Kinetic Study of DNA Topoisomerases by Supercoiling-Dependent Fluorescence Quenching.

DNA topoisomerases are essential enzymes for all living organisms and important targets for anticancer drugs and antibiotics. Although DNA topoisomerases have been studied extensively, steady-state kinetics has not been systematically investigated because of the lack of an appropriate assay. Previously, we demonstrated that newly synthesized, fluorescently labeled plasmids pAB1_FL905 and pAB1_FL924 can be used to study DNA topoisomerase-catalyzed reactions by fluorescence resonance energy transfer (FRET) or supercoiling-dependent fluorescence quenching (SDFQ). With the FRET or SDFQ method, we performed steady-state kinetic studies for six different DNA topoisomerases including two type IA enzymes (Escherichia coli and Mycobacterium smegmatis DNA topoisomerase I), two type IB enzymes (human and variola DNA topoisomerase I), and two type IIA enzymes (E. coli DNA gyrase and human DNA topoisomerase IIα). Our results show that all DNA topoisomerases follow the classical Michaelis-Menten kinetics and have unique steady-state kinetic parameters, K M, V max, and k cat. We found that k cat for all topoisomerases are rather low and that such low values may stem from the tight binding of topoisomerases to DNA. Additionally, we confirmed that novobiocin is a competitive inhibitor for adenosine 5'-triphosphate binding to E. coli DNA gyrase, demonstrating the utility of our assay for studying topoisomerase inhibitors.

Introduction

DNA topoisomerases are enzymes that catalyze DNA structural alterations that include relaxation of positively and negatively supercoiled DNA and resolution of DNA knots and catenanes.[1−4] These enzymes create transient DNA breaks to catalyze these changes in DNA topology.[1−4] Because topological issues are intrinsic to double-stranded DNA,[5,6] DNA topoisomerases are critical for DNA replication, transcription, recombination, and maintenance of the chromosome structure.[7−9] DNA topoisomerases are important targets for anti-bacterial agents as well as anti-cancer drugs.[10,11] Human topoisomerases I and II are targets of clinically important anticancer drugs, such as topotecan[10,12] and doxorubicin.[10] Bacterial DNA gyrase and Topo IV (another type II enzyme) are the targets of fluoroquinolones, such as ciprofloxacin, which are critically important antibiotics.[13−15] Bacterial topoisomerases continue to be useful targets for the discovery of novel antibacterial drugs that could avoid cross resistance with current antibiotics to counter the serious global health problem of multidrug-resistant bacterial pathogens.

DNA topoisomerases are classified into type I and II families according to whether they make a transient single-stranded nick or double-stranded break during catalysis.[1] Each family is further divided into different subfamilies depending on catalytic mechanisms.[1,4] Type I enzymes cleave only one strand of the DNA templates, while type II enzymes transiently cleave both strands. During catalysis, type IA DNA topoisomerases (bacterial DNA topoisomerases I and III) link to the 5′-phosphate. In contrast, type IB topoisomerases (human topoisomerase I and poxvirus DNA topoisomerase I) link to the 3′-phosphate of DNA.[1,4] Although most DNA topoisomerases can relax (−) and/or (+) DNA supercoiling,[1,4] bacterial DNA gyrase is the only enzyme that can actively introduce (−) DNA supercoiling to its DNA substrates.[1,4]

One major assay for DNA topoisomerase activity is gel electrophoresis.[16,17] Although gel electrophoresis is quick and convenient for resolving topoisomers of closed circular plasmid DNA molecules,[16,17] it is less suitable for probing kinetics because gel electrophoresis cannot readily provide quantitative analysis of DNA supercoiling in real time. More recently, single-molecule techniques, such as magnetic tweezers, have been used to analyze mechanisms of DNA topoisomerases-catalyzed supercoiling/relaxation reactions.[18−21] Indeed, these elegant techniques provided mechanistic insights into various DNA topoisomerase-catalyzed supercoiling/relaxation reactions;[22−29] however, DNA molecules must be physically linked onto solid surfaces.[18,30] It is also not straightforward to derive steady-state kinetic parameters, such as KM and kcat, from these studies. Additionally, constraint onto magnetic beads could influence diffusion or other aspects of radial DNA access of a topoisomerase before, during, or after the breaking/resealing reaction.[18,30]

Recently, we synthesized a type of unique fluorescently labeled DNA molecules that can be used to study DNA topoisomerases by fluorescence resonance energy transfer (FRET) or supercoiling-dependent fluorescence quenching (SDFQ).[31] This SDFQ method stems from the fact that alternating adenine–thymine sequences (AT) in the closed circular plasmids undergo very rapid cruciform formation/deformation depending on the supercoiling density of the plasmids.[32,33] A pair of fluorophore–quencher were inserted in the (AT) sequence, so that the distance between the fluorophore and quencher is dramatically changed when the plasmids adopt supercoiled (sc) or relaxed (rx) status.[31] As a result, the fluorescence intensity of the plasmids is also greatly changed upon supercoiling transition.[31] We demonstrated that these DNA molecules are excellent tools to examine relaxation/supercoiling kinetics of DNA topoisomerases and can be configured into rapid and efficient high-throughput screening assays to identify topoisomerase inhibitors.[31] In this work, we report an improved procedure to rapidly and efficiently synthesize and purify fluorescently labeled plasmid DNA molecules. Using the SDFQ assays, we find that all 6 DNA topoisomerases examined show characteristics of classical Michaelis–Menten kinetics. We also demonstrate that novobiocin is a competitive inhibitor for adenosine 5′-triphosphate (ATP) binding to Escherichia coli DNA gyrase.

