cycle, there may be as many as four states that contribute to forward motion and strand separation. These are (i) the apo state which has no nucleotide bound, (ii) ATP-bound state, (iii) a complex wherein the Pβ-O-Pγ bond has been broken, (iv) release of Pi (the ADP-bound state), and once the release of ADP occurs, the protein returns to the apo state [95, 96]. Each of these states is coupled to changes in the structure of the enzyme as shown for UvrD [97, 98].

The Gibbs free energy of ATP hydrolysis varies from −28 to −34 kJ/mol (i.e. ≈12 k_BT) depending on the concentration of the cation Mg²⁺ [99]. This is sufficient to separate 9–12 bp of duplex DNA or 0.2 ATP molecules hydrolyzed per bp unwound. Yet most helicases are significantly less efficient as some enzymes (RecQ) hydrolyze one ATP to translocate 2 bases on ssDNA, others a single ATP to unwind one base pair (NS3, Pif1, PcrA, UvrD, Dda), and RecBCD hydrolyzes 2–3 ATP molecules to unwind a single base pair [12, 100–105]. In contrast, RecG unwinds 4 base pairs per ATP hydrolyzed [106].

Assays to measure nucleoside triphosphate hydrolysis and product release (ADP and/or P_i) have been used extensively to determine helicase activity, ascertain energy efficiency and determine the role of NTP utilization in the biochemical mechanism of DNA helicases. This makes sense because NTP hydrolysis is coupled to movement and strand separation and in its absence helicases can melt only a few bases [107]. Assays are either radioactive and discontinuous or, light-based and continuous. Radioactive assays employ radiolabeled nucleoside triphosphates labeled at either one of the three phosphate positions or on the base, followed by thin-layer chromatography (TLC) to separate substrates (NTP) from products (NDP and/or P_i) [108]. ³H-labeled NTP substrate assays require scintillation counting of spots cut from the TLC sheets whereas ³²P-NTP TLC sheets can be exposed to phosphorimager screens and quantitated [109–111].

Continuous assays have significant advantages over discontinuous ones as they are less laborious, the data are observed directly and in addition, reaction conditions can be adjusted on the fly. One version of the assay is a coupled spectrophotometric assay [12]. Here, the hydrolysis of ATP is coupled to the oxidation of NADH, while simultaneously regenerating the ATP so that the concentration remains constant and inhibition by product is not observed (Fig. 2A). Coupling is achieved by adding phosphoenolpyruvate (PEP), NADH, and two enzymes: pyruvate kinase and lactate dehydrogenase. These components have not been demonstrated to interfere with helicase ATPase activity and provided, PEP and NADH are present above the K_m^PEP and K_m^NADH, efficiently coupling occurs. Importantly, and to permit direct visualization of hydrolysis of ATP, NADH absorbs at 340nm whereas NAD⁺ does not. Thus by measuring absorbance changes as a function of time one can “observe” ATPase activity and using the extinction coefficient of NADH, the rate of ATP hydrolysis is easily calculated. Performing this assay in a spectrophotometer, it is routine to determine kinetic parameters, substrate preferences, metal ion dependency, and relative affinity for various DNA molecules [12, 19, 76]. It is important to note that these assays do not provide the K_d or K_m for DNA. Instead, and since ATP hydrolysis is being measured, an apparent K_m for DNA is obtained, written as K_m^{DNA, apparent} [19, 76].

Figure 2. A continuous, real-time ATPase assay to monitor energy consumption by DNA helicases.

A, Schematic of the assay that couples the hydrolysis of one ATP to the utilization of one phosphoenolpyruvate (PEP) to one nicotinamide adenine dinucleotide + hydrogen (NADH) [12, 19]. As NADH absorbs at 340nm and NAD⁺ does not, the oxidation reaction can be monitored in a spectrophotometer in real-time. In the presence of DNA the helicase will translocate and/or unwind and in the process hydrolyze ATP to ADP and inorganic phosphate (P_i). Pyruvate kinase (PK) regenerates the ATP by transferring phosphate to the ADP (blue arrow). In the process, pyruvate is generated and utilized by lactate dehydrogenase (LDH) to oxidize NADH to NAD⁺ and lactate. B, raw data from an ATPase assay to monitor the effects of increasing [NaCl] on translocation by the RecG DNA helicase. Each dip in the time course corresponds to an addition of 1 μl of 1M NaCl. C, Assay to determine the DNA-dependence and substrate preference of the RecG enzyme [19]. D, Assay to determine the stoichiometry of RecG to forked DNA. Assays contained 100nM DNA molecules with activity saturating at 85–100nM RecG consistent with a 1:1 enzyme:DNA stoichiometry [76].

