Mechanisms for Defining Supercoiling Set Point of DNA Gyrase Orthologs

Background: The mechanisms controlling DNA supercoiling efficiency by gyrase are not understood.

Results: A nonconserved C-terminal tail in GyrA controls DNA binding, wrapping, and supercoiling set point.

Conclusion: The tail is a novel regulatory element that modulates gyrase activity.

Significance: Intrinsic properties of gyrase can be fine-tuned to boost enzyme output.

Abstract

DNA topoisomerases manage chromosome supercoiling and organization in all cells. Gyrase, a prokaryotic type IIA topoisomerase, consumes ATP to introduce negative supercoils through a strand passage mechanism. All type IIA topoisomerases employ a similar set of catalytic domains for function; however, the activity and specificity of gyrase are augmented by a specialized DNA binding and wrapping element, termed the C-terminal domain (CTD), which is appended to its GyrA subunit. We have discovered that a nonconserved, acidic tail at the extreme C terminus of the Escherichia coli GyrA CTD has a dramatic and unexpected impact on gyrase function. Removal of the CTD tail enables GyrA to introduce writhe into DNA in the absence of GyrB, an activity exhibited by other GyrA orthologs, but not by wild-type E. coli GyrA. Strikingly, a “tail-less” gyrase holoenzyme is markedly impaired for DNA supercoiling capacity, but displays normal ATPase function. Our findings reveal that the E. coli GyrA tail regulates DNA wrapping by the CTD to increase the coupling efficiency between ATP turnover and supercoiling, demonstrating that CTD functions can be fine-tuned to control gyrase activity in a highly sophisticated manner.

Introduction

DNA topoisomerases are essential enzymes that help counteract the topological effects of nucleic acid transactions such as transcription, replication, and repair (1). Nearly all bacteria, and some archaea, possess a unique topoisomerase, termed gyrase, which negatively supercoils DNA to maintain chromosomes in an underwound state (2, 3). Gyrase belongs to the type IIA family of topoisomerases, which are distinguished in part by their use of an ATP-dependent, duplex DNA strand-passage mechanism (1). Gyrase is an A₂B₂ heterotetramer whose supercoiling activity requires a domain (the “CTD”)² that resides at the C terminus of the GyrA subunit (see Fig. 1A) (4). The GyrA CTD forms a disc- or spiral-shaped DNA binding element that constrains a positive supercoil by wrapping a duplex around its surface (5–7). This bend, in the context of the gyrase holoenzyme, allows for the introduction of two negative supercoils upon strand passage (8, 9).

FIGURE 1.

Type II topoisomerase organization. A, primary domain structure of bacterial type IIA topoisomerases. Catalytic functions are indicated with horizontal bars. Vertical black bars mark the relative locations of the catalytic tyrosine (Cat. tyrosine) and the GyrA box. B, sequence alignment between representative CTD tails. GyrA residue numbers are as follows: E. coli GyrA 838–875; M. tuberculosis (Mtube) GyrA 817–838; Vibrio cholerae (Vchol) GyrA 862–894; Yersinia pestis (Ypest) GyrA 838–891; Staphylococcus aureus (Saure) GyrA 809–886; Salmonella typhimurium (Styph) GyrA 838–878; Caulobacter crescentus (Ccren) GyrA 899–919; Xanthomonas campestris (Xcamp) GyrA 872–899; Bacillus subtilis (Bsubt) GyrA 805–821; Mycobacterium smegmatis (Msmeg) GyrA 817–854; B. burgdorferi (Bburg) GyrA 807–810; and T. maritima (Tmari) GyrA 801–804.

The GyrA CTD constitutes a critical appendage that distinguishes gyrase from its other type IIA cousins. A paralogous bacterial enzyme, topo IV, possesses a domain that is related to the gyrase CTD (10); however, the topo IV CTD is degenerate and always lacks one or more structural features of its GyrA counterpart (11–13). Eukaryotic and phage T4 topo II appear to lack a version of the gyrase/topo IV CTD entirely. These structural differences bestow distinct activities on the different type IIA topoisomerase classes. For instance, topo IV preferentially decatenates interlocked DNA segments and removes positive DNA supercoils more efficiently than negative supercoils (14, 15). As with gyrase, this bias in activity is dependent upon its CTD (12). By contrast, topo II tends not to distinguish between the types of substrates on which it acts (16, 17), with human topo IIα being an exception (18). Thus far, it is unclear how specific evolutionary modifications to the CTD help modulate the substrate selectivity and functional output of bacterial type IIA topoisomerases. As a consequence, the extent to which the steady-state supercoiling level of the bacterial chromosome is influenced by the possession of a particular complement of type I and II topoisomerases (19, 20), versus specific augmentations to those topoisomerases, is poorly understood.

