Iddock Uptake en Saureus ME CITA

Journal of Antimicrobial Chemotherapy (2001) 47, 261–270

Introduction

For an antibacterial agent to be effective, it must penetratethe bacterial cell to reach its target. The primary target offluoroquinolones is DNA gyrase, an enzyme essential forDNA supercoiling, which is encoded by the gyrA and gyrBgenes. Topoisomerase IV (encoded by the grlA/parC andgrlB/parE genes) is also involved in quinolone action, andmay be the dominant target in Staphylococcus aureus.1 Thesusceptibility of an organism to a particular quinolone isdetermined by both the ability of the agent to penetrate thecell envelope and the inhibitory activity of the agent againstits target.2

The factors affecting the accumulation of quinolones byEscherichia coli and other Gram-negative bacteria havebeen studied, but less is known about accumulation by

S. aureus. The physical properties of quinolones (e.g. rela-tive hydrophobicity, charge or molecular mass) are import-ant for penetration into the Gram-negative bacterial cell.Increasing molecular mass, hydrophobicity and negativecharge hinder penetration of antibiotics into Gram-negative organisms through the porin channels, althoughhydrophobic molecules appear to enter via the lipopoly-saccharide or across the lipid bilayer.3,4 Gram-positive bac-teria do not possess an outer membrane, and so lack outermembrane proteins and lipopolysaccharide. Accumulationby S. aureus is thought to take place by simple diffusionacross the cytoplasmic membrane5,6 but an active effluxmechanism has been identified in staphylococci whichreduces accumulation. The norA locus codes for the pro-tein NorA, which appears to straddle the cell membraneand act as an active efflux pump of hydrophilic quinolones.7

261

Original articles

Effect of hydrophobicity and molecular mass on the accumulation of fluoroquinolones by Staphylococcus aureus

Laura J. V. Piddock*, Yu Fang Jin and Deborah J. Griggs

Antimicrobial Agents Research Group, Division of Immunity and Infection, The Medical School, University of Birmingham, Birmingham B15 2TT, UK

Ten novel fluoroquinolones, with similar chemical structures but differing antibacterial activities and hydrophobicities, were studied to evaluate the role of the physical properties ofquinolones on their accumulation and antibacterial activity for Staphylococcus aureus. Six ofthe 10 agents and tosufloxacin were more active against quinolone-susceptible and -resistantS. aureus than the remaining four agents and several piperazinyl fluoroquinolones. Changes tothe side chains of the pyrollidinyl substituent at the R7 position alone made little difference to the MICs. Comparison of MICs of agents that were structurally identical apart from the R1substituents, confirmed that a t-butyl group confers enhanced activity against S. aureus over acyclopropyl or ethyl group at this position. The steady-state concentrations (SSCs) of the 10novel quinolones accumulated by wild-type S. aureus did not correlate with their MICs or chemical structures. There was no apparent relationship between logP of the quinolone andaccumulation by S. aureus F77; however, accumulation was positively correlated with mo-lecular mass for 9/10 agents (r = 0.745) confirming that high molecular mass is not a barrier toaccumulation in S. aureus. For all 10 agents, the presence of carbonyl cyanide m-chlorophenyl-hydrazone (CCCP) increased the concentration of quinolone accumulated by SA-1199, suggesting that NorA was inhibited. The fold increase of the SSC in the presence of CCCP didnot correlate with hydrophobicity, but the SSC of agents with either an ethyl or cyclopropylgroup at R1 was increased two- to three-fold in the presence of CCCP, suggesting that affinityfor the NorA efflux pump may be influenced by quinolone structure.

*Corresponding author. Tel: �44-121-414-6966; Fax: �44-121-414-6966; E-mail: [email protected]

© 2001 The British Society for Antimicrobial Chemotherapy

JAC

L. J. V. Piddock et al.

NorA-mediated increased efflux has been described forquinolone-resistant S. aureus.8 Mutations in regulatorygenes controlling the expression of NorA have been shownto result in overexpression of the gene,9,10 and a regulatorylocus, arlR-arlS, has recently been described that is in-volved in the control of norA expression.11

Asuquo & Piddock12 compared the kinetics of quinoloneaccumulation in S. aureus with that into E. coli and Pseudo-monas aeruginosa. They concluded that accumulation wasmore rapid, and that quinolones were accumulated to ahigher concentration by S. aureus than by Gram-negativeorganisms. No correlation was found between the anti-bacterial activity of each agent and the amount of quino-lone accumulated in S. aureus, which confirmed earlierfindings.13,14 Takenouchi et al.15 investigated the relation-ship between the structure and hydrophobicity of quino-lones and their activity against a NorA-overproducingmutant of S. aureus and its parent strain. They found thatdecreased activity of the quinolones correlated with thebulkiness of the C7 substituent and the bulkiness andhydrophobicity of the C8 substituent, but not with hydro-phobicity. It has been proposed that for Gram-positiveorganisms, high molecular weight is not a barrier to penetra-tion.2,12