To demonstrate the ease with which helicase data can be obtained, three examples are presented. First, the effects of increasing [NaCl] on ssDNA translocation for RecG, that is the salt titration mid-point (STMP), can be determined in a single cuvette in a matter of minutes (Fig. 2B). Here the reaction is initiated by the addition of RecG and once a steady-state rate of ATP hydrolysis is achieved, a 1μl aliquot of NaCl is added to a final concentration of 5mM. This process is repeated 25–30 times until ATPase activity ceases. The rate after each addition is calculated as a fraction of the rate in the absence of NaCl to determine the STMP [112]. The effects of the E.coli SSB protein on the STMP can also be determined in the same way [19, 113]. Second, the DNA requirements for an enzyme can also be rapidly determined. Here, RecG was demonstrated to be a DNA-dependent ATPase, to translocate on ss-, ds- and negatively supercoiled DNA (Fig. 2C). Further, by using the same concentration of RecG in each assay it is clear (even without further calculation) that the enzyme displays the highest levels of activity on negatively supercoiled DNA [19]. Finally, this assay was also used to determine stoichiometry where the RecG helicase was titrated relative to a fork with duplex arms and a Holliday junction (all oligonucleotide-length). The reaction saturated at 1 enzyme per DNA molecule demonstrating that the enzyme preparation is 100% active (Fig. 2D) [75, 76, 114, 115].

Alternative continuous NTPase assays have been developed where the signal changes in fluorescent nucleoside triphosphates are monitored in a spectrofluorometer [116–119]. While these substrates are useful they must be used carefully as mantATP ((2′(3′)-O-(N-methylanthraniloyl)ATP) can be hydrolyzed much more slowly than unmodified ATP [106]. However, assays employing these analogues have an important benefit as they can provide insight into the pre-steady state whereas the coupled spectrophotometric assay only provides insight into steady-state kinetics. Additional fluorescence-based assays utilize reagentless biosensors that detect the formation of ADP or separately, P_i as soon as it is released from the helicase [120–124]. Here, ParM labeled with diethylaminocoumarin and separately, the phosphate binding protein of E.coli modified with N-[2-(1-maleimidyl)ethyl]-7-diethylaminocoumarin-3-carboxamide or rhodamine, were used, respectively. These sensors emit little to no fluorescence signal until they are complexed with either product of ATP hydrolysis. They have been used to provide insight into the biochemical mechanism of RecG, PcrA, RecD2, and AddAB by ascertaining the number of ATP molecules hydrolyzed per base pair unwound [106, 125–127].

3.4. Continuous, fluorescence-based, duplex DNA unwinding assays

The advantage of continuous assays over gel-based assays is that they are done in real-time and, as for the coupled spectrophotometric ATPase assay, data are observed directly, and reaction conditions can be adjusted rapidly. They require a spectrofluorometer with software that can control shutters and enables data collection, display, and analysis.

3.4.1. Using reporter proteins

In this assay single-strand DNA binding proteins (ssbs) are used as reporter molecules (Fig. 3A) [29]. These proteins (E.coli SSB or T4 gp32) have intrinsic fluorescence resulting from tyrosine or tryptophan residues [128–131]. This fluorescence is quenched upon binding to ssDNA. Therefore, in the absence of an NTP, the fluorescence signal is high (Fig. 3 A and B, phase I). As soon as an NTP is added and unwinding initiates, the ssb binds to the nascent single strands of DNA, resulting in the quenching of ssb intrinsic fluorescence (Phase II). As excess of ssb is used, the initial fluorescence signal decreases by 40–50% (Phase III). The initial rate of unwinding is calculated from the slope (Δf/Δt) in phase II. In the example shown, linear duplex DNA and the RecBCD enzyme are used but this assay can also be used for enzymes such as RecQ that can initiate unwinding at internal sites [132]. However, this requires the presence of a topoisomerase to release the accumulation of topological stress if covalently closed circular DNA is used as the substrate. In addition to providing data on the kinetic parameters for DNA unwinding, insight into the effects of mono- and divalent metal ions, and nucleoside triphosphate requirements, this assay can also be used to provide insight into enzyme processivity [34], the effects of anti-tumor drugs on helicase activity [50] and insight into the ability of helicases to unwind DNA within native and reconstituted chromatin [133]. While this assay is suitable for several helicases, enzymes that require a single-stranded tail to initiate unwinding cannot be studied as the ssb protein will bind to ssDNA and inhibit the reaction.