Materials and Methods

Materials

Restriction enzyme Nt.BbvCI, T4 DNA polymerase, T5 exonuclease, and T4 DNA ligase were purchased from New England Biolabs (Beverly, MA, USA). E. coli DNA topoisomerase I, E. coli DNA gyrase, Mycobacterium smegmatis DNA topoisomerase I, variola DNA topoisomerase I, human topoisomerase I, and human topoisomerase IIα were purified as described previously.[34−37] Ethidium bromide, buffer-saturated phenol, and isopropanol were purchased from Thermo Fisher Scientific, Inc. The synthetic oligonucleotides FL905 and FL924 were purchased from MWG-Biotech, Inc. (Huntsville, AL) and were described previously.[31] Plasmid pAB1 was described previously[31] and purified using commercial plasmid purification kits obtained from QIAGEN, Inc. (Valencia, CA).

Synthesis of rx and sc pAB1_FL905 and pAB1_FL924

The preparation of rx and sc pAB1_FL905 and pAB1_FL924 was described previously[31] with some modifications. Briefly, 1 mg of pAB1 (∼570 pmol) was digested by 2500 units of Nt.BbvCI in 20 mL of 1× CutSmart buffer for 1 h at 37 °C. After the digestion, 8000 pmol of phosphorylated FL905 or FL924 was added into the reaction mixture. The reaction mixture was incubated at 90 °C in a 4 L water bath for 2 min and then cooled down to room temperature in the water bath (∼4–5 h; usually this step was carried out overnight). To generate rx pAB1_FL905 or pAB1_FL924, 25 000 units of T4 DNA ligase were added in the presence of 10 mM of dithiothreitol (DTT) and 2 mM of ATP (final concentrations). The reaction mixtures were incubated at 37 °C for 60 min to seal the nicks, thereby yielding the rx pAB1_FL905 or pAB1_FL924. T4 DNA polymerase (750 units) and dNTPs (100 μM) were added to the ligation mixture to increase the yield of the rx pAB1_FL905 or pAB1_FL924. After the ligation step, 5000 units of T5 exonuclease were added into the reaction mixture at 37 °C for 60 min to digest the nicked pAB1 and oligomer FL905 or FL924. The rx pAB1_FL905 or pAB1_FL924 sample was extracted with 20 mL of phenol, precipitated with isopropanol, washed once with 70% ethanol, and dialyzed against a large excess of 1× TE buffer solution (10 mM Tris-HCl, 1 mM ethylenediaminetetraacetic acid (EDTA), pH 8.0). To generate sc pAB1_FL905 or pAB1_FL924, the ligation reaction was carried out in the presence of 25 μM of ethidium bromide (EB). The other steps are the same as the synthesis of rx pAB1_FL905 or pAB1_FL924. The sc pAB1_FL905 or pAB1_FL924 sample was extracted with 20 mL of phenol, precipitated with isopropanol, washed once with 70% ethanol, and dialyzed against a large excess of 1× TE buffer solution.

Fluorescence Spectroscopy

Fluorescence measurements were performed using a Horiba FluoroMax-3 spectrofluorimeter with an excitation wavelength of 494 nm or a Biotek Synergy H1 Hybrid Plate Reader with an excitation wavelength of 482 nm where plasmid pAB1_FL905 was used. Alternatively, fluorescence measurements were performed using a homemade spectrofluorimeter with a 532 nm laser for the excitation light source where pAB1_FL924 was used due to the fact that the fluorophore TAMRA of pAB1_FL924 was stable under this condition.

DNA Supercoiling Density Determination

sc pAB1_FL924 (4.6 μg) was relaxed by variola DNA topoisomerase I in the presence of various concentrations of EB at 37 °C in 1× CutSmart buffer for 1 h. Subsequently, the relaxation reaction was stopped by addition of an equal volume of phenol. The reaction mixtures were extracted one more time with phenol (a total of 2 times of phenol extraction) and dialyzed against 1 L of 10 mM Tris-HCl, pH 8.0 twice overnight. The topological status of each DNA sample was analyzed by electrophoresis in a 1% agarose gel in 1× TAE buffer [40 mM Tris-acetate (pH 7.8) and 1 mM EDTA] containing different concentrations of chloroquine. After electrophoresis, agarose gels were stained with EB, destained, and photographed under UV light. The DNA linking number change (ΔLk) was determined by analyzing the distributions of the topoisomers, and the supercoiling density (σ) was calculated as followswhere Lk0 and Lk represent the DNA linking number for the relaxed and the supercoiled DNA, respectively.

The fluorescence intensity of these DNA samples was measured by using a microplate reader with λex = 550 nm and λem = 580 nm. DNA concentration was determined by UV absorbance at 260 nm using a Cary 50 spectrophotometer.