In comparing the supercoiling properties of Mycobacterium tuberculosis (Mtb) and E. coli gyrase, we discovered that the isolated GyrA proteins of the two species differ dramatically in their respective abilities to wrap DNA (see accompanying article (46)). Further analysis of this distinction led us to probe the function of the nonconserved stretch of amino acids that follows the CTD in E. coli GyrA (see Fig. 1B). We unexpectedly found that removal of either the entirety or an internal portion of the tail bestows E. coli GyrA with the capacity to wrap DNA, an activity exhibited by MtbGyrA, but not by the wild-type E. coli protein. The isolated, full-length E. coli GyrA CTD is also unable to wrap, or even bind DNA, whereas ablating or trimming the tail restores these functions. Interestingly, alterations to the CTD tail have no effect on either basal or DNA-stimulated ATPase activity, but greatly reduce both the rate of negative supercoiling and the final level of superhelical density that can be introduced by gyrase. These findings indicate that species-specific appendages to the GyrA CTD can regulate its function in the context of the gyrase holoenzyme, thereby ensuring that ATP turnover is tightly coupled to supercoiling efficiency.

EXPERIMENTAL PROCEDURES

Protein Purification

A truncated E. coli GyrA CTD construct (residues 531–853) (5, 6), as well as full-length E. coli gyrA (1–875) and gyrB (1–804) genes, were cloned into pET28b. The full-length E. coli GyrA CTD (531–875), along with “insert-less” E. coli gyrA and the GyrA CTD (both missing residues 842–856), were amplified from the gyrA pET28b vector and cloned into a derivative of pET28b using an in-house ligation-independent cloning vector system (pLIC) behind an N-terminal, tobacco etch virus protease-cleavable hexahistidine tag. Proteins were expressed in E. coli BL21-CodonPlus(DE3)-RIL cells (Stratagene) by inducing log-phase cells grown in 2× YT broth with 0.25 mm isopropyl-β-d-thiogalactopyranoside overnight at 18 °C. Cells were harvested by centrifugation, resuspended in 20 mm Tris-HCl, pH 7.9, 800 mm NaCl, 30 mm imidazole, 10% glycerol, and protease inhibitors (1 μm leupeptin, 1 μm pepstatin A, and 1 mm phenylmethylsulfonyl fluoride), and then frozen dropwise in liquid nitrogen for storage at −80 °C.

For purification, cells were sonicated and centrifuged, and the clarified lysate was passed over an Ni²⁺ affinity column (Amersham Biosciences). The His-tagged protein was eluted with 20 mm Tris-HCl, pH 7.9, 100 mm NaCl, 500 mm imidazole, 10% glycerol, and protease inhibitors (1 μm leupeptin, 1 μm pepstatin A, and 1 mm phenylmethylsulfonyl fluoride), concentrated (Millipore Amicon Ultra 10/30), and exchanged into the same buffer containing 30 mm imidazole and then incubated overnight at 4 °C with 1–1.5 mg of hexahistidine-tagged tobacco etch virus protease (21). Following tobacco etch virus cleavage, the mixture was passed over an Ni²⁺ affinity column, and the flow-through was collected, concentrated, and passed over an S-200 or S-300 gel filtration column (Amersham Biosciences) in 50 mm Tris-HCl, pH 7.9, 500 mm KCl, 10% glycerol, and 2 mm 2-mercaptoethanol. Peak fractions (determined by UV absorbance) were pooled and concentrated by centrifugal filtration (supplemental Fig. S1) (Millipore Amicon Ultra-10/30).