Figure 3. Real-time DNA helicase assays.

A and B, helicase activity is monitored using a single-strand DNA binding (ssb) protein as the reporter [29]. C and D, The displacement of fluorescent dyes is used to monitor DNA unwinding [50, 178]. A, schematic of the ssb assay. In the absence of ssDNA, the intrinsic fluorescence of E.coli SSB is high (green spheres). As unwinding proceeds, SSB binds to nascent ssDNA resulting in the quenching of the fluorescence of the reporter protein (red spheres). B, Sample data from the SSB reporter assay to monitor DNA unwinding by the RecBCD enzyme [50]. Phase I, duplex DNA present, SSB fluorescence is high. Phase II, the reaction is initiated resulting in rapid unwinding. Phase III, all available duplex DNA has been unwound resulting in a quenching of ~40% of the available SSB protein. C, Sites of residues of the E.coli SSB modified to function as a reagent less sensor of duplex unwinding. The structure of SSB bound to ssDNA (blue) is shown with relevant residues indicated [179]. In separate studies, G26 and W88 were mutated to cysteine and then modified with thiol-reactive dyes. These modified SSB proteins exhibit low fluorescence in solution, but their fluorescence is enhanced when complexed with ssDNA [135]. D, Use of the dye-modified SSB proteins to monitor DNA unwinding by RecBCD [135]. E, Use of Cy3B-modified SSB proteins to monitor unwinding of single molecules of DNA using total internal reflectance microscopy [137]. Here, either biotinylated DNA can be attached to the surface (left), or biotinylated enzyme can be attached to the surface (right). Duplex unwinding in the presence of ATP results in the accumulation of fluorescence signal resulting from modified-SSB proteins binding to the nascent ssDNA at a location corresponding to each active DNA helicase molecule. In panels D and E, black ovals are non-fluorescent reporter proteins and green ovals represent reporter proteins bound to ssDNA resulting in fluorescence emission. F, A schematic of the helicase assay using fluorescent dyes as the reporter molecule. In Phase I, dye is bound to the DNA duplex resulting in the maximal fluorescence signal (green stars). As unwinding proceeds in Phase II, displaced dye molecules exhibit a much lower fluorescence signal (black stars). Dye molecules can rebind to available sites on remaining dsDNA limiting the total amount of DNA unwinding (Phase III). G, Typical data from fluorescent dye-reporter DNA helicase assays using RecBCD as the enzyme [139]. DNA unwinding is evaluated in the absence of dye, two monomeric cyanine dyes (Cyan-2 and DM-1) and an equivalent concentration of the dimeric cyanine dye, YOYO-1.

A closely-related version of the reporter protein assay uses a class of “reagentless biosensors”, which are dye-modified ssb proteins that report the formation of ssDNA. Here either bacteriophage T4 gp32 or E.coli SSB have been used [134–136]. Gp32 was labeled with 6-iodoacetamidofluorescein at a ratio of 1:1 (dye:protein) whereas SSB, which exists as a homotetramer, is labeled at 4:1 (coumarin dye:tetramer) (Fig. 3C). Both proteins exhibit an increase in fluorescence emission upon binding to ssDNA and SSB has been used to follow DNA unwinding by RecBCD (Fig. 3D). It is important to note that the fluorescence signal in this assay increases as DNA is unwound whereas, with wild-type SSB, the signal decreases as a function of the formation of ssDNA (Fig. 3B). The coumarin-modified SSB has also been used to report DNA unwinding in single-molecule experiments using AddAB as visualized using total internal reflectance microscopy (Fig. 3E). Here the accumulation of fluorescence emission in spots corresponding to single DNA molecules is observed using wild type helicase where the DNA is attached to the surface (left image) or with biotinylated AddAB attached to the surface (the image on the right) [137]. The increase in fluorescence is due to the binding of coumarin-SSB to ssDNA resulting in the conversion of a non-fluorescent protein (black ovals) to a reporter molecule that emits fluorescence (green ovals).

