Characterization of the interaction between DNA gyrase inhibitor and DNA gyrase of Escherichia coli.

Escherichia coli DNA gyrase is comprised of two subunits, GyrA and GyrB. Previous studies have shown that GyrI, a regulatory factor of DNA gyrase activity, inhibits the supercoiling activity of DNA gyrase and that both overexpression and antisense expression of the gyrI gene suppress cell proliferation. Here we have analyzed the interaction of GyrI with DNA gyrase using two approaches. First, immunoprecipitation experiments revealed that GyrI interacts preferentially with the holoenzyme in an ATP-independent manner, although a weak interaction was also detected between GyrI and the individual GyrA and GyrB subunits. Second, surface plasmon resonance experiments indicated that GyrI binds to the gyrase holoenzyme with higher affinity than to either the GyrA or GyrB subunit alone. Unlike quinolone antibiotics, GyrI was not effective in stabilizing the cleavable complex consisting of gyrase and DNA. Further, we identified an 8-residue synthetic peptide, corresponding to amino acids (89)ITGGQYAV(96) of GyrI, which inhibits gyrase activity in an in vitro supercoiling assay. Surface plasmon resonance analysis of the ITGGQYAV-containing peptide-gyrase interaction indicated a high association constant for this interaction. These results suggest that amino acids 89--96 of GyrI are essential for its interaction with, and inhibition of, DNA gyrase.

DNA gyrase is a type II topoisomerase found in all bacteria and is essential for viability. It is involved in the replication, repair, recombination, and transcription of DNA (1). DNA gyrase consists of two subunits, GyrA and GyrB, and the active holoenzyme is an A 2 B 2 complex (2). Mechanistic studies have revealed the steps involved in the gyrase supercoiling reaction (3)(4)(5). This process involves the wrapping of DNA around the A 2 B 2 complex, cleavage of this DNA on both strands, and the passage of a segment of DNA through the double strand break. Religation of the break results in the introduction of two negative supercoils. These processes require the binding and hydrolysis of ATP (6,7).
DNA gyrase is the target of several antibacterial agents, including the coumarin and quinolone families of antibiotics (8,9). Coumarins are potent inhibitors of gyrase supercoiling and ATPase reactions. Through the use of x-ray crystallography the molecular details of the ATPase reaction and its inhibition by coumarin have been revealed (10 -12). Quinolones, on the other hand, stabilize a conformation of the enzyme-DNA complex referred to as the "cleavable complex." This complex blocks the passage of RNA polymerases (13,14).
Although the majority of compounds known to target DNA gyrase are either quinolones or coumarins, natural toxins also exist such as microcin B17 (15) and CcdB (16). Microcin B17 is a peptidyl antibiotic secreted in the stationary phase by several strains of Escherichia coli and induces DNA cleavage in a fashion similar to that of the quinolones (17). The antibiotic efficacy of this polypeptide depends on the post-translational modification of eight of its cysteine and serine residues to thiazoles and oxazoles, respectively (18 -21). CcdB is a bacterial toxin encoded by the E. coli F plasmid (22). These natural toxins target domains of the GyrA and GyrB proteins distinct from those targeted by the quinolones and coumarins (23). Microcin B17 targets GyrB (24) and CcdB targets GyrA (23). No cross-resistance to quinolones has been observed.
DNA gyrase inhibitor (GyrI) 1 was the first chromosomally nucleoid encoded DNA gyrase regulatory factor identified in E. coli. In previous work, we found that this purified protein inhibited DNA supercoiling in vitro and that abnormally high or low GyrI levels resulted in the suppression of cell proliferation in vivo (25). The promoter activity of gyrI increases in the late exponential phase and peaks in the stationary phase (25). Independently, Baquero et al. (26) reported that the gyrI gene (identical to the sbmC gene) belongs to the SOS regulon. The inactivation of GyrI causes cells to be more sensitive to microcin B17, whereas its overexpression yields microcin B17-resistant cells (26).