Topological Footprinting Assays

The introduction of DNA writhe by various gyrase domains and subunits was assessed in a buffer containing 15 mm Tris-HCl, pH 7.5, 13% glycerol, 6 mm MgCl₂, 0.1 mg/ml BSA, 70 mm KCl, and 300 ng of (6 nm) nicked pSG483 as a DNA substrate. Varying amounts of protein were added to reaction mixtures (30 μl total volume) and equilibrated at 37 °C for 20 min, after which 60 units of ligase and 1 mm ATP were added. Reactions were allowed to incubate at 37 °C for an additional 30 min and then stopped by the addition of SDS (1% final concentration), EDTA (10 mm final concentration), and proteinase K (50 μg/ml final concentration) followed by incubation at 37 °C for 30 additional minutes. Reactions were analyzed by electrophoresis through 1.0% agarose gels (Invitrogen) with 1× TAE (40 mm Tris-acetate 1 mm EDTA) running buffer. Gels were run at 1.7 V/cm for 19–21 h, stained with ethidium bromide (EtBr), and visualized by UV trans-illumination.

To capture a gyrase holoenzyme “topology footprint,” we used an ATP-independent assay technique to prevent gyrase from supercoiling the plasmid DNA substrate. In these reactions, gyrase holoenzyme was incubated with relaxed plasmid and eukaryotic topoisomerase IB, which relaxes DNA supercoils in the absence of ATP (22, 23). Holoenzyme writhe assays were performed in the presence and absence of 2 mm AMP-PNP using relaxed pSG483 plasmid and topoisomerase IB purified in-house from wheat germ (24).

DNA Binding Assays

DNA binding was determined by fluorescence anisotropy using a random, 37-bp segment with 40% GC content as a substrate. Annealed oligonucleotides 5′-TAA AGT CTA GAG ACA CGC ATA GTC AAT GAC GGA GTT A-3′ and 5′-|56-FAM|TAA CTC CGT CAT TGA CTA TGC GTG TCT CTA GAC TTT A-3′ (where 56-FAM indicates the position of a carboxyfluorescein dye for visualization) were purchased from Integrated DNA Technologies and resuspended in TE buffer (10 mm Tris, pH 8.0, 1 mm EDTA, pH 8.0). Varying amounts of E. coli GyrA full-length CTD, insert-less CTD, and tail-less CTD were incubated with 20 nm of fluorescently labeled duplex oligonucleotide at room temperature in the dark in 20 mm Tris-HCl, pH 7.5, 70 mm KCl, 10% glycerol, and 1 mm MgCl₂. Fluorescence anisotropy measurements were performed using a Victor 3V (PerkinElmer Life Sciences) multilabel plate reader. Data points represent the average of three independent measurements, where error bars represent the S.D. between measurements (see Figs. 3 and 7). Binding curves were fit to a simplified version of the single site binding equation that holds for our experimental conditions using KaleidaGraph version 4.0 (Synergy software)

where θ represents the fraction of ligand binding sites filled, K_d,app is the apparent dissociation constant, and L is the ligand concentration.

FIGURE 3.

DNA binding by CTD constructs. Binding of a fluorescein-tagged, double-stranded 37-mer oligonucleotide (20 nm), monitored as a function of CTD concentration by fluorescence anisotropy, was determined. The apparent K_d values are as follows: tail-less 90 nm ± 21 nm; insert-less, 42 nm ± 8 nm; full-length, not determined.

FIGURE 7.

ATP hydrolysis by wild-type and insert-less E. coli gyrase. A, basal ATPase activity plotted as a function of starting ATP concentration in mm (x axis) and nmol of phosphate produced per minute (y axis). Data were fit to an apparent Michaelis-Menten model (R_wt = 0.996; R_Δtail = 1.000). B, DNA stimulated ATPase activity plotted as a function of sheared salmon sperm DNA concentration in μm (x axis) and nmol of phosphate produced per minute (y axis). Data were fit to the Hill equation (R_wt = 1.000; R_Δtail = 1.000). C, ATPase activity time course. Wild-type and insert-less E. coli gyrase ATPase activities are plotted as a function of nmol phosphate produced (x axis) and time in minutes (y axis). Phosphate produced during a 50-min time course is approaching the saturation point for the assay.