The proportion of charged and uncharged microspeciesof quinolones has been shown to affect accumulation in the bacterial cell.16 Quinolone species with no net charge(uncharged HQ0 and zwitterion HQ± forms) penetrate thecytoplasmic membrane of S. aureus more readily thancharged species (HQ+ and HQ–).17,18 Alovero et al.19 foundthat accumulation of sulphanilyl fluoroquinolones into S. aureus was decreased three-fold in the presence of magnes-ium ions and increased at lower pH. They suggested thatcompounds like the sulphanilyl fluoroquinolones, whichhave only one ionizable group, are accumulated in the cellmore favourably than are zwitterionic quinolones,18 andthat this may be important for their enhanced in vitro activ-ity against S. aureus, since the inhibition of DNA gyrase issimilar for both groups of agents.

The aim of this study was to determine whether thephysical properties of quinolones affect their accumulationand play a role in antibacterial activity against S. aureus, bycomparing the accumulation of 10 structurally similar fluoroquinolone molecules, with differing antibacterialactivities and hydrophobicities. The effect of the NorAefflux pump inhibitor, carbonyl cyanide m-chlorophenyl-hydrazone (CCCP) on accumulation of the 10 agents wasalso determined.

Materials and methods

Bacterial strains and media

S. aureus F77 (NCTC type strain 8532) and E. coli I113(NCTC 10418) were obtained from the National Collectionof Type Cultures, Colindale, UK. S. aureus F145 was a labor-

atory mutant derived from F77, which was ciprofloxacinresistant but susceptible to nalidixic acid. SA-1199 (wildtype), SA-1199B (grlA mutation and constitutive over-producer of NorA) and SA-1199-3 (inducible overproducerof NorA) were kindly donated by G. Kaatz, WashingtonState University.20 Ten structurally related fluoroquino-lones were provided by Parke Davis–Warner Lambert(Ann Arbor, MI, USA). All other antibacterial agentswere provided as gifts from the respective manufacturers.Bacteriological media were supplied by Unipath (Basing-stoke, UK) and chemicals by BDH or Sigma (Poole, UK).

Antibiotic susceptibility testing

The MICs of the 10 related fluoroquinolones (PD com-pounds) were determined by the microbroth doubling dilution method; MICs of all other agents were deter-mined by the agar doubling dilution technique as describedpreviously.12,21

Hydrophobicity and fluorescence of PD quinolones

The partition coefficient between 1-octanol and 0.1 Msodium phosphate buffer pH 7 was determined for eachquinolone as described by Asuquo & Piddock.12 The excitation (�ex) and emission (�em) maxima of the 10 PDcompounds were determined in 0.1 M glycine pH 3 and in50 mM phosphate buffer at pH 6, 7 and 8 (with and withoutboiling for 10 min) using a scanning spectrophotometer,and compared with other quinolones. The amount of fluor-escence emitted by 0.1 mg/L quinolone in 0.1 M glycine pH 3 and in 50 mM phosphate buffer at pH 6, 7 and 8 wasmeasured in a LS30 luminescence spectrophotometer(Perkin Elmer, Beaconsfield, UK).

Quinolone accumulation

The accumulation of each quinolone was determinedbetween three and six times, using the modified fluor-escence method.12 To investigate the role of the NorA effluxpump on accumulation of the PD agents by SA-1199,CCCP was added to a parallel set of tubes to a final concen-tration of 100 �M. In addition, the accumulation assayswere repeated and, instead of using 0.1 M glycine pH 3 tolyse the cells, the cell pellet was resuspended in 50 mMphosphate buffer pH 7 and boiled for 10 min. The concentra-tion of quinolone accumulated was determined by spectro-fluorimetry using the appropriate excitation and emissionmaxima for the agent.