3.4.2. Using dye displacement

To circumvent the inhibition of helicase activity by ssb proteins, a fluorescent dye displacement assay was developed [138]. Here, the dye molecules used must exhibit low fluorescence in the absence of DNA or the presence of ssDNA and, most importantly, demonstrate significant fluorescence enhancement in the presence of dsDNA (Fig. 3F, phases I and II). In this helicase assay, the initial fluorescence signal is high due to the binding of dyes to the DNA duplex. Once unwinding occurs, displaced dye molecules exhibit lower fluorescence, and this is observed as a decrease in the emission signal as a function of time (Phase II). 100% unwinding is not achieved in this assay as displaced dye molecules rebind to the unwound substrate and the binding of the additional dye molecules perturbs the DNA structure making it an unfavorable substrate for helicase activity (Phase III). These assays can be done in the presence and absence of an ssb protein and differences in rates are observed with the rates in the absence being more rapid. The reason for this is currently unknown.

The initial study to establish this assay used the monomeric minor groove binding dyes DAPI and Hoechst 33258 and monomeric intercalating dye thiazole orange, demonstrating comparable unwinding rates for RecBCD. Similar results were obtained using other monomeric cyanine dyes [139]. However, at the same dye concentration, the dimeric intercalating dye YOYO-1 significantly inhibited helicase activity of RecBCD, possibly due to distortion of the DNA double helix (Fig. 3G) [140]. This demonstrates a key feature of this assay and that is the concentration of dye used must provide an optical signal to noise ratio and at the same time, must not inhibit the activity of the helicase being studied. When optimized, this assay provides details on helicase kinetic parameters, enzyme processivity (provided heparin is used to trap dissociated enzymes) and can be used for helicases that initiate on tails, dsDNA ends, or at internal sites [132, 141].

3.4.3. Monitoring changes in fluorescence signal in the DNA substrate

The final set of methods to monitor duplex unwinding use changes in a fluorophore to show unwinding. The first of these used an oligonucleotide with adenine bases replaced with 2-aminopurine (2-AP) [84]. The fluorescence of 2-AP is quenched in dsDNA (i.e., when base-paired to thymine), and thus as DNA strands are separated, fluorescence increases. An alternative is to use oligonucleotides with either fluorophore pairs on the opposite strand at the same ends of a dsDNA molecule or fluorophore-quencher pairs in the same positions. In the former, fluorescence resonance energy transfer (FRET) between the pairs will occur in dsDNA and this signal will be lost on DNA unwinding, whereas in the latter case, fluorescence signal increase will be observed when strands are separated as the fluorophore and quencher have become separated [18, 80, 142].

4. Single-molecule approaches

The advantage of single-molecule studies is that individual helicases molecules can be studied to reveal the stochastic properties of single enzymes which are obscured by the averaging of ≥10⁸ molecules inherent in ensemble experiments. Not surprisingly, when the activity of a group of enzymes functioning under identical conditions is averaged, this average is identical to that observed in bulk phase. This was demonstrated using RecBCD [13]. In this section, two single-molecule approaches and their variations are presented.

4.1. Optical tweezers and fluorescence

This assay is the single-molecule version of the bulk-phase, dye displacement assay [138]. Experiments are done within the confines of multi-stream, microfluidic chambers positioned on the stage of an inverted fluorescence microscope (Fig. 4A) [13]. To isolate single molecules of DNA, optical tweezers are used to trap 1 μm streptavidin-coated beads (Fig. 4B) [143]. Single molecules of asymmetrically-biotinylated bacteriophage lambda DNA are attached to beads at a ratio of 1 DNA/bead. To permit visualization of the DNA, it is bound to the bis-intercalating, cyanine dye YOYO-1 [140]. Then, in the presence of Mg²⁺ ions, the RecBCD enzyme is bound to the free end of DNA and the complex trapped by the optical tweezers under the conditions of laminar flow. Fluid flow stretches out the DNA to its B-form length and the polynucleotide lattice appears as a white string against a black background (Fig. 4B). Next, the trapped bead-DNA-enzyme complex is moved into the adjacent fluid stream containing ATP by translating the microscope stage is perpendicular to the flow (Fig. 4A, inset). This initiates translocation and DNA unwinding by RecBCD which moves from the free end of DNA towards the bead. During strand separation, fluorescent dye molecules are displaced and washed away in the flow. Thus helicase activity is visualized directly as a shortening of the white string as a function of time.

Figure 4. Optical tweezers provide insight into translocation and DNA unwinding.

