at SciVerse ScienceDirect

European Journal of Medicinal Chemistry 54 (2012) 591e596

Contents lists available

European Journal of Medicinal Chemistry

journal homepage: http: / /www.elsevier .com/locate/ejmech

Original article

Design and synthesis of peptides from bacterial ParE toxin as inhibitorsof topoisomerases

Luiz Carlos Bertucci Barbosa, Saulo Santesso Garrido, Anderson Garcia, Davi Barbosa Delfino,Laura do Nascimento Santos, Reinaldo Marchetto*

UNESP e Institute of Chemistry, Department of Biochemistry and Technological Chemistry, Caixa Postal 355, 14800-900 Araraquara, São Paulo, Brazil

a r t i c l e i n f o

Article history:Received 27 April 2012Received in revised form4 June 2012Accepted 5 June 2012Available online 15 June 2012

Keywords:PeptidesTA systemsParE toxinDNA topoisomerasesEnzyme inhibition

* Corresponding author. Tel.: þ55 16 3301 9670; faE-mail address: marcheto@iq.unesp.br (R. Marche

0223-5234/$ e see front matter � 2012 Elsevier Mashttp://dx.doi.org/10.1016/j.ejmech.2012.06.008

a b s t r a c t

Toxineantitoxin (TA) proteic systems encode a toxin and an antitoxin that regulate the growth and deathof bacterial cells under various stress conditions. The ParE protein is a toxin that inhibits DNA gyraseactivity and thereby blocks DNA replication. Based on the Escherichia coli ParE structure, a series of linearpeptides were designed and synthesized by solid-phase methodology. The ability of the peptides toinhibit the activity of bacterial topoisomerases was investigated. Four peptides (ParELC3, ParELC8, Par-ELC10 and ParELC12), showed complete inhibition of DNA gyrase supercoiling activity with an IC100

between 20 and 40 mmol L�1. In contrast to wild-type ParE, the peptide analogues were able to inhibit theDNA relaxation of topoisomerase IV, another type IIA bacterial topoisomerase, with lower IC100 values.Interestingly only ParELC12 displayed inhibition of the relaxation activity of human topoisomerase II. Ourfindings reveal new inhibitors of bacterial topoisomerases and are a good starting point for the devel-opment of a new class of antibacterial agents that targets the DNA topoisomerases.

� 2012 Elsevier Masson SAS. All rights reserved.

1. Introduction

DNA topoisomerases are good targets for antibacterial chemo-therapy due to their nature and to their mechanisms of action inessential processes that control the topological state of DNA in cells[1e3]. Bacterial DNA gyrase and topoisomerase IV (topo IV) areexamples of type II topoisomerases, enzymes that act by breakingboth strands of duplex DNA and catalyzing passage of another DNAduplex through the break [1]. All topoisomerases can relax super-coiled DNA, but only DNA gyrase can also introduce negativesupercoils into DNA in the presence of ATP [4]. On the other hand,topo IV is the only topoisomerase significantly involved in deca-tenation and unknotting of DNA molecules [5,6]. Whereaseukaryotic topoisomerase II enzymes are large single-subunit,active as homodimers, DNA gyrase and topo IV are composed oftwo subunits: GyrA and GyrB for gyrase and ParC and ParE for topoIV [3,7]. For each enzyme, these subunits combine into a hetero-tetrameric (gyrase GyrA2GyrB2 and topo IV ParC2ParE2) complex toform the active enzymes [7]. The ParE subunit and the corre-sponding gyrase B-subunit (GyrB) are the subunits responsible forATP binding and hydrolysis and GyrA and ParC are responsible forDNA binding and the cleavage and religation reaction [8,9].

x: þ55 16 3322 2308.tto).

son SAS. All rights reserved.