DNA Relaxation and Supercoiling Assays

pSG483, a derivative of pUC19 containing a unique Nb.BbvCI nicking site, was used for supercoiled and relaxed DNA substrates. Negatively supercoiled plasmid was purified from E. coli with a maxiprep kit (Macherey-Nagel). Relaxed plasmid was made using nick ligation by first nicking maxiprepped plasmid DNA with Nb.BbvCI and then ligating with T4 DNA ligase. DNA supercoiling or relaxation assays (30 μl) were performed in a buffer containing 15 mm Tris-HCl, pH 7.5, 13% glycerol, 6 mm MgCl₂, 0.1 mg/ml BSA, 70 mm KCl, 1 mm dithioerythritol, 2 mm ATP, and 300 ng (6 nm) of DNA substrate; ATP was omitted to assess nucleotide-independent relaxation. Varying amounts of reconstituted holoenzyme were added to reaction mixtures, and reactions were allowed to proceed for 30 min at 37 °C. Time course reactions contained 4 mm ATP; for times longer than 30 min, 2 mm ATP was further added after 60 min. Reactions were stopped by the addition of SDS (1% final concentration) and EDTA (10 mm final concentration) and were analyzed by electrophoresis through 1.0% agarose gels with 1× TAE running buffer. Chloroquine gels contained an additional 3 μg/ml chloroquine (Sigma) in the gel and running buffer. Gels were run at 1.7–5.0 V/cm for 4–12 h, stained with EtBr, and visualized by UV trans-illumination.

ATPase Assays

ATPase assays were performed using a malachite green/molybdate assay for free phosphate. Reactions were run under supercoiling assay conditions and contained 50 nm gyrase. In the ATP titration assay, sheared salmon sperm DNA was held constant at 50 μm, and reactions were supplemented with a titration of ATP (0–4 mm). The DNA titration assays were started with 2 mm ATP and were supplemented with varying amounts of sheared salmon sperm DNA (0 to100 μm). ATP was added last to initiate the reactions followed by a 30-min incubation at 37 °C. 200 μl of malachite green reagent (0.034% malachite green, 10 mm ammonium molybdate, 1 n hydrochloric acid, 3.4% ethanol, and 0.01% Tween 20) was then added followed by a 5-min incubation at room temperature in the dark. Malachite green reagent was made fresh daily and sterile-filtered before use. Absorbance at 620 nm was measured using a PerkinElmer Life Sciences Victor 3V plate reader. All reactions were performed in triplicate. The ATP titration data were fit to the Michaelis-Menten model, V = (V_max [S])/(K_m+[S]), where V is the reaction rate, V_max is the maximum rate, [S] is the total starting substrate concentration, and K_m is the Michaelis-Menten constant. The DNA titration data were fit to the Hill equation, V = V_max [S]^h/(K_d^h+[S]^h) where V is the reaction rate, v_max is the maximum rate, [S] is the total substrate concentration, h is the Hill coefficient, and K_d is the apparent dissociation constant.

RESULTS

Control of DNA Wrapping by Negatively Charged E. coli GyrA Tail

The ability of gyrase to wrap large segments of DNA (∼120–140 bp) has been well established for over 30 years (9). Wrapping is known to be a property of the GyrA subunit and its C-terminal domain (5, 6); however, biochemical studies of E. coli GyrA by Reece and Maxwell (5) have shown that the full-length protein is unable to introduce writhe into DNA on its own. Although this finding at the time did not elicit additional investigation, it is surprising in retrospect as several studies have shown that isolated GyrA CTDs are readily capable of wrapping DNA, in some instances by nearly a full turn (5–7, 25).

The dichotomy between the wrapping behaviors of E. coli GyrA and its CTD came to our attention when we found that MtbGyrA imparts writhe on its own as efficiently as its isolated CTD (see accompanying article (46)). In the course of investigating this distinction, we realized that one source of variation between the CTDs of the two species, and between bacterial GyrA proteins in general, resides in the extreme C terminus of the subunit (Fig. 1B). This element, referred to hereafter as the GyrA tail, is highly acidic in both E. coli and MtbGyrA (the predicted pI of each is ∼3.0), but is significantly shorter in the latter. In other species, we found that the GyrA tail greatly varies in length, or is absent entirely (e.g. Borrelia burgdorferi and Thermotoga maritima). We might have dismissed these small differences as trivial were it not for a second source of variation that came to our attention; namely, every published study concerning the wrapping behavior of the isolated E. coli GyrA CTD has been conducted with constructs that lack a portion of the full-length domain, several of which include tail truncations (5, 6, 25).

To determine whether the tail plays a role in modulating wrapping by E. coli gyrase, we first examined its effect on the biochemical properties of the isolated CTD and of GyrA. Secondary structure predictions using PSIPRED indicate that the GyrA CTD tails are disordered (not shown), explaining why they have been either truncated or unobservable in the available crystal structures (6, 25, 26) (see also accompanying article (46)). We therefore created a series of tail deletions that lacked either an internal acidic segment or the entirety of the nonconserved region.