A, a 3-channel flow cell is viewed from the top. A standard flow cell is made of borosilicate glass and has 3 physical channels that combine into a single channel where fluid streams flow parallel to one another with minimal mixing due to laminar flow. Fluids are introduced by one of 3 methods indicated on the left. Typically, 1 μm beads coated with streptavidin are attached at low density to asymmetrically biotinylated lambda DNA. These complexes are stained with a bis-intercalating dye such as YOYO-1 and then introduced into one stream of the flow cell where they are optically trapped. Once they are trapped, the stage is moved perpendicular to the flow, translating the trapped molecule from one stream to the next to initiate reactions in a controlled manner (inset). B, The combination of optical tweezers, microfluidics, and the dye displacement assay that was first used to visualize unwinding by RecBCD [13]. In the schematic, the optical trap is viewed from the side with fluid flow stretching the DNA out to reveal B-form DNA coated with YOYO-1 and bound by RecBCD. Inset: a single molecule of DNA positioned in the optical trap under flow. C, Side view of the flow cell to demonstrate the parabolic flow profile. Here fluid flow near the center is rapid, resulting in DNA stretching whereas fluid flow near the surfaces is much slower and does not stretch the DNA as shown in the inset. D, Side view of a flow cell to show trap positions relative to the coverslip surface. Panels A, C, and D were adapted from [145]. E, Dual optical tweezers for force measurements on the RecB and D subunits of RecBCD simultaneously tracking on opposing strands of the duplex [151].

This assay provides insight into the ATP-dependence of helicase activity, reaction velocity of single molecules which occurred at a maximum rate of 972 ± 172 bp/sec or 300nm/sec, as well as enzyme processivity (42,300 bp unwound) by a single molecule of RecBCD [13]. At each ATP concentration, the rate of DNA unwinding varied as much as four-fold. Despite the individual variation, when the averaged data from the unwinding of 42 DNA molecules were approximated by the Michaelis-Menten equation, a V_max of 521±60 bp/sec and a K_m^ATP of 142±58 μM were obtained. These values are identical, within experimental error, to those obtained in bulk-phase assays where values of 586±45 bp/sec and 130±30μM ATP were obtained, respectively [12, 29]. Similar results were obtained when processivity (N) was examined. Here the N_max for the same 42 molecules of enzyme was calculated to be 29,670±4,256 bp with K_N^ATP of 158±78 μM whereas in bulk-phase assays processivity was calculated at 32,000 ± 1,800 bp per end, and K_N^ATP was 41±9 μM [34].

The keys to success in this assay are several-fold. First, 100% active enzyme was used [30] and this was determined in protein titrations relative to dsDNA ends in bulk-phase assays [29]. Second, knowledge of the effects of the dye on DNA structure and helicase activity was critical [138, 140, 144]. Third, a clear understanding of laminar flow within microfluidic devices is required [145]. This follows because the flow is parabolic occurring rapidly in the vertical center of the flow cell and slowest at its edges (Fig. 1C). When flow is too slow, DNA molecules are not extended making visualization of enzyme activity impossible (Fig. 1C, inset). When flow is too rapid, the DNA molecule is extended beyond B-form which is not an ideal substrate for a translocating enzyme and reactions can be completely inhibited [146]. Finally, a knowledge of optics and lasers are critical so that the focus of the optical trap(s) are 10–20 μm from the coverslip surface so that flow perturbations are eliminated, and surface effects minimized (Fig. 4D).

Further studies using this assay system revealed the recognition and response to the recombination hotspot chi by RecBCD elicited multiple changes in the enzyme: it induced a pause of up to 5 sec and elicited changes in both translocation velocity and processivity [147]. Furthermore, studies using RecBCD with a small bead attached to the RecD subunit revealed that this subunit does not dissociate from the complex at chi as previously proposed [148, 149]. The final change elicited by chi is a change in motor function within the complex. Before encountering the hotspot sequence, RecD is the lead motor whereas, after chi, it is RecB [150]. This was proposed to play important roles in DNA degradation and recombinational repair, respectively.

A non-fluorescent variation of this assay was used where optical trap position was determined using Position-Sensitive Detectors [151]. This system was developed to reveal the properties of the RecB and D subunits within the translocating holoenzyme. Here a T-shaped DNA substrate was held in opposing optical traps and RecBCD was allowed to unwind the short arm of the T, up to the junction (Fig. 1E). Because RecBCD has dual motors (RecB and RecD) that track on opposite strands of the duplex, this novel experimental format permitted the tracking of both subunits of the same enzyme complex [152, 153]. The authors found that RecBCD can generate forces up to ~40 pN, that the RecD subunit is a weak but rapidly moving DNA helicase with limited processivity and unwinds DNA as a Brownian ratchet.