Besides the structural similarity and the catalytic mechanism,gyrase and topo IV also have similarities in their sensitivity to bothsynthetic and natural topoisomerase inhibitors, such as quinolonesand coumarins [10]. Quinolones are synthetic drugs based on the 4-oxo-1,4-dihydroquinolone skeleton and are by far the mostsuccessful antibacterials targeted to topoisomerases. Quinolonesinteract with the enzyme and DNA, forming an Mg2þ-mediatedternary complex and in so doing disrupt the DNAcleavageereligation process. However the exact interactions of thedrugs with the proteins are not entirely clear but it is likely thatthese involve the Ser and Asp (or Glu) residues in GyrA that arecommonly mutated in quinolone-resistant bacteria and a Mg2þ ion[11e13]. Although gyrase is the target for quinolones in Gram-negative bacteria, topo IV can be the preferred target in someGram-positive organisms [13]. The coumarins, also referred to asaminocoumarins, are natural products of Streptomyces species,which inhibit ATPase activity of DNA gyrase and topo IV bycompeting with ATP for binding to the GyrB and ParE subunits ofthe enzymes respectively [14e17]. Other inhibitors as cyclo-thialidine [18], albicidin [19] and the plasmid-encoded proteinsmicrocin B17 [20], CcdB [21] and ParE [22] inhibit DNA gyrase butdo not inhibit topo IV or inhibit only at concentrations above thoseinhibiting gyrase [10].

ParE is a 12 kDa protein that is produced as a part of a two-component toxineantitoxin system by the RK2 plasmid of E. coli[22]. A bacterial toxineantitoxin (TA) system generally is composed

Fig. 1. Ribbon diagram of the E. coli ParE. The C-terminal a-helix and the three-stranded b-sheets, b2, b3 and b4, used in the peptides design, are colored in red,orange, blue and magenta, respectively. LNIES terminal sequence is in yellow. (Forinterpretation of the references to colour in this figure legend, the reader is referred tothe web version of this article.)

Fig. 2. Structure-based C-terminal sequence alignment and synthesized peptides. Thealignment was performed using C-terminal sequences of E. coli and C. crescentus ParEtoxins. E (in blue): b-sheet representations; H (in red): a-helices representations.Conserved residues are highlighted in gray. ParELC1 to ParELC12: primary structures ofthe synthesized peptides from C-terminal E. coli ParE. (For interpretation of thereferences to colour in this figure legend, the reader is referred to the web version ofthis article.)

L.C.B. Barbosa et al. / European Journal of Medicinal Chemistry 54 (2012) 591e596592

of two genes organized in an operon encoding a toxin and anantitoxin that antagonizes it [23]. TA systems enhance plasmidstability in cell populations by a mechanism called post-segregational killing. This mechanism relies on the differentialstabilities of the antitoxin and toxin proteins and leads to the killingof daughter cells that did not receive a plasmid copy at the celldivision [23]. The operon parDE encode a TA system formed by ParEtoxin and its antitoxin ParD. ParD is able to neutralize ParE actionand is effective in autoregulation of the parDE operon [24]. ParE isincluded in RelE superfamily which is composed by several smallertoxins such as RelE, YoeB, YafQ, HigB and ParE [25,26]. The toxins ofthe RelE superfamily are characterized by a core of b-sheet struc-tures which are flanked on one side by two N-terminal alpha-helixes and on the other side by one C-terminal alpha-helix [26].Crystallographic studies of ParDeParE complex from Caulobactercrescentus showed that ParE and RelE has a high homology at thelevel of primary sequence and tertiary structure, as predicted by themolecular model proposed for ParE from E. coli [27,28]. Unlike RelEthat inhibits translation by promoting cleavage of mRNA in theribossomal A-site, ParE blocks DNA replication by inhibiting DNAgyrase [22,29,30]. However the interactions between ParE andgyrase subunits and the mechanism by which ParE inhibits gyrasehave not been explored. In addition, no evidence of the inhibitoryactivity of ParE has been found for topo IV. So, based on ParEstructure we have designed and synthesized a series of linearanalogues of ParE and investigated by an in vitro assay the ability ofpeptides to inhibit bacterial topoisomerases activity as a means toelucidate the region of ParE responsible for proteineproteininteractions.

2. Results and discussion

2.1. Peptide design

Previous studies of ParE toxin from plasmid RK2 have shownthat C-terminal truncations attenuate protein toxicity [29]. Studieswith C. crescentus ParE have also shown that residues at the C-terminus of ParE are critical for its stability and toxicity [30]. Thecrystal structure of C. crescentus ParDeParE complex revealedimportant three-stranded b-sheet in the C-terminal residues [27].For E. coli ParE (Fig. 1), the C-terminal structure predicted bymolecular modeling showed a core of three-stranded b-sheets thatencompasses the residues L61 to F87 as well as one a-helixbetween M91 and R100 residues [28]. The structure-based align-ment (Fig. 2) shows a high degree of similarity between thesecondary structure of C. crescentus ParE and the structure pre-dicted for E. coli ParE, despite a low level of primary identity.