Steady-State Kinetic Measurements

All steady-state kinetic measurements were performed in 60 μL of a buffer solution containing either sc pAB1_FL905 (pAB1_FL924) or rx pAB1_FL905 (pAB1_FL924). CutSmart buffer [1× is 50 mM KAc, 20 mM Tris-Ac, 10 mM Mg(AC)2, 100 μg/mL bovine serum albumin (BSA), pH 7.9] was used for E. coli DNA topoisomerase I, M. smegmatis DNA topoisomerase I, and variola DNA topoisomerase I. Human Top1 buffer (1× is 10 mM Tris-Cl, pH 7.9, 150 mM NaCl, 0.1% BSA, 0.1 mM spermidine, 5% glycerol) was used for human DNA topoisomerase I. Human Top2 buffer (1× is 50 mM Tri-HCl, pH 8, 100 mM KCl, 1 mM EDTA, 8 mM MgCl2, 35 mM β-mercaptoethanol, 0.5 mg/mL BSA, and 2 mM ATP or indicated in the assay) was used for human DNA topoisomerase II. DNA gyrase buffer (1× is 35 mM Tris-HCl, pH 7.5, 24 mM KCl, 4 mM MgCl2, 2 mM DTT, 5 mM spermidine, 0.1 mg/mL BSA, 6.5% glycerol, and 1.75 mM ATP or indicated in the assay) was used for E. coli DNA gyrase. Kinetic reaction mixtures were assembled on ice (without enzyme) and equilibrated to 37 °C usually for 5 min (in a cuvette inside the spectrofluorimeter). Then, one of DNA topoisomerases was added directly to the cuvette and mixed with other components of reaction mixtures by pipetting three times. The fluorescence intensities of the reaction mixture at 521 nm (for pAB1_FL905) or at 582 nm (for pAB1_FL924) were recorded every 5 s. The initial velocity of the reactions was calculated from linear-fitting of the first 5–10 data points. The steady-state kinetic parameters KM, Vmax, and kcat were obtained by fitting the Michaelis–Menten equationwhere V0, [S], KM, Vmax, [E], and kcat represent the initial velocity, substrate concentration, Michaelis constant, maximum velocity, enzyme concentration, and turn-over number, respectively.

Results and Discussion

A rapid and efficient procedure to synthesize and purify fluorescently labeled plasmid DNA molecules.

Recently, we reported the synthesis of a type of unique fluorescently labeled plasmid DNA molecules that can be used to study DNA topology and topoisomerases by FRET or SDFQ.[31] Although the synthesis yield of the fluorescently labeled plasmid DNA molecules was high, agarose gel electrophoresis or CsCl–EB ultracentrifugation banding had to be used to purify the labeled products from unlabeled nicked DNA, the oligomer, and ATP.[31] Both purification methods have unavoidable disadvantages. Specifically, gels are not suited for purifying mg quantities of plasmid DNA, and although CsCl–EB ultracentrifugation banding is capable of purifying the labeled products in the milligram range, it is a lengthy, expensive, and labor-intensive procedure. Additionally, ∼20% products may be lost during this purification procedure and handling circular DNA in the presence of EB causes significant nicking of the substrate DNA. Therefore, it was desirable to develop a method to avoid these two labor-intensive purification methods. We developed a simple and efficient procedure to synthesize and purify fluorescently labeled rx and sc plasmid DNA molecules (Figure ). We used T5 exonuclease to digest nicked and gapped plasmids, and also oligomers FL905 and FL924 because T5 exonuclease cannot digest either rx and sc plasmid DNA molecules[38] (compare lanes 1 and 2 with lane 4, and also lanes 5 and 6 with lane 8 of Figure A). We also found that T4 DNA polymerase could increase the yield of the labeled products (compare lane 1 with lane 2 and lane 5 with lane 6 of Figure A). A possible reason is that T4 DNA polymerase repaired some damaged DNA templates during the nicking reaction. Additionally, 2 times of phenol extraction removed most of the digested labeled oligomers FL905 and FL924 (Figure A: compare lanes 1 and 2 with lanes 3 and 4; also compare lanes 5 and 6 with lanes 7 and 8). After isopropanol precipitation and dialysis steps, the fluorescently labeled plasmids pAB1_FL905 and pAB1_FL924 are free of the labeled oligomers and nicked or gapped pAB1. To generate sc pAB1_FL905 or pAB1_FL924, 25 μM of EB was added directly into the ligation reactions. We found that EB did not significantly affect ligation reactions (Figure S1). T5 exonuclease was able to remove nicked and gapped pAB1, and also oligomers FL905 and FL924 (Figure A). With this new method, we were capable of synthesizing milligrams of rx or sc pAB1_FL905 and pAB1_FL924 in 3–4 days with high yields.

Figure 1

Rapid and efficient procedure to produce rx or sc fluorescently labeled plasmid pAB1_FL905. The experimental procedure was described under the Materials and Methods. Abbreviations: T4 DNAP, T4 DNA polymerase; DNA ligase, T4 DNA ligase; sc, supercoiled; nk, nicked, rx, relaxed.