Results

Comparison of quinolone structures

The chemical structures of the 10 PD quinolones were com-pared with those of common fluoroquinolones and nali-dixic acid (Table I) for similarities in the R1, R7 and R8

262

Quinolone accumulation by S. aureus

263

Tab

le I

.C

hem

ical

str

uctu

res,

phy

sica

l pro

pert

ies,

ant

ibac

teri

al a

ctiv

ity

and

accu

mul

atio

n by

S. a

ureu

san

d E

. col

ifor

10

PD

age

nts

and

com

mon

qui

nolo

nes

Subs

titu

ent

Hyd

roph

obic

ity

S. a

ureu

sF

77E

. col

iI11

3

Mol

ecul

arM

ICac

cum

ulat

iona

MIC

accu

mul

atio

na

Age

ntN

ucle

usR

1R

7R

8m

ass

Pap

plo

gP(m

g/L

)(n

g/m

g)(m

g/L

)(n

g/m

g)

PD

1129

56A

—C

2H5

H31

9.33

0.30

6–0

.514

115

6 (8

9)0.

1284

(79

)

PD

1144

56A

—C

2H5

H35

6.4

0.11

9–0

.924

113

4 (1

12)

0.5

38 (

86)

Nor

floxa

cin

A—

C2H

5H

319

0.02

2b–1

.657

172

(N

D)

0.03

ND

Nal

idix

ic a

cid

NA

—C

2H5

CH

3H

232

20.0

2b1.

301

128

54 (

ND

)2

ND

PD

1249

79A

H40

8.78

0.24

5–0

.611

0.12

172

(125

)0.

0674

(79

)

PD

1303

90A

H48

5.39

0.45

3–0

.344

0.25

97 (

127)

0.06

79 (

90)

Cip

roflo

xaci

nA

H33

1.4

0.03

1b–1

.509

0.5

96 (

ND

)�

0.03

ND


264

Tab

le I

.(C

ontin

ued)

Subs

titu

ent

Hyd

roph

obic

ity

S. a

ureu

sF

77E

. col

iI11

3

Mol

ecul

arM

ICac

cum

ulat

iona

MIC

accu

mul

atio

na

Age

ntN

ucle

usR

1R

7R

8m

ass

Pap

plo

gP(m

g/L

)(n

g/m

g)(m

g/L

)(n

g/m

g)

Spar

floxa

cin

AF

–N

D0.

12N

D0.

06N

D

PD

1354

98B

—C

(CH

3)3

H41

5.85

1.05

20.

022

0.12

130

(116

)0.

0648

(54

)

PD

1356

71B

—C

(CH

3)3

H41

5.08

0.60

1–0

.221

0.12

91 (

57)

0.06

63 (

50)

PD

1380

86B

H48

5.08

0.36

9–0

.433

0.12

165

(121

)0.

0335

(42

)

PD

1645

34B

H48

0.95

0.15

6–0

.807

0.06

46 (

121)

0.03

84 (

44)

Tem

aflox

acin

AH

419

0.74

6b–0

.127

184

(N

D)

0.25

ND


265

Tos

uflox

acin

AH

403

1.69

b0.

228

0.06

152

(ND

)0.

06N

D

PD

1644

13B

H38

6.94

0.30

3–0

.518

232

6 (1

02)

0.5

111

(76)

PD

1645

32B

—C

H(C

H3)

2H

391.

90.

148

–0.8

304

118

(124

)0.

581

(80

)

Oflo

xaci

nA

oxaz

ine

ring

link

age

see

R1

360

0.08

8b–1

.055

0.5

67 (

ND

)0.

12N

Dof

N1

and

C8

Nuc

leus

B is

a s

tere

oiso

mer

of n

ucle

us A

. Pap

p, p

arti

tion

coe

ffici

ent i

n 1-

octa

nol–

0.1

M s

odiu

m p

hosp

hate

buf

fer

pH 7

; NA

, not

app

licab

le (

the

nucl

eus

of n

alid

ixic

aci

d re

sem

bles

A b

ut is

no

n-flu

orin

ated

at R

6 po

siti

on);

ND

, not

don

e.a A

ccum

ulat

ion

was

mea

sure

d us

ing

two

met

hods

of c

ell l

ysis

: usi

ng 0

.1 M

gly

cine

pH

3 a

nd, i

n br

acke

ts, b

y bo

iling

in 5

0 m

M p

hosp

hate

buf

fer

pH 7

for

10 m

in.

b Dat

a fo

r flu

oroq

uino

lone

s ot

her

than

the

PD

age

nts

from

Asu

quo

& P

iddo

ck.12


substituents. All 10 PD compounds had only hydrogen atthe R8 position (like ciprofloxacin, norfloxacin, tema-floxacin and tosufloxacin), and all had a pyrrolidinyl groupat position R7 similar to that of tosufloxacin. At the R1position, PD112956 and PD114456 had an ethyl substituentlike norfloxacin, and PD124979 and PD130390 had a cyclopropyl ring like ciprofloxacin and sparfloxacin. BothPD138086 and PD164534 had a 2,4-difluorophenyl sub-stituent at the R1 position like temafloxacin and tosu-floxacin. PD135498 and PD135671 had a t-butyl substituentat R1, PD164413 a cyclobutyl and PD164532 an isopropylgroup (Table I).