We initially assessed the ability of the CTD to wrap DNA using topology footprinting. In this assay, the amount of writhe imparted by a protein is assessed by incubating varying amounts of the protein with nicked plasmid DNA. The subsequent addition of ligase seals the nicked plasmid, after which treatment with SDS/proteinase K removes bound molecules to leave the DNA with an “afterimage” of the number of supercoils originally constrained upon binding. Notably, the full-length E. coli GyrA CTD was completely unable to introduce any writhe into plasmid DNA (Fig. 2A). By contrast, both the insert-less CTD and complete tail deletion exhibited the robust DNA wrapping activity seen by other groups. A similar examination of E. coli GyrA reproduced the lack of observable wrapping by the full-length protein as first noted by Reece and Maxwell (5) (Fig. 2B). As with the CTD alone, removal of the internal tail region resulted in extensive wrapping by the GyrA truncated construct (Fig. 2B), as did complete ablation of the tail (not shown).

FIGURE 2.

GyrA and CTD topology footprint assays. A, nicked pSG483 (6 nm) was incubated with varying amounts of the E. coli GyrA CTD (indicated as a molar excess of CTD over plasmid) and ligase. B, nicked pSG483 (6 nm) was incubated with varying amounts of E. coli GyrA dimer (indicated as a molar excess of GyrA dimer over plasmid) and ligase. The positions of relaxed and negatively supercoiled DNA species are labeled with graphic representations on the right.

Regulation of DNA Binding by GyrA Tail

Because removal of the GyrA tail actually led to a gain of wrapping function, we surmised that our alterations were not impairing the activity of the CTD, but rather eliminating an internal repressive element. The fact that the CTD responded to the deletions in a manner analogous to full-length GyrA in turn implied that at least part of the control was at the level of the CTD itself, and not simply due to an ability of GyrA to sequester the CTD from DNA in a tail-dependent manner.

To test this idea further, we set out to compare the abilities of the wild-type and truncated CTDs to bind DNA using fluorescence anisotropy. In this assay, short, fluorescently labeled DNA oligonucleotides were incubated with varying amounts of each purified E. coli CTD construct, and binding affinity was measured as a function of the change in the relative amount of circularly polarized light that is emitted by the sample (27). Analysis of the data from this assay showed clearly that both the tailless and the insert-less CTD constructs were competent to bind the DNA oligonucleotide with similar affinities (K_d,app = 90 ± 21 nm and 42 ± 8 nm, respectively) (Fig. 3). By contrast, no binding was detectable for the full-length CTD. These results demonstrate that the tail directly blocks the ability of the GyrA CTD to engage DNA and that its removal relieves this autoinhibitory function.

Effects of Tail on DNA Wrapping by Gyrase Holoenzyme

Because the E. coli gyrase holoenzyme readily wraps DNA (28) and because GyrA alone is unable to do so (5), we reasoned that the binding of GyrB might be capable of relieving the repressive action of the CTD tail. An additional line of evidence supporting this idea comes from a study showing that wild-type E. coli gyrase is unable to introduce writhe into DNA in the presence of AMP-PNP (29, 30). As AMP-PNP is known to act through the GyrB subunits, triggering their dimerization (31), this result suggests that a nucleotide-actuated switch in the ATPase subunits might control wrapping; however, the molecular mechanism of this switch has not been established.

To test whether the GyrA tail might be a control linkage affected by the nucleotide status of GyrB, we measured the ability of tail-truncated E. coli gyrase holoenzyme to introduce writhe in the presence and absence of AMP-PNP. Because both the insert-less and the tailless CTD constructs display similar biochemical properties, we investigated this phenomenon with the more conservative, insert-less GyrA construct. To capture a topology footprint by gyrase, an ATP-independent assay technique must be utilized to prevent the enzyme from supercoiling the plasmid DNA substrate. We therefore incubated gyrase with relaxed plasmid DNA and eukaryotic topoisomerase IB, which relaxes supercoils in the absence of ATP (22, 23). As seen previously (29), wild-type gyrase strongly wraps DNA in the absence of nucleotide and loses the ability to wrap when AMP-PNP is present (Fig. 4). By contrast, removal of a portion of the tail allowed gyrase to wrap DNA regardless of whether nucleotide was present. Interestingly, the degree of wrapping exhibited by the insert-less holoenzyme was equivalent under AMP-PNP-free and -bound conditions, but significantly less than that manifested by apo wild-type gyrase. These findings demonstrate that the acidic GyrA tail is necessary for GyrB to regulate DNA wrapping by the E. coli holoenzyme and that the tail allows for a more significant introduction of writhe when ATP is absent.