Figure 2

(A) Effects of different enzymes on the production of rx and sc pAB1_FL905. rx (lanes 1–4) and sc (lanes 5–8) pAB1_FL905 were generated according to Figure and also as described in “Materials and Methods.” The DNA molecules were isolated and subjected to 1% agarose gel electrophoresis in 1× TAE buffer. Symbols and abbreviations: T4 DNAL, T4 DNA ligase; T4 DNAP, T4 DNA polymerase; T5 exo, T5 exonuclease; nk, nicked plasmid; sc, supercoiled plasmid; rx, relaxed plasmids; FL905, oligomer FL905. Lane 9 is the NEB 1 kb DNA ladder. (B) Fluorescence spectra of sc (σ = −0.06; dotted line) and rx (σ = 0; solid line) of pAB1_FL924 with λex = 532 nm. (C) Fluorescence intensity of pAB1_FL924 is dependent of supercoiling density (σ). DNA topoisomeras with a mean ΔLk were generated as described under Materials and Methods and used here. The fluorescence intensity of the same concentration of pAB1_FL924 samples was measured using a microplate reader with λex = 550 nm and λem = 580 nm. (D,E) Analysis of DNA topoisomers using 1% agarose gel electrophoreses in the absence (D) and presence of 2.5 μg/mL chloroquine (E) to determine the supercoiling density of different pAB1_FL924 samples. Lanes 1 to 9 are DNA samples relaxed by variola DNA topoisomerase I in the presence of 0, 0.5, 1, 1.5, 2, 2.5, 3.75, 5, and 7.5 μM of EB, respectively. Lanes 10 and 11 are supercoiled and relaxed pAB1, respectively.

In order to establish the relationship between the fluorescence intensity and the supercoiling density (σ) of these fluorescently labeled plasmids, we relaxed sc pAB1_FL924 in the presence of increasing concentrations of EB using variola DNA topoisomerase I. After removing EB by phenol extraction and extensive dialysis, the fluorescence intensity of these DNA samples was measured and σ was determined.[39]Figure B–E shows our results. As expected, the fluorescence intensity was dependent on σ. When σ of pAB1_FL924 ≤ −0.04 with a ΔLk ≤ −11, the fluorescence intensity reached the minimum. This is consistent with previous results showing that the AT cruciform/hairpin was fully formed at σ ≤ −0.04.[40] As a result, the fluorescence was significantly quenched (Figure B). At σ ≥ −0.02 with ΔLk ≥ −5, the hairpin structure disappeared and the fluorescence intensity increased sharply (Figure C). The supercoiling-dependent fluorescence transition supports that these fluorescently labeled plasmids can be used to study kinetics of different DNA topoisomerases.

Steady-State Kinetics of Type I DNA Topoisomerases

With rx or sc pAB1_FL905 and pAB1_FL924, we carried out steady-state kinetic studies for 6 different DNA topoisomerases. The first DNA topoisomerase we examined is E. coli DNA topoisomerase I, a 96 kDa monomeric type IA topoisomerase.[1] We kept E. coli DNA topoisomerase I concentration at 17 nM and varied pAB1_FL905’s concentration from 0.36 to 6.39 nM (Figure ). This condition is consistent with previously published conditions for testing E. coli topoisomerase I’s activity using agarose gel electrophoresis.[34,41] Note that plasmid pAB1 is much larger than E. coli DNA topoisomerase I (1800 kDa vs 96 kDa) and sc pAB1_FL905 has ∼20 (−) supercoils. Considering the fact that multiple E. coli DNA topoisomerase I can bind and remove (−) supercoils from each pAB1_FL905, our testing conditions should meet the requirements for steady-state kinetics. Nevertheless, addition of E. coli DNA topoisomerase I into a solution containing sc pAB1_FL905 caused a rapid increase of fluorescence intensity of the solution at 521 nm that reached plateau within 200–300 s (Figure B). Initial velocities were calculated from these time courses and successfully fitted to the Michaelis–Menten equation to yield a KM of 1.2 ± 0.3 nM, Vmax of 50 ± 4 pM/s, and kcat of 3.0 × 10–3 s–1. These values are lower than those that we obtained from agarose gel electrophoresis published previously.[34] Possible reasons for this disparity include the following: (1) gel electrophoresis is not as quantitative for analyzing the kinetics of DNA topoisomerases. For example, the disappearance of the “most sc species” used in the agarose-gel based assays may not represent the real initial velocity of E. coli DNA topoisomerase I because the 1% agarose gels in the absence of chloroquine cannot differentiate the initial, less sc DNA products from the “most sc species”; (2) a different plasmid DNA template, that is, pXX6 (4483 bp) was used in our previous studies and is significantly larger than pAB1_FL905 used here; and (3) we used more E. coli DNA topoisomerase I in the current studies than in the previous studies[34] (17 vs 8 nM). The low turnover number (kcat, 3.0 × 10–3 s–1) suggests that E. coli DNA topoisomerase I is a “slow” enzyme. It should be noted however that the kcat values determined here represent the whole plasmid molecule involving multiple catalytic cycles along with the dissociation rate constant as the enzyme is released from relaxed DNA products. We also carried out a time course of the relaxation reaction of E. coli DNA topoisomerase I using agarose gel electrophoresis (Figure D). At 300 s, all pAB1_FL905 is fully relaxed, which is consistent with our fluorescence results.