Antibacterial activity

The antibacterial activity of the 10 PD agents against thefive strains of S. aureus was compared. Six PD agents(PD124979, PD130390, PD135498, PD135671, PD138086and PD164534) and tosufloxacin were more active againsteach S. aureus strain than the other four PD agents(PD112956, PD114456, PD164413 and PD164532), nalidixicacid, ciprofloxacin, norfloxacin and temafloxacin (TableII). The MICs of pairs of agents that had identical groups atthe R1 position but that differed only in the side chains onthe pyrrolidinyl ring at position R7, were very similar:PD112956 and PD114456 (ethyl group at R1); PD124979and PD130390 (cyclopropyl ring at R1); PD135498 and135671 (t-butyl group at R1), PD138086 and PD164534

(2,4-difluorophenyl group at R1) (Tables I and II). Thestructures of PD112956, PD135498 and PD164532 differedonly in the R1 substituent and showed varying activitiesagainst S. aureus. PD135498 was the most active of thesethree agents with a t-butyl group at R1 (MICs 0.03–1 mg/L),PD112956 was less active (ethyl group; MICs 0.25–16mg/L), while PD164532 was the least active with an iso-propyl group at R1 (MICs 0.5–32 mg/L).

F145 was two- to eight-fold less susceptible than its parent F77, to all the PD quinolones except PD135498. TheMICs of the PD agents were similar for SA-1199-3(inducible NorA overproducer) and its parent SA-1199;however, SA-1199B (constitutive NorA overproducer) waseight- to 64-fold less susceptible than SA-1199.

Physical properties of quinolones

The partition coefficient (Papp) determined for each PDagent (Table I) ranged from 0.119 (logP –0.924; PD114456)to 1.052 (logP 0.022; PD13548). The fluorescence excitationand emission maxima at pH 3, 6, 7 and 8 were determinedfor each PD agent and the comparative fluorescence ateach pH was measured in arbitrary units. The excitationmaxima differed by no more than 6 nm at different pH values for any of the PD compounds. The emission maximafor all the PD agents shifted to a shorter wavelength byapproximately 25–50 nm, as the pH increased. Six of thePD agents emitted more than twice the fluorescence at

266

Table II. Susceptibility to quinolones and accumulation of PD agents by mutant S. aureus

SA-1199 steady-state MIC (mg/L) concentration (ng/mg)

Agent F77 F145 SA-1199 SA-1199B SA-1199-3 –CCCP �CCCP fold increase

PD112956 1 4 0.25 16 0.25 14 43 3.1PD114456 1 4 0.5 16 0.5 54 147 2.7PD124979 0.12 1 0.12 2 0.12 40 71 1.8PD130390 0.25 1 0.12 1 0.12 62 105 1.7PD135498 0.12 0.12 0.03 1 0.06 52 66 1.3PD135671 0.12 0.25 0.06 1 0.06 42 58 1.4PD138086 0.12 0.25 0.03 0.5 0.03 30 51 1.6PD164413 2 4 0.25 16 1 39 63 1.6PD164532 4 8 1 32 0.5 87 107 1.2PD164534 0.06 0.12 0.03 0.5 0.06 25 34 1.4Ciprofloxacin 0.5 4 0.3 16 1.6 – – –Nalidixic acid 128 16 �512 �512 512 – – –Norfloxacin 1 16 0.5 16 4 – – –Sparfloxacin 0.12 0.12 0.1 0.2 0.1 – – –Temafloxacin 1 2 0.25 4 0.25 – – –Tosufloxacin 0.06 0.5 0.12 2 0.12 – – –

–CCCP and �CCCP, concentration of quinolone accumulated in the absence and presence of 100 �M CCCP, respectively. Bacterial cells werelysed in 0.1 M glycine pH 3 following accumulation assay.


267

pH 7 than at pH 3 (PD114456, PD124979, PD130390,PD138086, PD164413 and PD164534). The remaining fouragents (PD112956, PD135498, PD135671 and PD164532)fluoresced equally well at both pH values.