4.2. Magnetic tweezers

Magnetic tweezers, developed by the Bensimon and Croquette groups are a more sophisticated version of the tethered particle motion assay [15, 154, 155]. Here ss- or dsDNA molecules, 500 – 50,000 bases in length are positioned between a superparamagnetic bead and a coverslip surface within a flow cell, commonly using avidin-biotin binding on one end and digoxygenin-antibody binding at the other (Fig. 5). A pair of magnets is placed above the sample chamber but as close as possible to apply the magnetic field gradient. As the gradient increases, the bead experiences an upward force, which holds the DNA in place. Forces ranging from femto- to nanoNewtons can be generated, well within the range required to study helicases that function in the pN range [15]. Changes in DNA length, induced by a DNA helicase are recorded as changes in the 3-dimensional bead position [156, 157]. A variety of DNA substrates, dictated by the DNA helicase being studied, have been used including nicked linear DNA (UvrD), linear duplex (AddAB), and various fork and hairpin configurations (T4 gp41, UvrD, RecG, UvsW, SMARCAL1, and RecQ) as well as ssDNA (RecQ) [44, 45, 158–161]. These studies have revealed unique aspects of helicase mechanism, ATP-dependence of translocation and DNA unwinding, step size, the effect of force on enzyme movement, and the ability of helicases to dislodge protein obstacles in their path.

Figure 5. Magnetic tweezers monitor helicase activity with single base-pair resolution.

A, Arrangement of linear duplex DNA substrates between a super-paramagnetic bead held in place by a magnetic field and a coverslip surface. Left, a nicked DNA molecule served as the substrate for UvrD [158]. Right, intact dsDNA containing the Bacillus recombination hotspot chi to study the effects of chi recognition on translocation and DNA unwinding by AddAB [159]. B, Transverse magnetic tweezers used to study translocation by UvrD [171]. C, Hairpin DNA substrates to study fork DNA processing by RecG and UvsW [44]. D, Disturbance free flow cells were developed to enable pre-steady state studies for magnetic tweezer studies [174]. Here, a PDMS layer containing microwells is sandwiched between a slide and coverslip. A reference bead and test, DNA-bead complex are introduced into each well. When fluid flow is introduced, flow occurs above each well and does not perturb the DNA substrate while simultaneously allowing the rapid exchange of reaction components. This enables pre-steady state measurements to be made.

As UvrD can initiate DNA unwinding from a nick, a nicked linear duplex DNA was positioned in the magnetic tweezers (Fig. 5A, left) [158, 162]. To minimize reannealing, the system was operated at a higher force (i.e., 35pN) so that the unwound and unattached single strand is free in solution. Under these conditions, UvrD is bound to the nick and unwound in the 3ˋ–5ˋ direction towards the bead. Surprisingly, the helicase was observed to switch strands and track in the reverse direction, i.e., towards the surface, thereby allowing DNA strands to reanneal in its wake. Strand switching has been observed for several helicases including PriA [163]. In addition, these studies also revealed that the processivity of UvrD is 255 ±20 bp, that the K_m^ATP was within experimental error, the same as that observed in bulk-phase, and that the step size was 6 ± 1.5 bp, consistent with bulk-phase measurements of 4–5 bp [158, 164]. These unwinding step sizes are larger than that observed in the UvrD crystal structure (1 bp per ATP hydrolyzed) and likely result from the inability of the bulk-phase assays to observe key intermediates and the sensitivity of the single-molecule assay which detect distinct step sizes that are coupled to inefficient ATP utilization [97, 158].

Previous work had shown that the Bacillus homolog of E.coli RecBCD recognized its cognate chi sequence and formed a stable complex requiring the latch domain [165, 166]. To understand how this might occur, linear duplex DNA with biotinylated AddAB bound to one end was held in the magnetic tweezer setup (Fig. 5A, right). Following the addition of ATP, the enzyme translocated towards the surface. Upon encountering the Bacillus chi sequence, a conformational change coupled to pausing at chi was induced in the enzyme, resulting in latch opening thereby permitting unwound DNA to exit through an alternative channel bypassing the nuclease domain similar to what was observed for RecBCD [167, 168]. This change in the enzyme converts it from a destructive helicase-nuclease to a recombinogenic enzyme [169].

