International Journal of Antimicrobial Agents 28 (2006) 19–24

Antibacterial evaluation of a collection of norfloxacin andciprofloxacin derivatives against multiresistant bacteria

J. Vila a,∗, J. Sanchez-Cespedes a, J.M. Sierra a, M. Piqueras b,E. Nicolas b, J. Freixas c, E. Giralt b

a Servei de Microbiologia, Centre de Diagnostic Biomedic, Hospital Clinic, Villarroel 170, 08036 Barcelona, Spainb Departament de Quımica Organica, Facultat de Quımiques, Universitat de Barcelona, Barcelona, Spain

c Cenavisa, S.A., Reus, Tarragona, Spain

Received 5 December 2005; accepted 3 February 2006

Abstract

The objective of this study was to analyse an array of ciprofloxacin and norfloxacin derivatives in order to determine those with good activityagainst bacteria that already present fluoroquinolone resistance associated with mutations in the gyrA and/or parC genes. Four norfloxacinabmlsd©

K

1

tbtitioeflosg

0d

nd 20 ciprofloxacin derivatives were synthesised and tested against quinolone-susceptible and -resistant Escherichia coli, Acinetobacteraumannii, Stenotrophomonas maltophilia and Staphylococcus aureus strains using a microdilution test. Among the derivatives, the 4-ethyl-7-piperazine ciprofloxacin derivative showed a minimum inhibitory concentration for 50% of the organisms that was 16- and 8-fold

ower than ciprofloxacin for A. baumannii and S. maltophilia, respectively. When the methyl group at position 4 in the piperazine ring wasubstituted by ethyl, butyl or heptyl groups, activity against A. baumannii steadily decreased. The 7-(4-methyl)-piperazine ciprofloxacinerivative (UB-8902) showed very good activity against these multiresistant microorganisms including A. baumannii and S. maltophilia.

2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

eywords: Multiresistant bacteria; Acinetobacter baumannii; Stenotrophomonas maltophilia; Norfloxacin derivatives; Ciprofloxacin derivatives

. Introduction

Acinetobacter baumannii and Stenotrophomonas mal-ophilia may be considered the paradigm of multiresistantacteria [1–5]. To date, A. baumannii has been reportedo be resistant to all commercialised antimicrobial agents,ncluding colistin [1]. Likewise, S. maltophilia resistanto all antibiotics except trimethoprim/sulphamethoxazoles frequently isolated [1,6,7]. Among the different classesf antimicrobial agents, the fluoroquinolones are the mostxtensively used. Since the development of norfloxacin,uoroquinolones have gained prominence in the therapyf bacterial infections owing to their broad antibacterialpectrum and excellent bioavailability. However, the emer-ence of quinolone resistance has been steadily rising. This

∗ Corresponding author. Tel.: +34 93 227 5522; fax: +34 93 227 9372.E-mail address: jvila@ub.edu (J. Vila).

emergence of resistance has been associated with the useof quinolones both in humans and animals [8–10]. Themechanisms of resistance to quinolones are associated with[10]: (i) mutations mainly in the gyrA and parC genes,encoding the A subunits of DNA gyrase and topoisomeraseIV (quinolone protein targets), respectively. However, insome microorganisms mutations in the gyrB and parE genes,encoding the B subunits of DNA gyrase and topoisomeraseIV, respectively, may also play a role in quinolone resistance;(ii) a decrease in drug accumulation related to a decreasein quinolone permeability or to increased efflux of thequinolone from the cell owing to overexpression of someefflux pumps; and (iii) the presence of the qnr gene thatencodes a DNA gyrase protection mechanism [11].

The development of new antibiotics can be achievedfrom derivatives of known antimicrobial agents or byidentification of novel agents active against previouslyunexploited targets [12]. The former procedure is mainly

924-8579/$ – see front matter © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.oi:10.1016/j.ijantimicag.2006.02.013

20 J. Vila et al. / International Journal of Antimicrobial Agents 28 (2006) 19–24

based on the knowledge of the mechanisms of resistanceand the derivatives are designed to avoid these mechanisms.The main objective of this study was to analyse an arrayof ciprofloxacin and norfloxacin derivatives in order todetermine those with good activity against bacteria thatalready present fluoroquinolone resistance associated withmutations in the gyrA and/or parC genes.