Quinolone accumulation

The steady-state concentrations (SSCs) of the PD quino-lones accumulated by wild-type S. aureus SA-1199 did notcorrelate with the MICs of the agents, e.g. PD138086 (MIC0.03 mg/L) was accumulated to approximately one-thirdthe concentration of PD164532 (MIC 1 mg/L) by SA-1199(Table II). Similarly, the SSCs of drugs accumulated by SA-1199 or F77 bore no correlation with the structure of the R1or R7 substituent. For example, the SSCs of some agentswith identical substituents at R1 accumulated by SA-1199were broadly similar (e.g. for PD138086 and PD164534,which have a 2,4-difluorophenyl group at R1, the SSCswere 30 and 25 ng/mg, respectively; and for PD135498 andPD135671, which have a t-butyl group, the SSCs were 52and 42 ng/mg). In contrast, other agents, such as PD112956and PD114456, which have an ethyl group at R1, have SSCsof 14 and 54 ng/mg, while PD124979 and PD130390 (cyclo-propyl group) have SSCs of 40 and 62 ng/mg, respectively.The concentrations of PD112956, PD135498 and PD-164532 accumulated (which were structurally identical apartfrom the substituents at the R1 position), were quite differ-ent: 14, 52 and 87 ng/mg, respectively.

For all 10 PD agents, the presence of CCCP increasedthe concentration of quinolone accumulated by S. aureusSA-1199 (Table II). The accumulation of agents with anethyl group at R1 (PD112956 and PD114456) and with acyclopropyl group at this position (PD124979 and PD-130390), was increased by CCCP approximately three- andtwo-fold, respectively. Although the SSC was increased2.7-fold by the presence of the NorA efflux inhibitor CCCPfor the least hydrophobic agent (PD114456; logP –0.924)compared with 1.3-fold for the most hydrophobic agent(PD135498; logP 0.022), there was no apparent correlationbetween logP and the fold increase of the SSC for the otheragents in the presence of CCCP.

Six of the 10 PD agents emitted 50% less fluorescence at pH 3 than at pH 7. However, only for PD164534 was the accumulation by F77 significantly lower after cell lysisat pH 3 (46 ng/mg) than at pH 7 (121 ng/mg) (Table I).Accumulation of the 10 PD agents by SA-1199 was deter-mined only using cell lysis at pH 3, but none of the sixagents that emitted less fluorescence at pH 3 were accumu-lated to lower concentrations compared with the otheragents.

There was no apparent relationship between the accu-mulation of the 10 PD agents, measured following cell lysisin pH 3 glycine or in pH 7 phosphate buffer, with boiling,and the logP of each agent for S. aureus F77 (r � 0.042 andr � 0.320, respectively, where r is the correlation coeffi-cient) or E. coli I113 (r � 0.162, r � 0.339; data not shown).

A weak positive correlation (r � 0.399) was shown betweenaccumulation into S. aureus F77 and molecular mass aftercell lysis at pH 7. However, the omission of the data pointfor PD135671 yielded a correlation equation of coefficientof 0.745 (Figure 1). However, there was no apparentrelationship between accumulation and molecular mass forS. aureus following lysis in pH 3 glycine (r � 0.361), or E. coli I113 following cell lysis at either pH 3 or pH 7 (r � 0.151, r � 0.492), nor between accumulation and MICfor S. aureus F77 (r � 0.290, r � 0.115) or E. coli I113 (r � 0.233, r � 0.534) with lysis at either pH (data notshown).

The relationship between the accumulation of the 10 PDagents by S. aureus F77 and SA-1199 and logP, molecularmass and MIC was compared and was similar for the twostrains, although SA-1199 accumulated each fluoroquino-lone to a lower concentration than F77.

Discussion

The comparison of MICs of the 10 PD compounds andtheir different structures suggested that changes in the sidechains of the R7 pyrrolidinyl ring alone had little effect onthe activity of these agents against S. aureus, but that thesubstituent at the R1 position was more important. For thisseries of fluoroquinolones, the t-butyl substituent was themost active group at the R1 position, and other workershave shown this substitution to be important for activityagainst Gram-positive bacteria.22,23 Although S. aureusF145 (ciprofloxacin-resistant mutant) and SA-1199B (con-stitutive NorA overproducer) were two- to 64-fold less sus-ceptible to the PD agents than their respective wild-typeparent strains, all the strains were more susceptible to thesix most active PD agents and tosufloxacin, than to nali-dixic acid, ciprofloxacin, norfloxacin and temafloxacin. Thisconfirms the observation that quinolones with an amino-pyrrolidine ring at R7 tend to be more active against Gram-positive organisms than piperazinyl quinolones.24,25

Figure 1. Relationship between molecular mass and accumula-tion of nine PD quinolones by S. aureus F77. Accumulation wasdetermined by the modified fluorescence method12 and cells werelysed in 50 mM phosphate buffer pH 7 with boiling for 10 min(SSC, steady-state concentration).