FIGURE 4.

Gyrase holoenzyme topology footprint assays. Relaxed pSG483 (6 nm) was incubated with varying amounts of reconstituted holoenzyme (Holo) and topo IB in the absence (top panel) and presence (bottom panel) of AMP-PNP (2 mm) (indicated as a molar excess of holoenzyme over plasmid). The positions of relaxed and negatively supercoiled DNA species are labeled with graphic representations on the right.

Impact of GyrA Tail on DNA Supercoiling by Gyrase

Given the marked effect on the control of DNA wrapping by the GyrA tail, we next assayed the impact of tail deletions on strand passage. To assess this function, we reconstituted tetrameric gyrase using either insert-less or wild-type GyrA, together with wild-type GyrB, and compared titrations of the mutant and wild-type enzymes in gyrase activity assays. We first looked at the ATP-dependent introduction of negative supercoils into relaxed plasmid DNA using native agarose gel electrophoresis over a range of holoenzyme concentrations. Notably, comparison of wild-type and insert-less gyrase showed that deletion of the CTD tail insert reduced the specific activity of supercoiling by more than 50-fold (Fig. 5A, upper). Deletion of the entire CTD tail produced a similar effect to that seen by the insert-less mutant (not shown). Because the insert-less GyrA protein used in this assay binds and wraps DNA extensively on its own (Fig. 2B) and because the same GyrB preparation was used for both the mutant and the wild-type supercoiling experiments, the most parsimonious explanation for this effect is that the GyrA tail is necessary for the robust activity normally displayed by the full-length E. coli holoenzyme.

FIGURE 5.

Comparison of supercoiling extent by wild-type and insert-less E. coli gyrase. A, negative supercoiling activity. Protein concentrations are listed in nm holoenzyme, and asterisks denote the minimal concentrations of enzyme needed to supercoil the substrate in 30 min. B, negative supercoiling time course assay using 20 nm holoenzyme. Time points are listed in minutes. A portion of each sample was run on a 1% agarose gel in the absence (44) and presence (bottom) of 3 μg/ml chloroquine. DNA topoisomers are depicted with graphic representations at the right side of the native gel, and the topological state indicated as relaxed (Rel.) or negatively supercoiled DNA species ((-) sc) at the bottom of the chloroquine gel. All reactions were started with 6 nm relaxed pSG483 plasmid DNA substrate.

To gain more insights into this reduction in catalytic ability, we asked whether the supercoiled DNA produced by the wild-type and insert-less enzymes was underwound to a similar extent. This property was assessed by running the same supercoiling reactions on agarose gels in the presence of 3 μg/ml chloroquine. Chloroquine is a weak intercalating agent that overwinds DNA to permit the separation of supercoiled topoisomers; at appropriate concentrations, chloroquine can help resolve negatively supercoiled species that would otherwise cluster in a single high mobility band on a native gel (32). Analysis of the topoisomer distribution by this method shows that the activity of the insert-less enzyme gave rise to DNA species that were much less negatively supercoiled than the products of wild-type gyrase (Fig. 5A, lower).

To discern whether the apparent reduction in supercoiling levels attained by the insert-less gyrase was due to a slow rate of enzyme turnover, which could prohibit the reaction from going to completion within our standard reaction time (30 min), we assayed supercoiling over a range of different periods. The reconstituted mutant E. coli gyrase was added to relaxed plasmid DNA and ATP at a concentration where complete supercoiling was seen in our enzyme titration experiments (20 nm) and incubated at 37 °C for 5, 10, 30, and 90 min; wild-type E. coli gyrase was compared as a control. Time points that exceeded 30 min were supplemented with additional ATP to prevent ATP-independent relaxation events from occurring, and the results of this assay were again visualized using agarose gel electrophoresis in the absence and presence of 3 μg/ml chloroquine. We found that even when given extremely long amounts of time, the insert-less E. coli gyrase was unable to negatively supercoil the plasmid DNA to the same extent as the wild-type enzyme (Fig. 5B). Interestingly, the degree of supercoiling manifested by the mutant approximated that seen for wild-type gyrase at the shortest time assayed (5 min), indicating that deletion of the tail did not simply slow the rate of strand passage. Rather, this element appears to have a direct and profound effect on the superhelical density “set point” of the enzyme.