2. Methods

Three clinical isolates of Escherichia coli, A. bauman-nii and Staphylococcus aureus with different minimuminhibitory concentrations (MICs) of ciprofloxacin, from sus-ceptible to resistant, were used in the first screening. None ofthe isolates were epidemiologically related [13]. The geno-

typic characteristics of the isolates are listed in Table 1. Inaddition, 30 A. baumannii and 31 S. maltophilia clinical iso-lates were used to test the compound with the best activity.Ciprofloxacin, norfloxacin and their derivatives were syn-thesised as described previously [14,15] and were providedby Cenavisa, S.A. Laboratories (Reus, Spain). Antimicro-bial susceptibility was determined by the broth microdilutionmethod according to National Committee for Clinical Labo-ratory Standards guidelines [16].

3. Results

Ciprofloxacin and norfloxacin as well as four norfloxacinderivatives and ten ciprofloxacin derivatives (Fig. 1) weretested against three microorganisms representative of

Fig. 1. Norfloxacin and ciproflo

xacin derivatives 16–41.

J. Vila et al. / International Journal of Antimicrobial Agents 28 (2006) 19–24 21

Table 1Genomic description of the strains

Strain gyrA parCa

Ser-83b Asp-87 Ser-80 Glu-84

Escherichia coliC-20 Ser Asp Ser GluC-4 Leu Asp Ser Glu1273 Leu Tyr Ser Lys

Acinetobacter baumannii58 Ser Asp Ser Glu661 Leu Asp Ser Glu31 Leu Asp Ser Lys

Staphylococcus aureus5-61 Ser Asp Ser Glu5-96 Ser Asp Phe Glu4-10 Leu Asp Phe Glu

a grlA in S. aureus.b Ser-80 in S. aureus.

Enterobacteriaceae (E. coli), Gram-negative non-fermentative bacilli (A. baumannii) and Gram-positivecocci (S. aureus). The structure of ciprofloxacin (16) andnorfloxacin (29), two closely related fluoroquinolones,can be described as derived from a bicyclic 1-alkyl-6-fluoro-4-oxo-1,4-dihydro-quinoline-3-carboxylic acid witha piperazine ring at position 7. Ciprofloxacin and norfloxacindiffer in the nature of the 1-alkyl group (R1 = cC3H5and Et, respectively) (see Fig. 1). We have designed andsynthesised a collection of four norfloxacin derivatives andten ciprofloxacin derivatives that share the 1-alkyl-6-fluoro-4-oxo-1,4-dihydro-quinoline-3-carboxylic acid backbonebut differ in the nature of the heterocyclic substituentat position 7. The rationale of this design is to try tomodulate the specificity of ciprofloxacin and norfloxacin

against resistant microbial strains through modulation ofthe chemical nature of this substituent. In derivatives 17,19, 21, 22, 24, 25, 26, 31 and 32, the piperazine ringwas replaced by a different heterocycle. The remainingcompounds incorporate more conservative changes, i.e.N-monosubstituted or N,C-disubstituted piperazines insteadof the original unsubstituted piperazine ring.

The genetic features of the strains used in this study arelisted in Table 1. The three strains each of E. coli, A. bau-mannii and S. aureus used were: (1) wild-type E. coli (strainC-20), A. baumannii (strain 58) and S. aureus (strain 5-61)with no mutations in the gyrA or parC genes and with MICs ofciprofloxacin of 0.06, 0.125 and 0.125 �g/mL, respectively;(2) a strain with a mutation in the gyrA gene in E. coli (strainC-4) and A. baumannii (strain 661) and in the grlA gene of S.aureus (strain 5-96), which generated MICs of ciprofloxacinof 0.5, 8 and 1 �g/mL, respectively; and (3) an E. coli strainwith two mutations in the gyrA gene and one mutation in theparC gene (MIC of ciprofloxacin, 16 �g/mL) (strain 1273),an A. baumannii strain with a mutation in the gyrA gene andanother in the parC gene (MIC of ciprofloxacin, 32 �g/mL)(strain 31), and a S. aureus strain carrying one mutation inthe gyrA gene and one mutation in the grlA gene, producinga MIC of ciprofloxacin of 8 �g/mL (strain 4-10).