The concentration of quinolone accumulated did notcorrelate with the MIC for S. aureus F77 or SA-1199, assome agents that were accumulated to a high concentrationalso had a high MIC for S. aureus. This is in agreement withearlier studies.2,12,26 While the MIC data demonstrated thatthe R1 group was important for antibacterial activity, therewas no clear relationship between the chemical structureand the concentration accumulated for the 10 agents. Thedata concur with earlier studies which have concluded thatwhile quinolone structure does affect the interaction withthe target, it has little influence on the concentration of theagent accumulated.2,14

The concentration of all the PD agents accumulated bySA-1199 was increased in the presence of CCCP, suggest-ing that NorA was inhibited. However, the concentrationof those agents with an ethyl or cyclopropyl group at the R1position accumulated by SA-1199 was two- to three-foldhigher in the presence of CCCP, which may suggest a linkbetween quinolone structure and the affinity for the NorAefflux pump. Piddock et al.26 reported variation in the foldincrease of the concentration of quinolone accumulated byE. coli in the presence of CCCP for different agents, andproposed that the variance could reflect different affinitiesof an efflux pump for each quinolone. McCaffrey et al.6

found that CCCP had little effect on the accumulation ofofloxacin and fleroxacin by S. aureus, while others haveshown increased activity or accumulation of the hydro-philic quinolones, norfloxacin and ciprofloxacin, in the presence of efflux inhibitors.7,8,27,28 Takenouchi et al.15

demonstrated an increase in norfloxacin accumulation bySA-1199 and SA-1199B in the presence of CCCP, while theaccumulation of CS-940 was not affected. The variability inthe reported effects upon efflux by CCCP and other effluxinhibitors appears to be dependent on the quinolone andthe bacterial strain used and may reflect the affinity of thequinolone as a substrate for NorA, where those quinolonesthat are a poor substrate for the carrier protein are expelledfrom the bacterial cell less efficiently.15

Although NorA has been proposed to have a greateraffinity for hydrophilic quinolones,7 in this study, the foldincrease of the SSC for SA-1199 in the presence of CCCPdid not correlate with logP. Some studies have found a correlation between the concentration of fluoroquinoloneaccumulated and molecular hydrophobicity for S. aureus,2,6,7

but none was observed in the present study or others.12 Itremains to be established whether accumulation is deter-mined primarily by hydrophobicity or by quinolone struc-ture. Further investigation of how the quinolone structureaffects affinity for the NorA pump would be interesting.

In previous studies in this laboratory it had been ob-served that some agents were accumulated to low concentra-tions in bacteria, and that some of these agents alsofluoresced poorly. To address the question of whether thelow SSCs were an artefact of the method using lysis at pH 3,the fluoresence was determined at a range of pHs. In addition, the SSC was determined for S. aureus F77 after

cell lysis at both pH 3 and pH 7. Although six of the 10 PDquinolones emitted less than half the fluorescence at pH 3than at pH 7, the differences observed in the fluorescencewere not reflected in differences in the measured accumula-tion using the two methods of cell lysis. This is as would beexpected, as standard solutions used for calibration wereprepared in the same pH medium used for cell lysis for eachmethod. The preparation of quinolone standard solutionsin the lysis buffer should negate any differences due to lowfluorescence at pH 3. However, any errors in measuring theconcentration of quinolone in the lysed cell solution wouldbe greater where less fluorescence is emitted. Therefore,assuming cell lysis was complete under both conditionsused, the measurement of accumulation should be moreaccurate at pH 7, where 6/10 PD agents were more fluores-cent, than at pH 3.

The accumulation of the PD agents by S. aureus bore norelationship to either logP or the MIC. However, previousworkers have reported a positive correlation betweenquinolone accumulation and logP for S. aureus.2,6 Asuquo& Piddock12 found no direct relationship between hydro-phobicity or molecular mass and accumulation or withantibacterial activity. In addition, Bazile et al.2 found that amore bulky and hydrophobic quinolone (BMY 40397; logP1.05) was accumulated to a high concentration by S. aureus,demonstrating that molecular mass is not a limiting factorfor accumulation. In the present study, a positive correla-tion (r � 0.745) was shown between accumulation of 9/10PD agents in S. aureus F77 and molecular mass followingcell lysis at pH 7. The cause of this apparent associationobserved when cell lysis was carried out at pH 7 and not pH3, is not clear, but may be a consequence of the differentialfluorescence of the agents at the two pHs.