In this context, for peptide designwe considered the 45 residuesof C-terminal sequence of E. coli ParE that include the C-terminal a-helix (absent in the C. crescentus ParE) and the three-stranded b-sheets (b2, b3 and b4). The peptides size ranged from 8 to 45 aminoacids residues, including or changing secondary structures in orderto identify the region with more importance for the toxin activity.Therefore, the two first peptides (ParELC1 and ParELC2) were builtso as to contain the C-terminal a-helix (a3). The inclusion of the b-sheet b4 (P80 e F87) to ParELC2 resulted in the third peptidesequence ParELC3. To verify if the five C-terminal residues of E. coliParE (Leu, Asn, Ile, Glu and Ser) play a role in the complex formationwith enzymes, as described for CcdB analogues [31] this sequencewas introduced in the first three peptides, yielding ParELC4, Par-ELC5 and ParELC6, respectively (Fig. 2). ParELC7 and ParELC8 werebuilt as ParELC6 but introducing one (b3) or two b-sheets (b2 and b3)in the N-terminal end, respectively. Finally, the last four peptideswere prepared to contain only b-sheets: ParELC9 (b3 and b4), Par-ELC10 (b2, b3 and b4), ParELC11 (only b2) and ParELC12 (b2 and b3).

All peptide analogues described and presented in Fig. 2, weresynthesized by the solid-phase method employing the Fmoc-strategy and considering the usual polymer solvation parameters[32].

2.2. Inhibition activities of designed peptides

2.2.1. DNA gyraseGel electrophoresis assay was employed to evaluate the ability

of the peptides to inhibit the supercoiling reaction of gyrase. DNAgyrase introduces negative supercoiling into closed-circular DNAgenerating a molecule which migrates more rapidly compared torelaxed closed-circular DNA during gel electrophoresis [33]. Theassay was performed by titrating ParE analogues into a fixedconcentration of gyrase and DNA and determination of IC100 (theminimum concentration that produces complete inhibition ofsupercoiling activity). In the standard supercoiling assay at 37 �C aninitial screening (Fig. 3A) selected ParELC3, ParELC8, ParELC10 and

Fig. 3. Screening of synthesized peptides with inhibitory activity on supercoiling of gyrase (A) and relaxation of topo IV (B) reactions. Controls were no inhibitor (lanes C) andtreatment with ciprofloxacin (10 mmol L�1), the known topoisomerase inhibitor (lanes CFX). Lanes 1 to 12: presence of ParELC1 to ParELC12 peptides (100 mmol L�1), respectively.

L.C.B. Barbosa et al. / European Journal of Medicinal Chemistry 54 (2012) 591e596 593

ParELC12 as better gyrase inhibitors (complete inhibition at100 mmol L�1). Subsequently the IC100 for the selected peptideswere determined (Fig. 4) and showed that ParELC10was the better,inhibiting completely the supercoiling reaction of gyrase with anIC100 value of 20 mmol L�1 (Table 1).

One possibility to explain the inhibitory activity of the selectedpeptides is that the core of three-stranded b-sheets that encom-passes the residues L61 to F87 is important for interactions with thegyrase. ParELC1, ParELC2, ParELC4 and ParELC5, peptides that do notcontain any one of the three b-sheets, did not inhibit DNA gyrase.Moreover, ParELC10 (three-stranded b-sheets) had the lowest IC100value. Comparatively, the b4-sheet seems plays a key role in the

Fig. 4. IC100 determination of ParELC3, ParELC8, ParELC10 and ParELC12 for DNAsupercoiling reactions of DNA gyrase (3.4 nmol L�1). Controls were no peptide (lane 0)and treatment with ciprofloxacin 10 mmol L�1 (lane CFX).