The oligomeric form of an enzyme is sometimes a hotly debated issue [97, 98]. The structure of UvrD is consistent with a monomer, while several biochemical studies support a dimer [170]. The work of Dessignes using magnetic tweezers did not settle this issue [158]. To further understand the biochemical mechanism of this helicase, DNA hairpin substrates were held in place by transverse magnetic tweezers and the distance between the ends of the two tails was monitored while holding the ssDNA tail under constant force (Fig. 5B) [171]. The authors found that the rate of unwinding by UvrD decreases as the force increased, similar to other helicases, they observed unwinding and reannealing as before and obtained a comparable step size (4.2±1.7 bp). Under the conditions of the study, and as free UvrD was monomeric, they proposed that monomers bound to the DNA and met at the ss- dsDNA junction to form a dimer that more accurately approximated their data resulting in the proposal of a strained inchworm [172]. Differences between the work of Dessignes and this study were noted and were attributed to the assay setup. In the Sun study, the external tension is directly exerted on the 3′-ssDNA, whereas in the Dessignes study the displaced 3′-ssDNA tail near the nick is free of tension. This likely contributes to significantly lower unwinding rates and reduced processivity in the transverse setup.

Stalled replication fork regression requires an atypical DNA helicase that must couple DNA unwinding to duplex rewinding [57]. To understand how this might occur, a 1,200 bp hairpin was positioned in the magnetic tweezers with the arms attached to the bead and surface respectively (Fig. 5C). When force is applied the complementary arms are separated whereas when force is decreased, the arms reanneal. If the hairpin is partially unwound and an enzyme bound to the ss-dsDNA junction, that is at the fork, the ability of the helicase to regress (rewind DNA strands causing a corresponding decrease in Z-height) or reverse the fork (unwind the duplex region resulting in a corresponding increase in Z-height) can be observed. The addition of RecG and separately, UvsW resulted in ATP hydrolysis-dependent decrease in Z, corresponding to fork regression [44]. By placing short oligonucleotides on each ssDNA arm in separate experiments, the ability of each helicase to couple DNA unwind to duplex rewinding was revealed. Finally, RecG was found to be exclusively a fork regressing helicase and demonstrated the ability to work against >35pN of opposing force and to displace the SSB protein, which requires at least 10pN per tetramer [173]. In contrast, SMARCAL1 catalyzes both fork regression and fork rewinding of the same hairpin substrate and is readily inhibited by Replication Protein A (the eukaryotic equivalent of SSB) [45].

Under assay conditions in magnetic tweezer experiments, fluid flow disturbs DNA-bead complexes like seaweed in the tide, so in effect, only steady-state complexes can be studied. To circumvent this, disturbance-free flow cells were developed (Fig. 5D) [174]. In these flow cells, wells are made in PDMS which is then sandwiched between a coverslip and slide, with a spacer to enable fluid flow above the wells (Fig. 5D, blue stripe). DNA-bead complexes are positioned in the wells so that when fluid flows, components within the well can rapidly exchange (completely in 5–10 sec) without perturbing the DNA position (bottom right panel, red arrow). Thus, immediate changes in DNA structure or protein-DNA complexes, induced by buffer exchange can be observed [174]. While these flow cells have not been used to study DNA helicases to date, this represents an important technological advance adding to the repertoire of approaches to study these DNA molecular motors.

5. Concluding remarks

DNA helicases are specialized motor proteins that couple the energy present in nucleoside triphosphates to translocation, DNA unwinding, and displacement of proteins bound to the nucleic acid. The list of assays described herein is not all-encompassing and new approaches are still being developed [175–177]. Furthermore, many of the methods described can and have been applied to RNA helicases. For both DNA and RNA enzymes, the tried-and-true bulk-phase approaches continue to provide beautiful data. The newer and more technically challenging single-molecule approaches add additional, previously unattainable insight into the mechanism of individual molecules of helicases.

Highlights.

• The importance of nuclease-free and active protein

• Gel-based assays show enzyme polarity and duplex unwinding activity

• Fluorescence-based helicase assays reveal DNA unwinding in real-time

• Continuous NTPase assays provide kinetic parameters for energy coupling

• Optical tweezers and fluorescence imaging enable visual biochemistry

• Magnetic tweezers provide single base-pair resolution