In this work, we examined the interaction of the GyrI protein with the DNA gyrase A 2 B 2 tetramer. We show that the interaction of GyrI with DNA gyrase requires both the GyrA and GyrB subunits and that a specific region of GyrI mediates its interaction with, and inhibition of, DNA gyrase. vectors used in the cloning of the gyrB gene were pET16b and pET24b. The pET16b and pET24b vectors contain a His tag sequence, which is a consecutive stretch of histidine residues which can be expressed at either the amino or carboxyl terminus of the target protein. The pET3a vector contains a T7-tagged sequence that can be expressed at the amino terminus of the target protein. We engineered recombinant genes that express GyrA and GyrB proteins containing histidine tags on their amino-and carboxyl-terminal ends, respectively. The expressed GyrI protein contains a T7 tag on its amino-terminal end. Genes were cloned by PCR using the following primers: for gyrA, 5Ј-CAGCATATG-AGCGACCTTGCGAGAGAA-3Ј and 5Ј-ATAGGATCCGTACGACAAAA-GCCCAGAC-3Ј; for gyrB, 5Ј-CGAGGATCCGATGTCGAATTCTTAT-3Ј and 5Ј-CGAGGATCCGCCATTAAATATCGAT-3Ј (pET16b), 5Ј-CCAGG-ATCCGATGTCGAATTCTTAT-3Ј and 5Ј-AATGGATCCATATCGATAT-TC-GGCG-3Ј (pET24b); for gyrI, 5Ј-TCCCATATGAACTACGAGATTA-AGCAGGA-3Ј and 5Ј-TATGCATATGATGTTTTGGCTGCACCGCAA-3Ј. 30 cycles of PCR were carried out at 94°C for 1 min, 51°C for 1 min, and 72°C for 3 min. The gyrA gene was inserted into the NdeI-BamHI sites of pET16b, the gyrB gene was inserted into the BamHI site of pET24b, and the gyrI gene was inserted into the NdeI site of pET3a, yielding pCA30, pCA31, and pCA32, respectively. All constructs were verified by DNA sequencing using an ABI Prism 377 DNA sequencer (PerkinElmer Life Sciences, Applied Biosystems).
Growth and Lysis of Cells-The pCA30, pCA31, and pCA32 plasmids were transformed separately into E. coli BL21(DE3)pLysS. Transformed cells were inoculated into 800 ml of LB medium and incubated at 30°C with vigorous shaking until an A 600 value of 0.5 was reached. Growth of BL21(DE3)pLysS transformed with pCA30, pCA31, or pCA32 was induced by the addition of 0.5 mM isopropyl 1-thio-␤-Dgalactopyranoside. After 2 h, cultures were cooled rapidly on ice, and cells were harvested. Cell lysates were prepared upon the addition of lysozyme and Brij-58 (27).
Protein Purification-The GyrI protein was purified as described previously (25). Originally, we used Talon to purify histidine-tagged GyrA and GyrB; however, the respective purities of these proteins were insufficient for this study. We therefore purified the GyrA and GyrB proteins according to the method of Mizuuchi et al. (28) with the following modifications. GyrA proteins purified on a DEAE-Sepharose column were loaded onto a Sepharose CL-4B column (Amersham Biosciences, Inc.) preequilibrated with TGED buffer (50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM DTT, 10% glycerol). For purification of the GyrB protein, the GyrB fraction eluted from a heparin-agarose column was loaded onto a Sepharose Q column (Amersham Biosciences, Inc.) equilibrated with 0.025 M NaCl in TGED-2 (50 mM Tris-HCl (pH 7.5), 0.2 mM EDTA, 1 mM DTT, 10% glycerol). Proteins were eluted with a 100-ml linear gradient of 0.025-0.8 M NaCl in TGED-2 buffer. The fractions containing GyrB were dialyzed for 4 h against 2 liters of TGED buffer and loaded onto a column of Sepharose CL-4B equilibrated with TGED buffer. The fractions containing GyrA were identified by their ability to enhance the supercoiling activity of fractions containing GyrB. Active fractions of GyrA, GyrB, and GyrI were frozen in liquid nitrogen and stored at Ϫ80°C.