peptideeenzyme interactions. All active peptides, except ParELC12,have in its primary structure, the amino acids corresponding to theb4-sheet. In fact, the lack of this amino acid sequence in the Par-ELC12 structure resulted in the less active peptide among thoseselected. The toxicity of ParELC3 can be also explained by thepresence of two of the three highly conserved amino acids (H88and M91) in the primary structure of ParE toxins of several bacteria(Fig. 5). Apparently, these amino acids also play an important role inthe molecular interactions between the enzyme and the toxin fromdifferent sources. In addition, the insertion of the LNIES sequence atthe C-terminal extreme has rendered peptides non-toxic. Thiseffect can be best viewed by comparing the structureeactivity ofParELC3 and ParELC6. Furthermore, the low toxic activity caused byLNIES sequence can be also evaluated by ParELC8 peptide. Struc-turally this analogue contains the core of three-stranded b-sheetsand the C-terminal a-helix. So, ParELC8 should be the best inhibitorof gyrase, which was not experimentally observed. LNIES sequencemay have been the responsible for lower activity of ParELC8 ifcompared to ParELC10 and ParELC3 (Table 1). The differencebetween IC100 values indicates a decrease in the binding with theenzyme. The random conformation of LNIES probably allows a closeproximity of its amino acid side chains to the two highly conservedamino acids (H88 and M91) preventing, particularly in smallerpeptides, the molecular interactions responsible for the inhibitionof the enzyme. CD spectra displayed a typical spectrum of a b-structure (one positive band at 195 nm and one negative band in218 nm) for all selected peptides, with highest percentage for theParELC10. There was no evidence of cleavage complex formation inthe assay conditions.

There are few studies involving E. coli ParE. Therefore, it is notpossible to compare results. The only available study involvinggyrase inhibition assay was carried out in relaxation conditions andnot supercoiling [22]. In this case, the minimum concentration forgyrase inhibition was not determined.

Table 1Inhibitory activities of peptide analogues of E. coli ParE on bacterial topoisomerases.

Peptide IC100 (mmol L�1)a

DNA gyraseb topo IVb

ParELC3 >20c 10ParELC8 35 50ParELC10 20 10ParELC12 50 25

a Concentration of the inhibitor required for complete inhibition of top-oisomerase activity.

b From Escherichia coli.c 93% inhibition at 20 mmol L�1 estimated by ImageJ software.

Fig. 5. Multiple alignment of several C-terminal sequences of bacterial ParE toxins. The high conserved amino acids are depicted in red. The number in the top refers to amino acidspositions in the E. coli ParE sequence. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

L.C.B. Barbosa et al. / European Journal of Medicinal Chemistry 54 (2012) 591e596594

2.2.2. Topoisomerase IVAs bacterial topoisomerases are the most validated targets in

antimicrobial therapies, the development of inhibitors that mightsimultaneously target both DNA gyrase and topo IV, with equalcapacity has the potential to reduce the emergence of target-basedresistance. Thus, we also tested the ability of the peptides to inhibitthe relaxation reaction of topo IV. Gel electrophoresis was used forthis purpose based on the ability of topo IV to relax positive ornegative closed DNA supercoils generating a molecule whichmigrates more slowly compared to supercoiled DNA during gelelectrophoresis [34]. The assay was performed following the sameprocedure described for gyrase: titrating the ParE analogues intoa fixed concentration of topo IV and supercoiled DNA followed byIC100 determination. As with DNA gyrase, an initial screening(Fig. 3B) selected ParELC3, ParELC8, ParELC10 and ParELC12 besidesthe ParELC7 as better topo IV inhibitors (complete inhibition at100 mmol L�1). The IC100 for the selected peptides were determined(Fig. 6) and was found that ParELC3 and ParELC10 were the best,

Fig. 6. IC100 determination of ParELC3, ParELC8, ParELC10 and ParELC12 for DNArelaxation reactions of topo IV (5.0 nmol L�1). Controls were no peptide (lane 0) andtreatment with ciprofloxacin 10 mmol L�1 (lane CFX).

completely inhibiting the relaxation reaction of topo IV with anIC100 value of 10 mmol L�1 (Table 1). The ParELC7 did not showinhibition at less than 75 mmol L�1 of concentration (data notshown). Therefore it was excluded from IC100 assays.

Like gyrase, the core of three-stranded b-sheets of ParE isimportant for interactions with the topo IV. Peptides that do notcontain any one of the three b-sheets also do not inhibit thisenzyme. Again, the amino acid sequence of b4-sheet seems to playa special role in the peptideeenzyme interactions. Compared togyrase the IC100 values were two times lower for the selectedpeptides, except for ParELC8which increased. In this case, the effectof LNIES sequence was more pronounced. Probably the five C-terminal amino acids of E. coli ParE are not involved in the mech-anism of inhibition of topo IV, which can explain the lack of studiesrelated with this enzyme and the ParE protein. These results showthat peptides analogues and probably wild type ParE, interactdifferently with these enzymes.