Recently, it has been proposed that the activity ofquinolones against S. aureus is dependent on the hydro-phobicity and/or bulkiness of the substituents at R7 andR8, and not the molecular hydrophobicity.15 Bulkiness ofthe R7 substituents of the PD agents used in this study wasestimated by comparison with the values determined byTakenouchi et al.15 for identical or similar substituents. Nocorrelation was observed between either the bulkiness ofthe R7 group and SSCs in F77, or the MIC increment ratioof SA-1199B and SA-1199 for this range of quinolones(data not shown).

Nikaido & Thanassi18 discussed the dynamics of quino-lone accumulation during the initial rapid phase and sug-gested that for resistant strains that actively efflux fluoro-quinolones from the cell (such as S. aureus expressingnorA), a higher rate of influx may enhance antibacterialactivity. Re-analysis of the data from Asuquo & Piddock12

shows that for S. aureus there is a direct relationshipbetween the rate of accumulation of fluoroquinolones dur-ing the initial rapid phase and hydrophobicity (r � 0.699;Figure 2), suggesting that the initial rate of influx is depend-ent on hydrophobicity. Further study of the complex relationship of quinolone accumulation by S. aureus,

268


particularly during the initial rapid phase, with hydro-phobicity, molecular mass and structure is warranted, andwill aid the development of new quinolone antibacterialagents with enhanced activity against S. aureus.

Acknowledgements

We acknowledge the assistance of Dr Tom Renau (previ-ously of Parke Davis–Warner Lambert, currently Micro-cide, Mountain View, CA) for providing information onthe PD compounds. This study was supported by a grantfrom the Medical Research Council ROPA scheme.

References

1. Ferrero, L., Cameron, B., Manse, B., Lagneaux, D., Crouzet, J.,Famechon, A. et al. (1994). Cloning and primary structure of Staphylococcus aureus DNA topoisomerase IV: a primary target offluoroquinolones. Molecular Microbiology 13, 641–53.

2. Bazile, S., Moreau, N., Bouzard, D. & Essiz, M. (1992). Relation-ships among antibacterial activity, inhibition of DNA gyrase andintracellular accumulation of 11 fluoroquinolones. AntimicrobialAgents and Chemotherapy 36, 2622–7.

3. Hirai, K., Aoyama, H., Irikura, T., Iyobe, S. & Mitsuhashi, S.(1986). Differences in susceptibility to quinolones of outer mem-brane mutants of Salmonella typhimurium and Escherichia coli.Antimicrobial Agents and Chemotherapy 29, 535–8.

4. Chapman, J. S. & Georgopapadakou, N. H. (1988). Routes ofquinolone permeation in Escherichia coli. Antimicrobial Agents andChemotherapy 32, 438–42.

5. Piddock, L. J. V., Hall, M. C. & Wise, R. (1990). Mechanism ofaction of lomefloxacin. Antimicrobial Agents and Chemotherapy 34,1088–93.

6. McCaffrey, C., Bertasso, A., Pace, J. & Georgopapadakou, N. H.(1992). Quinolone accumulation an Escherichia coli, Pseudomonasaeruginosa, and Staphylococcus aureus. Antimicrobial Agents andChemotherapy 36, 1601–5.

7. Yoshida, H., Bogaki, M., Nakamura, S., Ubukata, K. & Konno, M.(1990). Nucleotide sequence and characterization of the Staphylo-coccus aureus norA gene which confers resistance to quinolones.Journal of Bacteriology 172, 6942–9.

8. Kaatz, G. W., Seo, S. M. & Ruble, C. A. (1991). Mechanisms offluoroquinolone resistance in Staphylococcus aureus. Journal ofInfectious Diseases 163, 1080–6.

9. Kaatz, G. W., Seo, S. M. & Ruble, C. A. (1993). Efflux-mediatedfluoroquinolone resistance in Staphylococcus aureus. AntimicrobialAgents and Chemotherapy 37, 1086–94.

10. Ng, E. Y., Trucksis, M. & Hooper, D. C. (1994). Quinolone resist-ance mediated by norA: physiologic characterization and relation-ship to flqB, a quinolone resistance locus on the Staphylococcusaureus chromosome. Antimicrobial Agents and Chemotherapy 38,1345–55.

11. Fournier, B., Aras, R. & Hooper, D. C. (2000). Expression of themultidrug resistance transporter NorA from Staphylococcus aureusis modified by a two-component regulatory system. Journal of Bacteriology 182, 664–71.

12. Asuquo, A. E. & Piddock, L. J. (1993). Accumulation and killingkinetics of fifteen quinolones for Escherichia coli, Staphylococcusaureus and Pseudomonas aeruginosa. Journal of AntimicrobialChemotherapy 31, 865–80.