Figure 3

Steady-state kinetics of E. coli DNA topoisomerase I. (A) Proposed reaction scheme for the relaxation reaction catalyzed by E. coli DNA topoisomerase I. (B) Time courses of E. coli DNA topoisomerase I-catalyzed DNA relaxation reactions monitored by fluorescence intensity change. For the relaxation reaction, 60 μL of 1× CutSmart buffer (50 mM KAc, 20 mM Tris-Ac, 10 mM Mg(AC)2, 100 μg/mL BSA, pH 7.9) containing different concentrations of sc pAB1_FL905 was prepared and equilibrated to 37 °C, and 17 nM of E. coli DNA topoisomerase I was used to relax the sc pAB1_FL905. The fluorescence intensity at λem = 521 nm was monitored with λex = 494 nm using a Horiba FluoroMax-3 spectrofluorimeter. (C) Initial velocities of relaxation reaction were calculated from (B), plotted against the substrate (pAB1_FL905) concentration, and fitted into the classical Michaelis–Menten equation to determine KM, Vmax, and kcat. (D) Time courses of E. coli DNA topoisomerase I-catalyzed DNA relaxation reactions monitored by 1% agarose gel electrophoresis; 465 μL of 1× CutSmart buffer (50 mM KAc, 20 mM Tris-Ac, 10 mM Mg(AC)2, 100 μg/mL BSA, pH 7.9) containing 2.13 nM of sc pAB1_FL905 was prepared and equilibrated to 37 °C, and 17 nM of E. coli DNA topoisomerase I was used to relax the sc pAB1_FL905. The reactions were stopped by adding 20 mM EDTA and 0.1% SDS into the reaction mixtures. Lanes 1–7 represent DNA samples from 0, 30, 60, 120, 300, 450, and 600 s of the relaxation assay, respectively. rx and sc represent relaxed and supercoiled DNA molecules, respectively.

Previous studies showed that high salt concentrations inhibited the relaxation activities of E. coli DNA topoisomerase I and also changed the final topological status of the relaxed topoisomers.[42] Here, we performed kinetic studies of E. coli topoisomerase I using plasmid pAB1_FL924 in two different solution conditions, 1× CutSmart buffer and 1× CutSmart buffer plus 100 mM NaCl. Figure S2 shows the results. Indeed, high salt concentration decreased the fluorescence intensity of the final DNA topoisomers, presumably because of the topological status difference of the final relaxed topoisomers (Figure S2A), and also inhibited the relaxation activities of E. coli DNA topoisomerase I (Figure S2B). After fitting our kinetic data to the Michaelis–Menten equation, we obtained the following kinetic parameters: KM of 1.2 ± 0.2 nM, Vmax of 51 ± 4 pM/s, and kcat of 3.0 × 10–3 s–1 for 1× CutSmart buffer and KM of 1.5 ± 0.2 nM, Vmax of 40 ± 3 pM/s, and kcat of 2.3 × 10–3 s–1 for 1× CutSmart buffer plus 100 mM NaCl. The KM value does not change significantly, while Vmax decreases in the high salt buffer condition, which is similar to the effect of a noncompetitive inhibitor of the enzyme.

M. smegmatis DNA topoisomerase I, a 110 kDa monomeric type IA topoisomerase,[43] was also examined. It is the only type IA DNA topoisomerase of M. smegmatis, essential for this bacterium,[44,45] and a validated target for antibiotic discovery.[13,44,46] In DNA relaxation time courses, we kept M. smegmatis DNA enzyme fixed at 13.3 nM and varied DNA concentrations from 0.37 to 4.41 nM (Figure S3). Similar to E. coli DNA topoisomerase I, M. smegmatis DNA topoisomerase I was able to fully relax sc pAB1_FL905 (Figure S3A,D). Initial velocities were calculated from these time courses and successfully fitted to the Michaelis–Menten equation to yield a KM of 4.3 ± 3.0 nM, Vmax of 130 ± 5 pM/s, and kcat of 9.8 × 10–3 s–1. The KM value is slighter higher than that of E. coli DNA topoisomerase I and may be due to the divergent C-terminal domains.[47] Again, the turnover number (kcat, 9.8 × 10–3 s–1) is rather low indicating that M. smegmatis DNA topoisomerase I is also a comparatively “slow” enzyme.

Additionally, two type IB DNA topoisomerases, variola and human DNA topoisomerase I, were examined. Both enzymes can relax (+) and (−) DNA supercoiling.[1,48] Variola DNA topoisomerase I is the smallest topoisomerase (314 aa residues, ∼36.6 kDa; variola DNA topoisomerase I is similar to another poxvirus vaccinia topoisomerase I except for 3 amino acids[48,49]). Figure shows the steady-state kinetics of variola DNA topoisomerase I. The KM, Vmax, and kcat were determined to be 2.4 ± 1.0 nM, 120 ± 23 pM/s, and 6.0 × 10–3 s–1, respectively. Human topoisomerase I is a 90.7 kDa monomer (765 aa residues)[1] and an important anticancer drug target.[50] Notably, topoisomerase I inhibitors irinotecan and topotecan, derivatives of camptothecin, are highly effective FDA approved anticancer drugs.[51] Interestingly, more recent studies showed that human topoisomerase I relaxation activity is related to autism spectrum disorder[52] and inflammation.[53]Figure shows the results of steady-state kinetic studies of human DNA topoisomerase I. Apparently, it follows the classic Michaelis–Menten kinetics. The KM, Vmax, and kcat were determined to be 2.7 ± 1.00 nM, 31 ± 5 pM/s, and 1.2 × 10–3 s–1, respectively.