Among the 14 derivatives studied initially, 10 derived fromciprofloxacin and 4 from norfloxacin, compound 38 with twomammE

sT

Table 2Minimum inhibitory concentrations (MICs; in �g/mL) of ciprofloxacin, norfloxacinand Staphylococcus aureusa

Compound E. coli A. bauman

C-20 C-4 1273 58

Ciprofloxacin (16) 0.06 0.5 16 0.12517 0.06 4 >128 218 2 >128 >128 1619 2 >16 >128 162 82 22 62 82 82 42 0.5N 0.53 63 0.253 0.253 0.06

0 2 32 >1281 0.12 4 >1282 0.25 8 >128 13 0.06 2 644 0.25 16 >128 >125 0.06 4 >1286 0.06 4 128orfloxacin (29) 0.06 1 320 8 32 32 11 0.25 16 >647 0.125 32 >648 0.06 0.5 32 <a See Table 1 for genomic description of the strains.

ethyl groups, at R3 and R6 of the piperazine substituentt position 7 of the 1,4-dihydroquinoline nucleus, showeduch better activity than ciprofloxacin against wild-type andutated strains of A. baumannii and S. aureus but not against. coli (Table 2).

A series of analogues of compound 38 were furtherynthesised to improve the activity of this derivative (Fig. 1;able 3) and were tested against A. baumannii. Compound

and some of their derivatives for Escherichia coli, Acinetobacter baumannii

nii S. aureus

661 31 5-61 5-96 4-10

8 32 0.125 1 84 16 0.03 0.12 8

16 16 0.5 1 816 32 0.06 0.12 816 16 0.12 0.5 8

2 32 0.25 0.5 6432 64 0.06 0.25 32

8 32 0.12 0.5 16>128 >128 0.03 0.25 8

4 32 0.016 0.12 864 128 0.25 0.25 3232 >64 0.5 16 3232 32 8 4 1632 >64 <0.06 0.25 >64

>64 >64 0.125 5 82 32 <0.03 0.5 8

22 J. Vila et al. / International Journal of Antimicrobial Agents 28 (2006) 19–24

Table 3Activity of compound 38 and some of its derivatives against Acinetobacterbaumanniia

Compound MIC (�g/mL)

58 661 31

Ciprofloxacin 0.125 8 3238 <0.06 2 817 2 4 1627 32 32 3228 0.125 4 833 (UB-8902) <0.06 0.5 434 0.06 0.5 1635 1 2 6436 8 8 6439 32 32 3240 0.06 2 1641 0.125 16 16

MIC, minimum inhibitory concentration.a See Table 1 for genomic description of the strains.

33, in which the methyl group at R3 from compound 38was removed, was of interest, with increased activity againstciprofloxacin-resistant A. baumannii strains being seen. TheMIC of ciprofloxacin for the A. baumannii strain carryinga substitution in GyrA (Ser83 → Leu) (strain 661) was8 �g/mL, whereas the MIC of compound 33 was 0.5 �g/mL(Table 3). Meanwhile, the MICs of ciprofloxacin andcompound 33 for the strain with a substitution in GyrA(Ser → Leu) and ParC (Glu-84 → Lys) (strain 31) were32 �g/mL and 4 �g/mL, respectively. When the methyl groupat R6 of compound 33 was substituted by an ethyl (compound34, enrofloxacin), butyl (compound 35) or heptyl (com-pound 36) group, the activity against A. baumannii steadilydecreased (Table 3). On determining the susceptibility of acollection of epidemiologicaly unrelated clinical isolates ofA. baumannii and S. maltophilia to compound 33 (so-calledUB-8902) compared with ciprofloxacin (Table 4), an MIC50of UB-8902 that was 16-fold lower than ciprofloxacin for A.baumannii and 8-fold lower for S. maltophilia was observed.MICs of wild-type and mutated strains of A. baumanniiand S. maltophilia to ciprofloxacin and UB-8902 are listedin Tables 5 and 6. Overall, the activity of compound 33compared with ciprofloxacin increased between 2- and8-fold for wild-type A. baumannii strains, between 8- and64-fold for A. baumannii strains with one mutation in thegyrA gene and between 8- and >32-fold for strains with onem