13. Bedard, J., Chamberland, S., Wong, S., Schollaardt, T. &Bryan, L. E. (1989). Contribution of permeability and sensitivity toinhibition of DNA synthesis in determining susceptibilities ofEscherichia coli, Pseudomonas aeruginosa and Alcaligenes faecalis to ciprofloxacin. Antimicrobial Agents and Chemotherapy33, 1457–64.

14. Diver, J. M., Piddock, L. J. & Wise, R. (1990). The accumulationof five quinolone antibacterial agents by Escherichia coli. Journal ofAntimicrobial Chemotherapy 25, 319–33.

15. Takenouchi, T., Tabata, F., Iwata, Y., Hanzawa, H., Sugawara,M. & Ohya, S. (1996). Hydrophilicity of quinolones is not an exclusivefactor for decreased activity in efflux-mediated resistant mutants ofStaphylococcus aureus. Antimicrobial Agents and Chemotherapy40, 1835–42.

16. Ross, D. L. & Riley, C. M. (1992). Physicochemical properties ofthe fluoroquinolone antimicrobials. II. Acid ionization constants andtheir relationship to structure. International Journal of Pharma-ceutics 83, 267–72.

17. Furet, Y. X., Deshusses, J. & Pechere, J. C. (1992). Transportof pefloxacin across the bacterial cytoplasmic membrane inquinolone-susceptible Staphylococcus aureus. Antimicrobial Agentsand Chemotherapy 36, 2506–11.

18. Nikaido, H. & Thanassi, D. G. (1993). Penetration of lipophilicagents with multiple protonation sites into bacterial cells: tetracyclineand fluoroquinolones as examples. Antimicrobial Agents andChemotherapy 37, 1393–9.

19. Alovero, F., Nieto, M., Mazzieri, M. R., Then, R. & Manzo, R. H.(1998). Mode of action of sulfanilyl fluoroquinolones. AntimicrobialAgents and Chemotherapy 42, 1495–8.

20. Kaatz, G. W. & Seo, S. M. (1995). Inducible NorA-mediatedmultidrug resistance in Staphylococcus aureus. AntimicrobialAgents and Chemotherapy 39, 2650–5.

21. Griggs, D. J., Hall, M. C., Jin, Y. F. & Piddock, L. J. (1994).Quinolone resistance in veterinary isolates of Salmonella. Journal ofAntimicrobial Chemotherapy 33, 1173–89.

22. Bouzard, D., Di Cesare, P., Essiz, M., Jacquet, J. P., Ledoussal,B., Remuzon, P. et al. (1992). Fluoronaphthyridines as anti-bacterial agents. 4. Synthesis and structure–activity relationships of 5-substituted-6-fluoro-7-(cycloalkylamino)-1,4-dihydro-4-oxo-1, 8-naphthyridine-3-carboxylic acids. Journal of Medicinal Chemistry35, 518–25.

269

Figure 2. Re-analysis of data for initial rate of accumulation offluoroquinolones into S. aureus F77 and logP.12


23. Bryskier, A. & Chantot, J. F. (1995). Classification and struc-ture–activity relationships of fluoroquinolones. Drugs 49, Suppl. 2,16–28.

24. Tillotson, G. S. (1996). Quinolones: structure–activity relation-ships and future predictions. Journal of Medical Microbiology 44,320–4.

25. Chu, D. T. & Fernandes, P. B. (1989). Structure–activity relationships of the fluoroquinolones. Antimicrobial Agents andChemotherapy 33, 131–5.

26. Piddock, L. J., Jin, Y. F., Ricci, V. & Asuquo, A. E. (1999).Quinolone accumulation by Pseudomonas aeruginosa, Staphylo-coccus aureus and Escherichia coli. Journal of AntimicrobialChemotherapy 43, 61–70.

27. Yoshida, S., Kojima, T., Inoue, M. & Mitsuhashi, S. (1991).Uptake of sparfloxacin and norfloxacin by clinical isolates of Staphylo-coccus aureus. Antimicrobial Agents and Chemotherapy 35,368–70.

28. Aeschlimann, J. R., Dresser, L. D., Kaatz, G. W. & Rybak, M. J.(1999). Effects of NorA inhibitors on in vitro antibacterial activitiesand postantibiotic effects of levofloxacin, ciprofloxacin, and nor-floxacin in genetically related strains of Staphylococcus aureus.Antimicrobial Agents and Chemotherapy 43, 335–40.

Received 25 May 2000; returned 6 September 2000; revised 23October 2000; accepted 12 November 2000

270

Iddock Uptake en Saureus ME CITA

Documents

Transcript of Iddock Uptake en Saureus ME CITA