Gyrase-mediated DNA Cleavage-Cleavage assays were performed as described by Bernard et al. (29). Briefly, 0.2 g of PstI-digested linear pBR322 DNA was incubated with 179 ng of gyrase in 20 l of assay buffer (35 mM Tris-HCl (pH 7.5), 47 mM KCl, 7 mM MgCl 2 , 1 mM ATP, 3 mM DTT, 100 g/ml bovine serum albumin, 40 g/ml E. coli tRNA) at 30°C for 50 min. After the addition of either GyrI (26, 52, or 105 ng) or oxolinic acid (4 g), samples were incubated further at 30°C for 50 min. Reactions were stopped by the addition of SDS and proteinase K at a final concentration of 0.2% and 0.1 mg/ml, respectively. DNA was analyzed by 1% (v/v) agarose gel electrophoresis.
DNA Supercoiling Activity of DNA Gyrase-The supercoiling activity was measured as described previously (25).
Immunoprecipitation-GyrI-gyrase complexes were formed by mixing 12 g/ml T7-tagged GyrI and gyrase (48 g/ml GyrA, 220 g/ml GyrB) in a buffer consisting of 25 mM Tris-HCl (pH 7.5), 67 mM KCl, 5 mM MgCl 2 , 1.25 mM spermidine hydrochloride, 1.7 mM ATP, 5 mM DTT, 75 g/ml bovine serum albumin, and 150 g/ml E. coli tRNA. Protein G-Sepharose beads (5 l) were added to the mixture. After incubating for 30 min at 4°C, the mixture was centrifuged at 10,000 rpm for 10 s, and 1 g of an anti-T7 antibody and 5 l of protein G-Sepharose beads were added to the supernatant. After gentle rotation for 60 min at 4°C, the protein G-beads were collected by centrifugation, washed four times with TBS buffer, and boiled in SDS sample buffer for 5 min. Immunoprecipitated proteins were resolved by 12% SDS-PAGE and assayed by Western blotting. Proteins were probed with either an anti-His or T7 antibody. Horseradish peroxidase-conjugated anti-rabbit IgG and anti-mouse IgG were used as secondary antibodies for the His and T7 antibodies, respectively. Immunocomplexes were revealed with an ECL kit according to the manufacturer's protocol (30).
Determination of Binding Constants Using SPR-SPR experiments were performed on a FAST IAsys Cuvette System (Affinity Sensors, Cambridge, U. K.) (32)(33)(34). GyrI was immobilized on the sensor chip surface using an IAsys carboxymethyl-dextran cuvette (FCD-5101; Affinity Sensors) according to the manufacturer's instructions. The immobilization procedure was done as follows. The cuvette was washed with PBS/Tween 20 buffer and then activated by a 7-min incubation with a mixture of 50 l of 13 mg/ml N-hydroxysuccinimide and 50 l of 76 mg/ml N-ethyl-NЈ- (3-(diethylamino)propyl)carbodiimide to form succinimidyl esters that react with amino groups. After washing with PBS/Tween 20 buffer, either 50 l of 200 g/ml GyrI or 1 mM of the synthetic peptide (corresponding to various sequences within GyrI) in 50 mM sodium phosphate buffer (pH 8.0) was added to the cuvette, and after a response plateau was reached signifying completion of the reaction, the cuvette was washed with PBS/Tween 20 buffer. Any remaining activated groups were blocked by incubating the cuvettes with 1 M ethanolamine (pH 8.5) for 2 min. After reequilibration with PBS/Tween 20 buffer, the cuvette was ready for the binding experiments. The cuvette matrix was washed with 10 mM HCl to remove noncovalently bound ligand (GyrI), and reequilibrated with PBS/Tween 20 buffer. Binding experiments were performed in 50 mM sodium phosphate buffer (pH 8.0). Various concentrations of either GyrA or GyrB were injected over the immobilized proteins, and molecular associations were observed for 6 min. At the end of the incubation, the surface was washed three times with PBS/Tween 20 buffer, and dissociation was followed for 5 min. The surface was regenerated by washing with 10 mM HCl for 2 min followed by reequilibration in PBS/Tween 20 buffer.