2.2.3. Human topoisomerase IISequence analysis of type II DNA topoisomerases suggests that

human topoisomerase II, like other eukaryotic type II top-oisomerases, evolved from bacterial topoisomerase II by fusion ofthe two gyrase subunits into a single peptide. The multiple activi-ties including DNA binding, double-strand DNA cleavage andreunion as well as ATP hydrolysis are combined in this singlepeptide [7]. Although relatively diverse at the molecular level,bacterial type II topoisomerases (DNA gyrase and topo IV) andhuman topoisomerase II demonstrate sufficient structural andfunctional homology which is themolecular basis for the conservedmode of interaction between enzyme and DNA, and also serves asa common cellular response mechanism to drug action [35]. Likebacterial topoisomerase II, which is an important therapeutic targetof synthetic and natural antibiotics, human topoisomerase II is alsothe cellular target of many potent anticancer drugs [36]. In vitrobinding studies have revealed a common domain of interaction ofhuman type II topoisomerase with both antineoplastic and anti-microbial agents [37]. Thus, the effect of the peptides in the DNArelaxation catalyzed by human topoisomerase II was evaluated alsousing the gel electrophoresis assay. All peptides, except ParELC12,were unable to inhibit the human topoisomerase II activity atconcentrations 10- to 50-fold higher than those required for gyraseand topo IV (data not shown). ParELC12 inhibited the human top-oisomerase II activity with an IC100 value very close to that forgyrase (50 mmol L�1 e Fig. 7), a contradiction of the fact thatantibacterial drugs target prokaryotic topoisomerases at concen-trations 100- to 1000-fold lower than human enzymes, thepreferred targets of anticancer drugs [38].

Apparently, the peptide activity on human topoisomerase IIdepends on an association of b2 and b3 sheets. Moreover, the b4-sheet inclusion reduces the ability of peptides to inhibit the humantopoisomerase II. In fact, ParELC3, ParELC8 and ParELC10, unable to

Fig. 7. ParELC12-mediated inhibition of DNA relaxation reaction of human top-oisomerase II (4.0 nmol L�1). Controls were no inhibitor (lane 0) and treatment withdoxorubicin (10 mmol L�1), the known anticancer drug (lane DXR). Lane 50: ParELC12 at50 mmol L�1.

L.C.B. Barbosa et al. / European Journal of Medicinal Chemistry 54 (2012) 591e596 595

inhibit the human topoisomerase II, have the b4-sheet in theirstructures. The effect of b4-sheet can be also viewed by comparingthe structureeactivity of ParELC10 and ParELC12. Both share two b-sheets (b2 and b3) and inhibited bacterial topoisomerases but Par-ELC10, which also has the amino acid sequence of the b4-sheet, didnot inhibit human topoisomerase II. Comparatively, CD spectrum ofParELC12 shows a more disordered structure while ParELC10a typical b-structure, which could explain the activity differences.

3. Conclusions

We have designed and synthesized a series of linear peptidesderived from E. coli ParE toxin and tested them for inhibition oftopoisomerases activity. Four peptides were active and completelyinhibited bacterial topoisomerases up to 40 mmol L�1 concentra-tion. The fact that the peptides ParELC3, ParELC8 and ParELC10 sharethe b4-sheet suggests that the amino acid sequence of thissecondary structure is very important for their inhibitory activityon bacterial topoisomerases. It is noteworthy that these peptidesinhibit preferentially the topo IV relaxation activity, evidence thatthe interaction mechanism is likely different for these bacterialtopoisomerases. Due to lack of evidence of cleavage complexformation, the peptides do not act as cleavage complex stabilizingagents, at least in the assay conditions. The removal of b4-sheetfrom ParELC10 rendered a peptide less active on bacterial top-oisomerases but with inhibitory activity on human topoisomeraseII. Therefore, the b4-sheet is not involved in the interactions withhuman topoisomerase II.

Thus, our results suggest a new class of molecules with simul-taneous inhibitory activity in DNA gyrase and topoisomerase IV.Furthermore, we have obtained the first example of a syntheticpeptide from a bacterial toxin with inhibitory activity on humantopoisomerase II. Although more studies are needed includinggenotoxicity and citotoxicity assays, the results of inhibitory activityassays give an idea of the potentiality of the peptides from ParE,especially ParELC10, as a future antibiotic agent. Likewise, ParELC12has great potential as anticancer drug.