Figure 4

Steady-state kinetics of variola DNA topoisomerase I. (A) Proposed reaction scheme for the relaxation reaction catalyzed by variola DNA topoisomerase I. (B) Time courses of variola DNA topoisomerase I-catalyzed DNA relaxation reactions monitored by fluorescence intensity change. For the relaxation reaction, 60 μL of 1× CutSmart buffer (50 mM KAc, 20 mM Tris-Ac, 10 mM Mg(AC)2, 100 μg/mL BSA, pH 7.9) containing different concentrations of sc pAB1_FL905 was prepared and equilibrated to 37 °C, and 20 nM of variola DNA topoisomerase I was used to relax the sc pAB1_FL905. The fluorescence intensity at λem = 521 nm was monitored with λex = 494 nm using a Horiba FluoroMax-3 spectrofluorimeter. (C) Initial velocities of relaxation reaction were calculated from (B), plotted against the substrate (pAB1_FL905) concentration, and fitted into the classical Michaelis–Menten equation to determine KM, Vmax, and kcat. (D) Time courses of variola DNA topoisomerase I-catalyzed DNA relaxation reactions monitored by 1% agarose gel electrophoresis; 450 μL of 1× CutSmart buffer (50 mM KAc, 20 mM Tris-Ac, 10 mM Mg(AC)2, 100 μg/mL BSA, pH 7.9) containing 2.2 nM of sc pAB1_FL905 was prepared and equilibrated to 37 °C, and 20 nM of variola DNA topoisomerase I was used to relax the sc pAB1_FL905. The reactions were stopped by adding 20 mM EDTA and 0.1% SDS into the reaction mixtures. Lanes 1–7 represent DNA samples from 0, 30, 60, 120, 300, 450, and 600 s of the relaxation assay, respectively. rx and sc represent relaxed and supercoiled DNA molecules, respectively.

Figure 5

Steady-state kinetics of human DNA topoisomerase I. (A) Proposed reaction scheme for the relaxation reaction catalyzed by human DNA topoisomerase I. (B) Time courses of human DNA topoisomerase I-catalyzed DNA relaxation reactions monitored by fluorescence intensity change. For the relaxation reaction, 60 μL of 1× human Top1 buffer (10 mM Tris-Cl, pH 7.9, 150 mM NaCl, 0.1% BSA, 0.1 mM spermidine, 5% glycerol) containing different concentrations of sc pAB1_FL905 was prepared and equilibrated to 37 °C, and 25 nM of human DNA topoisomerase I was used to relax the sc pAB1_FL905. The fluorescence intensity at λem = 521 nm was monitored using a Biotek Synergy H1 Hybrid Plate Reader with an excitation wavelength of 482 nm. (C) Initial velocities of relaxation reaction were calculated from (B), plotted against the substrate (pAB1_FL905) concentration, and fitted into the classical Michaelis–Menten equation to determine KM, Vmax, and kcat.

Steady-State Kinetics of Type IIA DNA Topoisomerases

Next, we studied the steady-state kinetics of the type IIA DNA topoisomerases, E. coli DNA gyrase, and human DNA topoisomerase IIα. Both enzymes require ATP for catalytic activity.[1]E. coli DNA gyrase is a tetrameric protein and contains two different kinds of subunits, gyrA and gyrB that form an active A2B2 complex.[1] DNA gyrase is the only known DNA topoisomerase that actively introduces (−) supercoils in DNA substrates under normal reaction conditions and therefore requires hydrolysis of ATP.[1−3] DNA gyrase only exists in bacteria, is an essential enzyme,[54−56] and could be used to identify inhibitors without significantly affecting host human enzymes.[10,57]Figures and S4 show our kinetic results. Addition of DNA gyrase to solutions containing rx pAB1_FL905 resulted in a significant decrease of the fluorescence intensity of the solution at 521 nm that reached the plateau within 300–400 s (Figure B) and a complete (−) supercoiling of the rx pAB1_FL905 (Figure C). Because E. coli gyrase has two substrates, rx DNA pAB1_FL905 and ATP, we determined the pseudo first-order kinetics for both substrates by fitting the initial velocity results to the Michaelis–Menten equation (Figure D,E). For rx DNA pAB1_FL905, we kept the ATP concentration at 1.75 mM and varied DNA concentration from 0.75 to 7.2 nM to yield a KM, Vmax, and kcat of 2.7 ± 0.2 nM, 50 ± 2 pM/s, and 1.1 × 10–3 s–1, respectively. For ATP, we kept the concentration of rx pAB1_FL905 at 7.2 nM and varied the ATP concentration from 0.1 mM to 2 mM to generate a KM and Vmax of 0.33 ± 0.14 mM and 50 ± 7 pM/s, respectively. The KM values of DNA gyrase are consistent with previously published results using ATPase assays[58,59] and single-molecule techniques.[60]