RESULTS
DNA Gyrase Supercoiling and Cleavage Assay Using Purified GyrI and DNA Gyrase-We engineered recombinant fusion genes expressing GyrA, GyrB, and GyrI proteins containing histidine or T7 tags on their COOH-or NH 2 -terminal ends. These recombinant proteins were purified by chromatography as described under "Experimental Procedures." SDS-PAGE analysis revealed that the sizes of GyrA, GyrB, and GyrI correlated with their estimated molecular masses (Fig. 1A). The holoenzyme, obtained by mixing recombinant GyrA and GyrB, possessed DNA supercoiling activity as determined using relaxed pBR322 DNA as substrate. DNA supercoiling was inhibited by novobiocin and oxolinic acid in a dose-dependent manner (Fig. 1B) and inhibited by the GyrI protein (Fig. 1C). We next verified the capacity of purified GyrI and oxolinic acid to facilitate DNA gyrase-mediated cleavage of pBR322 plasmid DNA. As shown in Fig. 1D, oxolinic acid facilitated the cleavage of pB322 DNA by gyrase (lane 6), whereas GyrI could not position DNA gyrase into a cleavable complex, such that upon the addition of SDS, a double strand cleavage of the DNA was not observed (Fig. 1D, lanes 2, 3, and 4). Similarly, when GyrI was assayed for in vitro DNA gyrase activity in the absence of ATP, no DNA cleavage was observed (Fig. 1D, lane 1). However, inhibition of DNA supercoiling was observed when GyrI was added to the supercoiling reaction (Fig. 1C) (25).
Interaction between GyrI and Gyrase-The next step was to clarify the interaction between GyrI and the DNA gyrase holoenzyme as well as with the individual subunits GyrA and GyrB. Purified T7-tagged GyrI was incubated with His-GyrA, His-GyrB, or His-holoenzyme, and the reaction mixtures were immunoprecipitated with a His tag-specific antibody. The immunoprecipitants were resolved by 12% SDS-PAGE and sub-jected to Western blotting using an antibody specific for the T7 tag. As shown in Fig. 2A, GyrI was detected in an ATP-independent manner in the immunoprecipitates containing GyrI and holoenzyme. However, in the immunoprecipitants recovered from reaction mixtures containing GyrI and either GyrA or GyrB, GyrI was only weakly detected (Fig. 2B). Because His-GyrB was undetectable by the His antibody (lane 2), we constructed a COOH-terminally His-tagged GyrB subunit. Although the His antibody C-recognized terminally His-tagged GyrB, no GyrI was found in the His-immunoprecipitates containing a mixture of GyrI and GyrB (Fig. 2C). These results indicate that GyrI binds only to the DNA gyrase holoenzyme and displays little or no binding activity toward the individual GyrA or GyrB subunits.