4. Experimental section

4.1. Materials

9-fluorenylmethoxycarbonyl (Fmoc)-amino acids, Rink AmideeMBHA resin (substituted at 0.68 mmol g�1), and Fmoc-Ser-Wang

resin (substituted at 0.64 mmol g�1) were purchased from Nova-biochem (San Diego, CA, USA). Chemicals for the peptide synthesiswere obtained from SigmaeAldrich Co. (St. Louis, MO, USA). Aceto-nitrile for HPLC was the product of J.T.Baker (Phillipsburg, NJ, USA).DNA gyrase, topo IV and human topoisomerase II were supplied asreaction kits by Inspiralis Limited (Norwich, UK). Ciprofloxacin (CFX)and doxorubicin (DXR) were obtained as the hydrochlorides fromSigmaeAldrich Co. (St. Louis, MO, USA).

4.2. Synthesis, purification and identification of the peptides

The peptides whose sequences are reported in Fig. 2 weresynthesized manually according to solid-phase synthesis method-ology using Fmoc chemistry with Rink AmideeMBHA resin andDIC/HOBt activation [39]. The functional side chains of Fmoc-aminoacids were protected by the following groups: But for Asp, Glu, Ser,and Tyr, Trt for His, Asn and Cys, and Pmc for Arg. Acetylation wasperformed with acetic anhydride and DIEA (10 eq. each). Depro-tection and cleavage were achieved by treatment with TFA/water/phenol/thioanisole/1.2-ethanedithiol (82.5:5:5:5:2.5) for 2 h atroom temperature. Then peptides were precipitated in ether,centrifuged, solubilized and dried under a vacuum. Peptides werepurified by semi-preparative Reverse-Phase High Pressure LiquidChromatography (RP-HPLC) on a Waters system using a Vydac-C18column (25� 2.5 cm; 10 mmparticles; 300 Å porosity) with a lineargradient of 30e70% of solvent B (A: water, 0.1% TFA; B: acetonitrile(MeCN) 75% in water, 0.1% TFA) over 90 min. The flow rate was10 mL min�1 and detection was carried out at 220 nm. AnalyticalRP-HPLC was carried out on a Varian ProStar apparatus employinga Nucleosil C18 reverse-phase column (25 � 0.46 cm; 5 mm parti-cles; 300 Å porosity) with a 30e70% linear gradient of solvent B (A:water, 0.04% TFA; B: MeCN, 0.04% TFA) over 30 min, flow rate1.0 mL min�1 and UV detection at 220 nm. Peptide purity wasestimated to be above 98% by amino acids analysis (Hydrolysis:6 mol L�1 aqueous HCl solution with 80 mL of 5% aqueous phenol at110 �C for 72 h) on a Shimadzu LC-10A/C-47A analyzer. The identityof the peptides was confirmed by positive electrospray massspectrometry (ESI-MS) on an LC-ESI-MS Waters model apparatusfrom Micromass Waters.

The secondary structure of the pure peptides was estimated byCircular Dichroism spectroscopy. CDmeasurements in 10 mmol L�1

Tris.HCl buffer (pH 7.5) were carried out on a JASCO J-715 CDspectrophotometer (Japan) on nitrogen flush using a quartz cell of1 mm path-length. Spectra were recorded at 25 � 1 �C from 190 to240 nm at 0.5 to 0.2 nm resolutions with a scan rate of 20 nm/minand a sensitivity of 50e100 mdeg, respectively.