Figure 6

Steady-state kinetics of E. coli DNA gyrase. (A) Proposed reaction scheme for the supercoiling reaction catalyzed by E. coli DNA gyrase. The reaction includes two substrates: ATP and DNA (pAB1_FL905). (B) Time courses of E. coli DNA gyrase-catalyzed DNA supercoiling reactions monitored by fluorescence intensity change. For the supercoiling reaction, 60 μL of 1× DNA gyrase buffer (35 mM Tris-HCl, 24 mM KCl, 4 mM MgCl2, 2 mM DTT, 1.75 mM ATP, 5 mM spermidine, 0.1 mg/mL BSA, 6.5% glycerol, pH 7.5) containing different concentrations of rx pAB1_FL905 was prepared and equilibrated to 37 °C, and 44.6 nM of E. coli DNA gyrase was used to supercoil the rx pAB1_FL905. The fluorescence intensity at λem = 521 nm was monitored with λex = 494 nm using a Horiba FluoroMax-3 spectrofluorimeter. (C) Time courses of E. coli DNA gyrase-catalyzed DNA supercoiling reactions monitored by 1% agarose gel electrophoresis. 384 μL of 1× DNA gyrase buffer containing 7.2 nM of rx pAB1_FL905 was prepared and equilibrated to 37 °C, and 44.6 nM of E. coli DNA gyrase was used to supercoil the rx pAB1_FL905. The reactions were stopped by adding 20 mM EDTA and 0.1% SDS into the reaction mixtures. Lanes 1–7 represent DNA samples from 0, 30, 60, 120, 300, 450, and 600 s of the supercoiling assay, respectively. rx and sc represent relaxed and supercoiled DNA molecules, respectively. (D) Initial velocities of supercoiling reaction were calculated from (B), plotted against the substrate (sc pAB1_FL905) concentration, and fitted into the classical Michaelis–Menten equation to determine KM, Vmax, and kcat for the DNA substrate (rx pAB1_FL905). (E) Initial velocities of supercoiling reaction were plotted against ATP concentration and fitted into the classical Michaelis–Menten equation to determine KM and Vmax for ATP.

Previously, it was demonstrated that novobiocin is a competitive inhibitor of DNA gyrase to prevent ATP binding to gyrase subunit B.[61] In this study, we also performed kinetic studies of E. coli DNA gyrase in the absence and presence of 60 nM of novobiocin. Figure shows our results. As expected, novobiocin significantly inhibited the supercoiling reaction of E. coli DNA gyrase (Figure A). Fitting of these kinetics to the Michaelis–Menten equation produced the following kinetic parameters: KM of 0.4 ± 0.1 mM and Vmax of 50 ± 3 pM/s in the absence of novobiocin and KM of 2.4 ± 0.1 mM and Vmax of 50 ± 2 pM/s in the presence of 60 nM novobiocin. Vmax is identical while KM is significantly higher in the presence of novobiocin. This result demonstrates that novobiocin is a competitive inhibitor of ATP for E. coli DNA gyrase. In the Lineweaver–Burk or double-reciprocal plot, the intercept on the y-axis of the plot of 1/V0 versus 1/[ATP] is the same in the presence or absence of novobiocin showing that novobiocin competes with ATP for its binding sites on E. coli DNA gyrase. The Ki value was calculated to be 11.6 nM which is consistent with the previous determination.[61]

Figure 7

Novobiocin is a competitive inhibitor of E. coli DNA gyrase to prevent ATP binding. (A) Inhibition of E. coli DNA gyrase catalyzed supercoiling reaction as a function of ATP. DNA supercoiling reactions were performed as described in “Materials and Methods” and also in the legend of Figure . The kinetic results were fitted to the classical Michaelis–Menten equation to yield KM, Ki, and Vmax. (B) Lineweaver–Burk plot or double-reciprocal plot of E. coli DNA gyrase in the absence or presence of novobiocin. Closed circles and squares represent supercoiling reactions in the absence and presence of 60 nM novobiocin, respectively.

Human DNA topoisomerase IIα is a homodimer with 1531 residues and 174 385 Da per monomer.[1,62] It can relax both (−) and (+) supercoiled DNA in the presence of ATP[1] and plays an essential role during mitosis and meiosis as a decatenase.[63] This enzyme is also a target of several important anticancer drugs, including doxorubicin and etopside.[10,64] Similar to DNA gyrase, human DNA topoisomerase IIα has two substrates, sc plasmid DNA and ATP. Therefore, we determined the pseudo first order steady-state kinetic parameters for both substrates (Figure S5). For sc DNA pAB1_FL924, we kept the ATP concentration at 2 mM and varied DNA concentration from 1.3 to 8 nM to yield KM, Vmax, and kcat of 5.7 ± 1.5 nM, 187 ± 25 pM/s, and 6.7 × 10–3 s–1, respectively. For ATP, we kept the concentration of sc pAB1_FL924 at 4 nM and varied the ATP concentration from 0.025 mM to 2 mM to generate a KM and Vmax of 0.22 ± 0.06 mM and 99 ± 8 pM/s, respectively. The KM value of human DNA topoisomerase IIα is consistent with previously published results using ATPase assays[65−68] and single-molecule techniques[22] for human and other eukaryotic DNA topoisomerase IIα.