SPR Kinetic Analysis of GyrI
Binding-To characterize further the interaction of GyrI with the holoenzyme, we determined the kinetic parameters governing the interaction of GyrI with either GyrA, GyrB, or the holoenzyme using SPR. For these experiments, GyrI was immobilized in the cuvette, and the binding of GyrA, GyrB, or the holoenzyme was measured. Analysis of these interactions using the FAST FIT program indicated that all of the association and dissociation phases were consistent with a monophasic reaction model. The association rate for the binding between the analyte (GyrA, GyrB, or the holoenzyme) and the immobilized substrate (GyrI) is expressed in the Equation 1  5). The reactions were stopped by the addition of SDS and proteinase K and were incubated further at 37°C for 30 min. After removal of proteins, the samples were analyzed on 1.0% agarose gels in TAE buffer.  (lanes 1, 4, 6, and 8), GyrB (lanes 2, 5, 6, and 9), and GyrI (lanes 3-7) were incubated with the reaction mixtures. Precipitates were subjected to Western blot analysis using anti-T7 or anti-His antibody. Panel C, GyrA (lanes 4 and 6), GyrB (lanes 2, 3, 4, and 7), and GyrI (lanes 1, 3, 4, and 5) were incubated with the reaction mixtures. The histidine tag of GyrB was positioned on its COOH-terminal rather than on its NH 2 -terminal side. Precipitates were subjected to Western blot analysis using anti-T7 or anti-His antibody.
where K a is the association rate constant, K d is the dissociation rate constant, R max is the maximum binding capacity of the immobilized substrate surface as determined by saturation with analyte, R t (Arc seconds) is the amount of bound analyte measured as the SPR response at time t, and [C] is the concentration of analyte added to the cuvette. The affinity constant, K D , is then calculated from K d /K a . In Fig. A, binding curves are depicted for the interaction of GyrA, GyrB, or the GyrA/B holoenzyme (250 nM) with immobilized GyrI. GyrI bound poorly to the individual GyrA and GyrB subunits compared with its binding affinity for the GyrA/B holoenzyme (Fig. 3A). In Fig.   3B, an overlay plot is shown for the association profiles used by the FAST FIT program to determine the K on value for each concentration of the holoenzyme (31.2-600 nM). As for the determination of K d , we obtained it directly by measuring the dissociation phase using the FAST FIT program in Fig. 3A. The dissociation rate constant of GyrA was similar to that of GyrB for the interaction with the immobilized GyrI. On the other hand, in the presence of GyrA/B holoenzyme, binding to the immobilized GyrI exhibited slow dissociation. A series of profiles for GyrA, GyrB, and the holoenzyme binding to GyrI was used to generate the kinetic constants shown in Table I. These results clearly support those of the immunoprecipitation experiments, indicating that the interaction of Gyr I with DNA gyrase is dependent on the presence of both the Gyr A and GyrB subunits.
Inhibition of DNA Gyrase by GyrI-derived Synthetic Peptides-To identify the region of GyrI involved in its inhibitory interaction with DNA gyrase, we used the DNA gyrase supercoiling reaction to examine the effect of overlapping synthetic peptides (Fig. 4A) corresponding to various sequences in GyrI that ultimately covered its whole sequence. The peptides underlined in Fig. 4A inhibited DNA supercoiling in a dose-dependent manner (12.5-100 M), reaching complete inhibition at a concentration of 100 M (final concentration) (Fig. 4B). Importantly, these peptides did not inhibit the activity of topoisomerase I (Fig. 4C). To define further the region mediating the inhibitory effect of GyrI on DNA gyrase, we synthesized peptides of various sizes within the region bordered by residues 81-112 of GyrI (Table II). It was found that peptides containing the motif ITGGQYAV inhibited DNA supercoiling with maximal efficacy. The concentrations of the ITGGQYAV-containing peptides required for 50% inhibition were remarkably similar (12.3-23.6 M), although the peptide corresponding to amino acids 89 -96 of GyrI was especially potent.