Table 5In vitro activity of ciprofloxacin (16) and the new fluoroquinolone UB-8902(33) against Acinetobacter baumannii

QRDR mutationa/A. baumannii strain 16 33

Wild-type77 2 1201 0.5 0.0688 64 893 <0.06 <0.0665580 0.5 <0.0658 0.12 <0.06

Ser-83 → Leu46I 64 2333028B >64 4Y-29 32 0.533167B 64 4661 8 0.574I 1 <0.06875 64 834 32 433 8 0.5167R 4 0.25175 16 2

Ser-83 → Leu; Ser-80 → Leu113321 >64 4105346 >64 8F-14 >64 4E-24 >64 484I 64 813714B >64 467384RC >64 45985V 64 4215I >64 8709R >64 4A5-22 64 2A5-45 32 4

Ser-83 → Leu; Glu-84 → Lys31 >64 4

a Mutations in the quinolone resistance-determining region: gyrA, Ser-83;parC, Ser-80, Glu-84.

The activity of compound 33 compared with ciprofloxacinfor most S. maltophilia strains increased from 4- to 533-fold.

4. Discussion

Hospital-acquired multiresistant microorganisms such asA. baumannii and S. maltophilia are of great concern.Acinetobacter baumannii clinical isolates resistant to allantimicrobial agents, even colistin, have been described

TI (33) against Acinetobacter baumannii and Stenotrophomonas maltophilia

33 (UB-8902)

MIC90 MIC range MIC50 MIC90

A >64 <0.06 to 8 4 8S 64 <0.06 to 32 0.5 8

M

utation in the gyrA gene and one mutation in the parC gene.

able 4n vitro activity of ciprofloxacin (16) and the new fluoroquinolone UB-8902

Ciprofloxacin

MIC range MIC50

. baumannii (n = 26) <0.06 to >64 64. maltophilia (n = 11) 2 to >64 4

IC, minimum inhibitory concentration (�g/mL).

J. Vila et al. / International Journal of Antimicrobial Agents 28 (2006) 19–24 23

Table 6In vitro activity of ciprofloxacin (16) and the new fluoroquinolone UB-8902(33) against Stenotrophomonas maltophilia

S. maltophilia 16 33

4834R 2 0.25N26-17 >64 8N30-41 16 42-19 2 0.12J19-15 2 0.5239V 4 0.526091V 2 0.25D2-10 32 415519H 64 0.12L9-5Rs 4 0.12L10-26Rs 64 0.125506R 8 0.5D1-22Rs 4 0.5J12-26 8 0.5L13-34 >64 32O29-23 16 0.12L9-13 64 0.5O27-34 8 12-20 2 23554F 2 1D7-14 8 211424-O 64 8N29-7Rs 2 2D6-7Rs 4 0.25N15-10CG 2 122122 >64 32D10-8Rs 8 1N26-34 2 0.12N8-40 2 0.51-26 4 0.5N13-8Rs 2 0.25

[1,5]. Moreover, the typical resistance phenotype of S.maltophilia is to all antimicrobial agents except trimetho-prim/sulphamethoxazole. Therefore, therapeutic alternativesfor the treatment of nosocomial infections caused by thesemicroorganisms are needed. One of the approaches to developa more active antimicrobial agent is to know the mech-anism(s) of resistance and to develop derivatives of thisantimicrobial agent that circumvent this mechanism of resis-tance. The acquisition of quinolone resistance is mainly dueto chromosomal mutations in genes (gyrA and parC) encod-ing the A subunits of the protein targets (DNA gyrase andtopoisomerase IV) as well as mutations causing reduceddrug accumulation, either by decreased uptake or increasedefflux [10]. The primary quinolone resistance determinantin Gram-negative microorganisms is located in a regioncalled the quinolone resistance-determining region (QRDR)in GyrA. Amino acid substitutions have been described indifferent positions in this region, with changes at position83 being the most frequent [10]. Although the interplaybetween low permeability and constitutive expression ofsome efflux pumps [10] can produce a low level of intrinsicresistance in A. baumannii, the main mechanisms of resis-tance to quinolones are mutations in topoisomerase genes(gyrA and/or parC) [17,18]. Meanwhile, in S. maltophilia,