In addition to negatively supercoiling DNA, gyrase can relax negatively supercoiled DNA in an ATP-independent manner (33). Although the physiological role of this reaction is unclear, it does provide another window into understanding the overall mechanism of the enzyme. Moreover, other type II topoisomerases such as topo II and topo IV fail to exhibit ATP-independent relaxation, suggesting that the unique DNA wrapping properties of gyrase may play a role in this activity. To assess the effect of the tail on the relaxation of negatively supercoiled DNA by gyrase, we titrated different amounts of either the insert-less mutant or wild-type gyrase against a fixed amount of DNA in the absence of ATP. After 30 min, the products of this reaction were visualized using native agarose gel electrophoresis. As with supercoiling, the insert-less mutant proved to be significantly less active (>10-fold) than wild-type enzyme. This finding indicates that either robust wrapping by the gyrase holoenzyme, the control of wrapping, or both is critical to its ATP-independent relaxation activity (Fig. 6).

FIGURE 6.

ATP-independent, negative supercoil relaxation by wild-type and insert-less E. coli gyrase holoenzyme. The positions of relaxed and negatively supercoiled DNA species are labeled with graphic representations on the right. All reactions were started with 6 nm pSG483 plasmid DNA substrate.

Effect of the Tail on Gyrase ATPase Activity

Because GyrB and its ATPase status were the only factors previously known to variably modulate DNA wrapping by E. coli gyrase, we next asked whether the tail was an important factor in nucleotide turnover. To address this question, the ATPase activities of wild-type and insert-less E. coli gyrase were assessed using a malachite green colorimetric assay in which ATP hydrolysis is measured by phosphate release (34). We first conducted these measurements by incubating reconstituted gyrase with increasing amounts of ATP at a fixed concentration of sheared salmon-sperm DNA; the addition of DNA is known to dramatically stimulate the ATPase activity of the enzyme (Fig. 7A) (35). We then varied the amount of DNA present in the reaction while holding the initial concentration of ATP constant (Fig. 7B). Finally, we performed the reactions for different amounts of time using a fixed starting concentration of DNA and ATP. In all instances, the insert-less gyrase hydrolyzed ATP comparatively to wild-type gyrase, demonstrating that alteration of the GyrA tail does not have an adverse effect on ATP turnover by gyrase. The findings further argue against the observed loss of supercoiling activity by the insert-less mutant as having arisen from simple inactivation of the enzyme. Instead, given our observation that gyrases lacking some or all of the tail are impaired for DNA wrapping and supercoil introduction in the context of the holoenzyme (but not with the isolated GyrA subunit or CTD), these data suggest that the tail is a linkage element that allows gyrase to efficiently couple ATP turnover to productive strand passage events.

DISCUSSION

Despite extensive study, the mechanisms used to modulate and tune topoisomerase function are still poorly understood. Here, we have uncovered a new control element that both promotes turnover efficiency and helps lock the set point for DNA supercoiling by an archetypal DNA remodeling enzyme, E. coli gyrase. Like any molecular machine, gyrase is composed of multiple moving parts and complementary activities that collaborate to support function. We have found that a nonconserved, acidic tail appended to GyrA regulates DNA binding and wrapping by an auxiliary domain at the C terminus of the subunit. The deletion of small internal segments of the tail, or its ablation, actually permits the CTD to engage and deform DNA, either on its own or in the context of GyrA, indicating that the tail is an autoinhibitory element (Figs. 2 and 3). It has been shown previously that GyrB can both restore and re-inhibit DNA wrapping by GyrA, depending on its ATPase status (29); we now find that this regulation depends on the presence of a fully intact GyrA tail (Fig. 4). Strikingly, removal of the tail decreases the rate and extent of DNA underwinding by gyrase, but does not significantly affect its ATP hydrolysis properties (Figs. 5 and 7). Hence, the tail helps to ensure that the consumption of ATP is tightly coupled to rapid and efficient strand passage.

How might the tail exert these effects physically? The finding that the negatively charged tail prevents the isolated CTD from binding DNA suggests that it acts in cis, perhaps by associating electrostatically with the rim of the domain, which is positively charged in all such domains imaged to date (6, 7, 25). This type of interaction could block DNA from accessing the CTD through steric occlusion; removal of small segments of the tail might prevent this interaction from forming stably. Alternatively, or in addition, the tail might help sequester the CTD away from DNA, perhaps by linking the domain to the rest of the GyrA subunit. The binding of GyrB could potentially relieve autoinhibition of the CTD directly, by engaging the tail to allow DNA wrapping, or allosterically, by inducing a conformational change in GyrA that promotes both CTD release and undocking of the tail to permit DNA engagement.