4.3. Gel electrophoresis assays

4.3.1. Gyrase inhibition assayThe standard reaction (30 mL) contained relaxed pBR322 DNA

(500 ng), DNA gyrase (3.4 nmol L�1) and indicated amounts of eachof a peptides. The reactions were carried out in the assay buffercontaining 30 mmol L�1 Tris.HCl, pH 7.5, 24 mmol L�1 KCl,4 mmol L�1 MgCl2, 2 mmol L�1 dithiothreitol (DTT), 1 mmol L�1

ATP, 1.8 mmol L�1 spermidine, 6.5% glycerol and 0.1 mg mL�1

albumin, incubated at 37 �C for 1 h. The reaction was stopped byadding 15 mL of STEB (20% sucrose; 0.05 mol L�1 Tris.HCl, pH 7.5,0.05 mol L�1 EDTA, 50 mg/mL bromophenol blue) and 60 mL ofchloroform: isoamyl alcohol (24:1 v/v) mixture. The mixture wascentrifuged and the supernatant was analyzed on a 1% agarose gelin TBE buffer (89 mmol L�1 Tris.HCl, 89 mmol L�1 boric acid,2 mmol L�1 EDTA, pH 8.2). The samples were electrophoresed at20 V for 6 h, and the gel was stained with ethidium bromidesolution (1 mg/mL) and analyzed by an Alpha Imager EP System of

L.C.B. Barbosa et al. / European Journal of Medicinal Chemistry 54 (2012) 591e596596

Alpha Innotech. The ImageJ software was used to process the gelimages and quantify the bands.

4.3.2. Topoisomerase IV inhibition assayThe standard reaction (30 mL) contained supercoiled pBR322

DNA (400 ng), topo IV (5.0 nmol L�1) and indicated amounts of eachof a peptides. The reactions were carried out in the assay buffercontaining 40 mmol L�1 HEPES/KOH, pH 7.6, 100 mmol L�1 KCl,10 mmol L�1 MgCl2, 10 mmol L�1 dithiothreitol (DTT), 1 mmol L�1

ATP, and 50 mg mL�1 albumin, incubated at 37 �C for 1 h. Thereaction was stopped by adding 15 mL of STEB and 60 mL of chlo-roform: isoamyl alcohol (24:1 v/v) mixture, centrifuged andanalyzed as described for the gyrase inhibition assay.

4.3.3. Human topoisomerase II inhibition assayThe reaction (30 mL) contained supercoiled pBR322 DNA

(500 ng), human topoisomerase II (4.0 nmol L�1) and 50 mmol L�1 ofpeptides. The reactions were carried out in the assay buffer con-taining 50 mmol L�1 Tris.HCl, pH 7.5, 125 mmol L�1 NaCl,10 mmol L�1 MgCl2, 5 mmol L�1 dithiothreitol (DTT), and100 mg mL�1 albumin, incubated at 37 �C for 1 h. The reaction wasstopped by adding 15 mL of STEB and 60 mL of chloroform: isoamylalcohol (24:1 v/v) mixture, centrifuged and analyzed as describedfor the gyrase inhibition assay.

4.4. Computational methods

All protein sequences were obtained into NCBI sequence data-base (http://www.ncbi.nlm.nih.gov). Sequence alignment wasperformed using CLUSTALW program (http://www.ebi.ac.uk/Tools/clustalw2/index.html). Additional analysis and manipulations ofthe sequences were performed by EBiAn package [40]. For pro-cessing and gel images analyses were used the public IMAGEJprogram (http://rsbweb.nih.gov/ij/download.html).

Acknowledgments

We gratefully acknowledge FAPESP, CNPq and CAPES for finan-cial support. R. Marchetto is recipient of research fellowship fromCNPq.

References

[1] K.D. Corbett, J.M. Berger, Annu. Rev. Biophys. Biomol. Struct. 33 (2004)95e118.

[2] A.J. Schoeffler, J.M. Berger, Biochem. Soc. Trans. 33 (2005) 1465e1470.[3] Y. Pommier, E. Leo, H. Zhang, C. Marchand, Chem. Biol. 17 (2010) 421e433.[4] A.D. Bates, A. Maxwell, Biochemistry 46 (2007) 7929e7941.[5] E.L. Zechiedrich, A.B. Khodursky, N.R. Cozzarelli, Genes Dev. 11 (1997)

2580e2592.[6] R.W. Deibler, S. Rahmati, E.L. Zechiedrich, Genes Dev. 15 (2001) 748e761.[7] J.J. Champoux, Annu. Rev. Biochem. 70 (2001) 369e413.[8] R.J. Reece, A. Maxwell, Crit. Rev. Biochem. Mol. Biol. 26 (1991) 335e375.[9] H. Peng, K.J. Marians, J. Biol. Chem. 268 (1993) 24481e24490.