quinolone-susceptible and -resistant strains presented identi-cal amino acid sequences in GyrA and ParC [19,20], suggest-ing that in this microorganism low permeability, overexpres-sion of efflux pumps or the interplay between both effectsmay play an important role in the acquisition of quinoloneresistance [21]. However, the amino acid of GyrA found inthe position equivalent to Ser-83 of E. coli was Gln insteadof Ser or Thr. Although this should be investigated, thisunusual amino acid may generate an intrinsic resistance toquinolones in S. maltophilia. In this study, a number of 7-substituted norfloxacin and ciprofloxacin derivatives weresynthesised and evaluated for their antibacterial activities. Wechose this position to generate ciprofloxacin and norfloxacinderivatives because it has been implicated in the interactionbetween quinolones and DNA gyrase, especially with theQRDR where the substitutions related to quinolone resis-tance have been located [10,22–24]. Therefore, we thoughtthat changes in this position would improve the activity. Wefound that the 7-(4-methyl)-piperazine ciprofloxacin deriva-tive UB-8902 showed very good activity against multiresis-tant microorganisms such as A. baumannii and S. maltophilia.

Acknowledgments

(CDmgaI

R

This work has been supported in part by Grants PETRIPETRI1995-0430-OP) from Comision Interministerial deiencia y Tecnologıa, Spain, and 2005 SGR00444 from theepartment d’Universitats, Recerca I Societat de la Infor-acio de la Generalitat de Catalunya, Spain (to J.V.), and

rant MCYT-FEDER (BIO-2002-2301) (to E.G.). J.S.-C. hasfellowship from Red Espanola de Investigacion en Patologıa

nfecciosa (REIPI) C14.

eferences

[1] Arroyo LA, Garcia-Curiel A, Pachon-Ibanez ME, et al. Reliability ofthe E-test method for detection of colistin resistance in clinical iso-lates of Acinetobacter baumannii. J Clin Microbiol 2005;43:903–5.

[2] Ribera A, Ruiz J, Jimenez de Anta MT, Vila J. Effect of an effluxpump inhibitor on the MIC of nalidixic acid for Acinetobacter bau-mannii and Stenotrophomonas maltophilia clinical isolates. J Antimi-crob Chemother 2002;49:697–8.

[3] Rodriguez-Hernandez MJ, Jimenez-Mejias ME, Pichardo C, CuberosL, Garcia-Curiel A, Pachon J. Colistin efficacy in an experimen-tal model of Acinetobacter baumannii endocarditis. Clin MicrobiolInfect 2004;10:581–4.

[4] Vila J. Mechanisms of antimicrobial resistance in Acinetobacter bau-mannii. Rev Med Microbiol 1998;9:87–97.

[5] Yoon J, Urban C, Terzian C, Mariano N, Rahal JJ. In vitro dou-ble and triple synergistic activities of polymyxin B, imipenem,and rifampin against multidrug-resistant Acinetobacter baumannii.Antimicrob Agents Chemother 2004;48:753–7.

[6] Caylan R, Kaklikkaya N, Aydin K, et al. An epidemiological analysisof Stenotrophomonas maltophilia strains in a university hospital. JpnJ Infect Dis 2004;57:37–40.

[7] Fadda G, Spanu T, Ardito F, et al.; Italian Epidemiological Obser-vatory. Antimicrobial resistance among non-fermentative Gram-

24 J. Vila et al. / International Journal of Antimicrobial Agents 28 (2006) 19–24

negative bacilli isolated from the respiratory tracts of Italian inpa-tients: a 3-year surveillance study by the Italian EpidemiologicalSurvey. Int J Antimicrob Agents 2004;23:254–61.