A scheme outlining these concepts is shown in Fig. 8. Importantly, it makes several predictions that are supported by various lines of experimental evidence. For instance, small x-ray scattering studies have suggested previously that the E. coli CTD may associate with the rest of GyrA in a manner that would prevent it from readily engaging a DNA bound to the active site region of the protein (36, 37). Our deletion studies demonstrate that the tail is not only critical for the GyrB-dependent modulation of DNA wrapping by gyrase, but that the extent of holoenzyme-induced wrapping is compromised when the tail is altered (Fig. 4). Given that partial or complete truncation of the tail enables robust wrapping by both the CTD and the GyrA on their own (Fig. 2), these data suggest that the presence of GyrB helps place the CTD in a productive conformation to bind DNA, introduce writhe, and correctly position a DNA segment in cis for strand passage. Direct tethering of the tail to GyrB would link wrapping to the ATPase cycle and help ensure that turnover and strand passage are efficiently interconnected. Both coupling phenomena have been seen biochemically (30, 38–40), although the CTD tail has not been implicated in the process until now.

FIGURE 8.

Model accounting for the action of CTD tail on E. coli gyrase. Arrow thickness indicates the qualitative probability of transitioning between states. Panel i, full-length GyrA cannot wrap DNA on account of the repressive action of the tail on CTD activity and/or localization. Panels ii and iii, binding of GyrB releases the repressive activity, either allosterically or by binding the tail directly, allowing wrapping. Panel iv, ATP binding induces strand passage, freeing the tail from the action of GyrB; hydrolysis allows reset back to state (panel iii). Successive rounds of turnover allow for DNA supercoiling, which is extensive due to the tight coupling of CTD wrapping with ATPase status. Panel v, futile cycle in which ATP binding is accidentally not converted to a strand passage event, which is possibly due to incomplete wrapping. In wild-type gyrase, this event is rare (38, 45), but it may become more prevalent as DNA becomes more negatively supercoiled. Panel vi, AMP-PNP permanently relieves the effect of GyrB on the tail, allowing re-engagement of the CTD to prohibit wrapping. Note that our data neither support nor exclude the possibility that a similar, tail-repressed CTD state might occur during normal enzyme cycling. Panel vii, removal of the tail frees the CTD to wrap DNA in the context of GyrA alone. Panel viii, GyrB does not potentiate, and may even hinder, wrapping in the context of the holoenzyme. Panel ix, futile cycles of ATPase activity occur with increased frequency on account of the loss of coupling between CTD activity and GyrB status. Panel x, occasional ATPase cycles result in strand passage. The degree of supercoiling obtained without the tail is not as extensive as in wild-type gyrase, either because the inefficiency of the enzyme permits a relaxation “back-reaction” to occur (i.e. tail-less gyrase acts like a nonsupercoiling type II topoisomerase) or because the ability of the enzyme to maintain a wrapped DNA state cannot compensate as its “load” is increased (i.e. the wrap on the CTD slips as DNA becomes more supercoiled). At present, we cannot distinguish between these two possibilities. Panel xi, AMP-PNP has no effect on wrapping by a tail-less enzyme because the CTD is no longer under the control of GyrB. NTD, N-terminal domain.

Because the GyrA tail is not conserved (Fig. 1), it is unlikely that the mechanism by which it controls E. coli gyrase will be universal. Indeed, differences in supercoiling between gyrase orthologs have been observed previously (41–43) (see also accompanying article (46)) and have been attributed to disparate factors such as alterations in GyrB sequence or CTD structure. These studies indicate that gyrase is not a monolithic entity and that there instead exists a range of gyrase activities that vary in their relative supercoiling efficiencies. They also suggest that rather than being the rule in terms of supercoiling efficiency, E. coli gyrase is a supercharged exception. It is presently unclear how the E. coli GyrA tail might selectively bind to different regions of gyrase to control holoenzyme conformation and function. It is also not understood how or why particular bacterial species adjust and select their respective topoisomerase retinue to achieve a particular level of steady-state supercoiling. Issues such as these are subjects for future investigations.