[10] D.C. Hooper, Clin. Infect. Dis. 27 (Suppl. 1) (1998) S54eS61.[11] R. Marchetto, E. Nicolás, N. Castilho, J. Bacardit, M. Navia, J. Vila, E. Giralt,

J. Pept. Sci. 7 (2001) 27e40.[12] C. Sissi, M. Palumbo, Nucleic Acids Res. 37 (2009) 702e711.[13] F. Collin, S. Karkare, A. Maxwell, Appl. Microbiol. Biotechnol. 92 (2011)

479e497.[14] C.D. Hardy, N.R. Cozzarelli, Antimicrob. Agents Chemother. 47 (2003)

941e947.[15] A.C. Scatigno, S.S. Garrido, R. Marchetto, J. Pept. Sci. 10 (2004) 566e577.[16] S.S. Garrido, A.C. Scatigno, E. .Trovatti, D.C. Carvalho, R. Marchetto, J. Pept. Res.

65 (2005) 502e511.[17] S. Alt, L.A. Mitchenall, A. Maxwell, L. Heide, J. Antimicrob. Chemother. 66

(2011) 2061e2069.[18] E. Goetschi, P. Angehrn, H. Gmuender, P. Hebeisen, H. Link, R. Masciadri,

J. Nielsen, Pharmacol. Ther. 60 (1993) 367e380.[19] S.M. Hashimi, M.K. Wall, A.B. Smith, A. Maxwell, R.G. Birch, Antimicrob.

Agents Chemother. 51 (2007) 181e187.[20] J.G. Heddle, S.J. Blance, D.B. Zamble, F. Hollfelder, D.A. Miller, L.M. Wentzell,

C.T. Walsh, A. Maxwell, J. Mol. Biol. 307 (2001) 1223e1234.[21] M. Couturier, E.M. Bahassi, L. Van Melderen, Trends Microbiol. 6 (1998)

269e275.[22] Y. Jiang, J. Pogliano, D.J. Helinski, I. Konieczny, Mol. Microbiol. 44 (2002)

971e979.[23] F. Hayes, Science 301 (2003) 1496e1499.[24] E.P. Johnson, A.R. Strom, D.R. Helinski, J. Bacteriol. 178 (1996) 1420e1429.[25] V. Anantharaman, L. Aravind, Genome Biol. 4 (2003) R81.1eR81.15.[26] D. Francuski, W. Saenger, J. Mol. Biol. 393 (2009) 898e908.[27] K.M. Dalton, S. Crosson, Biochemistry 49 (2010) 2205e2215.[28] L.C.B. Barbosa, S.S. Garrido, A. Garcia, D.B. Delfino, R. Marchetto, Bio-

information 4 (2010) 438e440.[29] R.C. Roberts, D.R. Helinski, J. Bacteriol. 174 (1992) 8119e8132.[30] A. Fiebig, C.M.C. Rojas, D. Siegal-Gaskins, S. Crosson, Mol. Microbiol. 77 (2010)

236e251.[31] E. Trovatti, C.A. Cotrim, S.S. Garrido, R.S. Barros, R. Marchetto, Bioorg. Med.

Chem. Lett. 18 (2008) 6161e6164.[32] R. Marchetto, E.M. Cilli, G.N. Jubilut, S. Schreier, C.R. Nakaie, J. Org. Chem. 12

(2005) 4561e4568.[33] C. Ullsperger, N.R. Cozzarelli, J. Biol. Chem. 271 (1996) 31549e31555.[34] N.J. Crisona, T.R. Strick, D. Bensimon, V. Croquette, N.R. Cozzarelli, Genes Dev.

14 (2000) 2881e2892.[35] S. Albertini, A.A. Chetelat, B. Miller, W. Muster, E. Pujadas, R. Strobel, E. Gocke,

Mutagenesis 10 (1995) 343e351.[36] Z. Topcu, J. Clin. Pharm. Ther. 26 (2001) 405e416.[37] S.H. Elsea, M. Westergaard, D.A. Burden, J.P. Lomenick, N. Osheroff,

Biochemistry 36 (1997) 2919e2924.[38] P. Heisig, Mutagenesis 24 (2009) 465e469.[39] G.B. Fields, R.L. Noble, Int. J. Pept. Protein Res. 35 (1990) 161e214.[40] L.C.B. Barbosa, S.S. Garrido, A. Garcia, D.B. Delfino, R.D. Gonçalves,

R. Marchetto, Bioinformation 5 (2010) 46e48.