Summary

Table summarizes the steady-state kinetic parameters for 6 DNA topoisomerases that we examined in this study. The Michaelis constants of these topoisomerases are all in the nanomolar range. Because KM can be used to estimate the strength of the enzyme substrate (ES) complex (Figure A), low KM values indicate strong binding of topoisomerases to DNA, which is consistent with previously published results.[34,69] Intriguingly, the kcat values of these 6 DNA topoisomerases are all low (Table ). To explain this, we are considering a kinetic pathway for topoisomerases (Figure B). The enzyme and substrate form an ES complex and become an enzyme-product complex through an activation ES* state. After relaxation or supercoiling steps, the topoisomerases may need to dissociate from the DNA product for the next round of catalysis. If topoisomerases still tightly bind to their product, koff should be low, which yields a low kcat. Indeed, the koff value of E. coli DNA topoisomerase I was determined previously and shown to be quite low[69] (0.017–0.043 s–1). In other words, the low values of kcat of E. coli DNA topoisomerase I stem from its tight binding to DNA. Recently, it was reported that Streptomyces coelicolor DNA topoisomerase I is a highly processive enzymes and can relax ∼150 (−) supercoils in a single burst in the single-molecule assay.[70] Steady-state kinetic parameters were also determined using a gel-based assay and pUC18[34] to yield a KM of 5.1 nM and a Vmax of 108 pM/s (6.5 nM/min).[70] Because 5 nM of DNA topoisomerase I was used, this produces a kcat of 2.2 × 10–2 s–1, comparable to kcat values determined in this article (Table ). Although the SDFQ method provides quantitative measurements for kinetic studies of DNA topoisomerases, we noticed some limitations for this method. This assay cannot be used to measure relaxation of (+) sc DNA. Additionally, this method only works in a relative narrow range of σ, that is, −0.02 ≥ σ ≥ −0.04 and cannot be used to study the transition of hypernegatively sc DNA to (−) sc DNA with σ ≤ −0.04.

Table 1

Steady-State Kinetic Parameters for DNA Topoisomerases

enzymea	subfamily type	KM (nM)	Vmax (pM/s)	kcat (s–1)
EcTopI	IA	1.2 ± 0.3b	50 ± 4b	3.0 × 10–3b
		1.5 ± 0.2c	40 ± 3c	2.3 × 10–3c
		7.5 ± 2.2 (ref (34))d	8.0 ± 1.0 (ref (34))d	1.2 ± 0.5 (ref (34))d
MsTopI	IA	4.3 ± 3.0	130 ± 5	9.8 × 10–3
vTopI	IB	2.4 ± 1.0	120 ± 23	6.0 × 10–3
hTopI	IB	2.7 ± 1.0	31 ± 5	1.2 × 10–3
EcGyrase	IIA	2.7 ± 0.2	50 ± 2	1.1 × 10–3
EcGyrasee (ATP, mM)	IIA	0.33 ± 0.14	51 ± 7
		0.3 (refs (58) and (61))d
		0.16 (ref (60))d
hTopIIα	IIA	5.7 ± 1.5	187 ± 25	6.7 × 10–3
hTopIIαe (ATP, mM)	IIA	0.2 ± 0.1	99 ± 8
		0.56 ± 0.17 (ref (65))d
		0.5 ± 0.3 (ref (68))d

EcTopI, E. coli DNA topoisomerase I; MsTopI, M. smegmatis DNA topoisomerase I; vTopI, variola DNA topoisomerase I; hTopI, human topoisomerase I; EcGyrase, E. coli DNA gyrase; hTopIIα, human DNA topoisomerase IIα.

These kinetic parameters were determined in 1× CutSmart buffer.

These kinetic parameters were determined in 1× CutSmart buffer plus 100 mM NaCl.

These values were determined in the cited references.

The concentration unit for KM of ATP is mM.

Figure 8

(A) Simplified reaction pathway for the classical Michaelis–Menten enzyme kinetics. (B) Proposed kinetic pathway for DNA topoisomerases. DNA topoisomerases may need to dissociate from the DNA product before the next round of catalysis. koff represents the off rate constant for DNA topoisomerases dissociating from the product.

In conclusion, we developed a rapid and efficient method to generate rx and sc fluorescently labeled plasmid DNA molecules for DNA topoisomerases. Utilizing the fluorescently labeled rx and sc pAB1_FL905 and pAB1_FL924, we found that all 6 DNA topoisomerases including type IA, IB, and IIA topoisomerases follow the classical Michaelis–Menten kinetics. We determined the steady-state kinetic parameters, that is, KM, Vmax, and kcat, for these DNA topoisomerases. Using this kinetic study, we also confirmed that novobiocin is a competitive inhibitor of DNA gyrase to prevent ATP binding.