[8] Bearden DT, Danziger LH. Mechanism of action of and resistanceto quinolones. Pharmacotherapy 2001;21:224S–32S.

[9] Hooper DC. Emerging mechanisms of fluoroquinolone resistance.Emerg Infect Dis 2001;7:337–41.

[10] Vila J. Fluoroquinolone resistance. In: Whitte DG, Alekshun MN,McDermott PF, editors. Frontiers in antimicrobial resistance: a tributeto Stuart B. Levy. Washington, DC: ASM Press; 2005. p. 41–52.

[11] Martınez-Martınez L, Pascual A, Jacoby GA. Quinolone resistancefrom a transferable plasmid. Lancet 1998;351:797–9.

[12] Vila J, Sanchez-Cespedes J, Giralt E. Old and new strategies forthe discovery of antibacterial agents. Curr Med Chemist Anti-InfectAgents 2005;4:337–53.

[13] Vila J, Ruiz J, Navia M, Becerril B, et al. Spread of amikacinresistance in Acinetobacter baumannii strains isolated in Spain dueto an epidemic strain. J Clin Microbiol 1999;37:758–61.

[14] Escribano E, Calpena AC, Garrigues TM, Freixas J, DomenechJ, Moreno J. Structure–absorption relationships of a series of6-fluoroquinolones. Antimicrob Agents Chemother 1997;41:1996–2000.

[15] Merino V, Freixas J, Bermejo MV, Garrigues TM, Moreno J, Pla-Delfina JM. Biophysical models as an approach to study passiveabsorption in drug development: 6-fluoroquinolones. J Pharm Sci1995;84:777–82.

[16] National Committee for Clinical Laboratory Standards. Performancestandard for antimicrobial susceptibility testing. Fifteenth Informa-tional Supplement. M100-S15. Wayne, PA: NCCLS; 2005.

[17] Vila J, Ruiz J, Goni P, Marcos A, Jimenez de Anta MT. Mutation inthe gyrA gene of quinolone-resistant clinical isolates of Acinetobacterbaumannii. Antimicrob Agents Chemother 1995;39:1201–3.

[18] Vila J, Ruiz J, Goni P, Jimenez de Anta MT. Quinolone resistancein the topoisomerase IV parC gene of Acinetobacter baumannii. JAntimicrob Agents 1997;39:757–62.

[19] Ribera A, Domenech-Sanchez A, Ruiz J, Benedı J, Jimenez de AntaMT, Vila J. Mutations in gyrA and parC QRDRs are not rele-vant for quinolone resistance in epidemiological unrelated Stenotro-phomonas maltophilia clinical isolates. Microb Drug Resist 2002;8:245–52.

[20] Valdezate S, Vindel A, Echeita A, et al. Topoisomerase II andIV quinolone-resistance determining regions in Stenotrophomonasmaltophilia in clinical isolates with different levels of quinolonesusceptibility. Antimicrob Agents Chemother 2002;46:665–71.

[21] Valdezate S, Vindel A, Saez-Nieto JA, Baquero F, Canton R.Preservation of topoisomerase genetic sequences during in vivoand in vitro development of high-level resistance to ciprofloxacinin isogenic Stenotrophomonas maltophilia strains. J AntimicrobChemother 2005;56:220–3.

[22] Shen LL, Mitscher LA, Sharma PN, et al. Mechanism of inhibitionof DNA gyrase by quinolone antibacterials: a cooperative drug–DNAbinding model. Biochemistry 1989;28:3886–94.

[23] Palumbo M, Gatto B, Zagotto G, Palu G. On the mechanism ofaction of quinolone drugs. Trends Microbiol 1993;1:232–4.

[24] Vila J, Ruiz J, Marco F, et al. Association between double muta-tion in gyrA gene of ciprofloxacin-resistant clinical isolates ofEscherichia coli and MICs. Antimicrob Agents Chemother 1994;38:2477–9.