SPR Kinetic Binding Analysis of Synthetic Peptides Corresponding to Various Sequences within GyrI-We next quantified the interaction between the ITGGQYAV-containing peptide and DNA gyrase using SPR. The peptide ITGGQYAVAVARVVGD (GyrI residues 89 -104) or the peptide LMMWVDSKNIVPKEWV (GyrI residues 33-48), used as a negative control, was immobilized in the cuvette, and the binding of gyrase to the immobilized peptides was measured. The time course of gyrase binding to both peptides is shown in Fig.  5A. The complex between the ITGGQYAV-containing peptide and gyrase was found to be relatively stable compared with that of the peptide corresponding to GyrI 33-48 (Table III), which had no inhibitory effect on DNA gyrase activity. In  taining peptide (Table III). The dissociation rate (K d ϭ 3.7 ϫ 10 Ϫ3 s Ϫ1 ), which can be directly obtained by analyzing the dissociation phase using the FAST FIT program, is shown in Table III. A K D value of 2.2 M was determined from the ratio of K d to K a for the ITGGQYAV-containing peptide, whereas a weaker dissociation constant K D (26.0 M) was observed for the GyrI residues 33-48 (Table III). DISCUSSION We previously discovered GyrI during the purification of DNA gyrase from E. coli (25). Here, we wished to characterize the molecular interaction between GyrI and DNA gyrase. To do so, DNA gyrase was partially purified by affinity chromatography on a novobiocine-Sepharose column (25,37,38). The GyrA subunit was eluted with 2 M KCl, and the GyrB subunit was eluted with a combination of 2 M KCl and 5 M urea. The DNA gyrase holoenzyme, consisting of two subunits (i.e. A (GyrA) and B (GyrB)), was eluted with 5 M urea. The GyrI protein coeluted with the fraction containing the two subunits. This result suggests that GyrI interacts with the holoenzyme consisting of GyrA and GyrB, but not with GyrA or GyrB alone.  Table II. Panel C, inhibitory effects of GyrI protein-derived synthetic peptides (100 M) on topoisomerase I activity.   To investigate the interaction of GyrI with gyrase, we used SPR and immunoprecipitation. The results of the immunoprecipitation experiments suggest that GyrI binds directly to the holoenzyme and that binding is also observed, albeit to a lesser degree, with the individual GyrA and GyrB subunits. We also showed that ATP is not required for the interaction between GyrI and the gyrase holoenzyme. Kinetic parameters from SPR experiments used to examine the interaction between GyrI and GyrA, GyrB, or the holoenzyme supported the immunoprecipitation results. The association rate constant for the formation of the GyrI-holoenzyme complex was determined to be ϳ10 4 M Ϫ1 s Ϫ1 . This is similar to that for the association of CcdB and GyrA. Like GyrI, CcdB is a natural toxin that targets gyrase. Analysis of the CcdB-gyrase interaction reveals that this complex forms through the direct binding of GyrA to CcdB (39). The inhibitory mode of action of CcdB appears to be similar to that of the quinolones on gyrase (40). It would be very interesting to determine the structure of gyrase complex containing GyrI.
We used GyrI-derived synthetic peptides to determine the regions of GyrI involved in its binding to DNA gyrase. The peptide corresponding to an 8-amino acid sequence in GyrI (ITGGQYAV; 89 -96 of GyrI) was sufficient to inhibit the supercoiling activity of DNA gyrase. This peptide did not affect topoisomerase I activity, demonstrating that the inhibitory effect of the synthetic peptide was not an artifact of the supercoiling assay.
Gyrase is a target of several antibiotics, some of whose mechanisms have been characterized in detail. Coumarin binds the amino-terminal domain of the GyrB subunit and inhibits gyrasedependent ATP hydrolysis. In contrast, quinolone antibiotics bind a cleavable complex of DNA bound to DNA gyrase. When we introduced the GyrI-peptide (ITGGQYAV (200 M); 89 -96 of GyrI) into the cleavage assay, no cleaved DNA-gyrase complexes were observed (data not shown). This result suggests that the inhibitory mechanism of the ITGGQYAV peptide might be different from that of the quinolones. It would be interesting to modify this peptide to develop a more effective inhibitor of gyrase activity. The discovery and analysis of natural molecules with anti-gyrase activity hold promise for further investigations into mechanism of DNA gyrase activity. Finally, the GyrI peptides we describe here may form the basis for a novel class of peptide-based antibacterial agents.