Energy-dependent accumulation of norfloxacin and porin expression in clinical isolates of Klebsiella pneumoniae and relationship to extended-spectrum β-lactamase production

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 2002, p. 3926–3932 Vol. 46, No. 12 0066-4804/02/$04.00⫹0 DOI: 10.1128/AAC.46.12.3926–3932.2002 Copy igh © 2002, Ame ican Socie y o Mic obiology. All Righ s Rese ed. Ene gy-Dependen Accumula ion o No loxacin and Po in Exp ession in Clinical Isola es o Klebsiella pneumoniae and Rela ionship o Ex ended-Spec um ␤-Lac amase P oduc ion Luis Ma ínez-Ma ínez, 1,2 * Al a o Pascual, 1,2 Ma ía del Ca men Conejo, 1 Isabel Ga cía, 1 P o idencia Joyanes, 1 An onio Dome´nech-Sa´nchez, 3 and Vicen e Ja ie Benedí 3 Depa men o Mic obiology, School o Medicine, Uni e si y o Se ille, 1 and Uni e si y Hospi al V. Maca ena, 2 Se ille, and Labo a o y o Mic obiology, IMEDEA (CSIC-UIB), Uni e si a de les Illes Balea s, Palma de Mallo ca, 3 Spain Recei ed 28 Janua y 2002/Re u ned o modi ica ion 18 May 2002/Accep ed 8 Sep embe 2002 The ela ionships be ween po in de iciency, ac i e e lux o luo oquinolones, and ex ended-spec um ␤-lac- amase (ESBL) p oduc ion we e de e mined o 53 clinical isola es o Klebsiella pneumoniae. Thi y- wo ESBL-posi i e s ains (including 22 s ains exp essing po ins and 10 s ains lacking po ins) and 21 ESBL- nega i e s ains we e e alua ed. Ac i e e lux o no loxacin was de ined as a >50% inc ease in he accumu- la ion o no loxacin in he p esence o ca bonyl cyanide m-chlo ophenylhyd azone (CCCP) in compa ison wi h he co esponding basal alue in he absence o CCCP. The quinolone esis ance-de e mining egions o bo h gy A and pa C om 13 s ains, ep esen ing all isola es wi h di e en po in p o iles and wi h o wi hou ac i e e lux, we e de e mined. Po in loss was signi ican ly mo e common among ESBL-posi i e s ains (10 o 32 [31.2%]) han among ESBL-nega i e s ains (0 o 2 [0%]) (P< 0.01). Ac i e e lux was obse ed in 7 o 10 (70%) s ains lacking po ins and in 4 o 43 (9.3%) s ains p oducing po ins (P< 0.001). The 11 s ains showing ac i e e lux co esponded o 3 o 21 (14.3%) ESBL-nega i e s ains and 8 o 32 (25.5%) ESBL-posi i e s ains (P> 0.05). Basal alues o no loxacin accumula ion we e highe in s ains lacking ac i e e lux han in hose ha had his mechanism (P< 0.05). In he absence o opoisome ase changes, he con ibu ion o ei he po in loss o ac i e e lux o luo oquinolone esis ance in K. pneumoniae was negligible. I is concluded ha among K. pneumoniae s ains o clinical o igin, po in loss was obse ed only in hose p oducing ESBL, and ha a signi ican numbe o po in-de icien s ains also exp essed ac i e e lux o no loxacin. In e ms o luo o- quinolone esis ance, bo h mechanisms a e signi ican only in he p esence o opoisome ase modi ica ions. Klebsiella pneumoniae s ains p oducing ex ended-spec um ␤-lac amases (ESBL) a e mo e equen ly esis an o luo o- quinolones han K. pneumoniae s ains lacking hese enzymes (22). ESBL-p oducing K. pneumoniae s ains in e media e o esis an o cip o loxacin con ain gy A mu a ions, alone o combined wi h pa C mu a ions (4, 5, 6, 12, 25). I is also possible ha he plasmid coding o he ESBL, o an addi ional plasmid wi hin he same s ain, may con ain a qn -like de e - minan as desc ibed o o ganisms exp essing plasmid-medi- a ed AmpC ype enzymes (14). Qn is esponsible o low-le el esis ance o quinolones and a o s inc eased esis ance o hese d ugs (14). An al e na i e explana ion o ␤-lac am– luo oquinolone co esis ance in ESBL-p oducing K. pneumoniae is a dec ease in he pe meabili y o he ou e memb ane o bo h classes o agen s because o po in al e a ions. Non-ESBL-p oducing s ains usually exp ess he wo majo po ins (OmpK35 and OmpK36) o he species, while ESBL-p oducing s ains com- monly exp ess ei he only one o hese (no mally OmpK36) o no po in (9). Po in loss in ESBL-p oducing K. pneumoniae causes esis ance o ce oxi in and inc eased esis ance o de- c eased suscep ibili y o all oxyimino cephalospo ins, zwi e i- onic cephalospo ins, and ␤-lac am–␤-lac amase combina ions (1, 11, 13). Loss o hese channels causes a mode a e ( wo- o ou old) inc ease in he le el o esis ance o luo oquinolones when modi ica ions o opoisome ase II (alone o combined wi h changes in opoisome ase IV) a e p esen (5, 12). The di ec ole o ei he OmpK35 o OmpK36 in he esis ance o K. pneumoniae s ains o luo oquinolones and o ␤-lac ams has been demons a ed. Exp ession o ei he po in om a ec o con aining wild- ype ompK35 o ompK36 in s ains lack- ing bo h OmpK35 and OmpK36 dec eases he le el o esis- ance (1, 13; A. Dome´nech-Sa´nchez, S. He na´ndez-Alle´s, L. Ma ínez-Ma ínez, and V. J. Benedí, Abs . 38 h In e sci. Con . An imic ob. Agen s Chemo he ., abs . C186, 1998). In addi ion o dec eased pe meabili y, i is possible ha ac i e e lux also con ibu es o luo oquinolone esis ance in K. pneumoniae. A ew clinical isola es o ESBL-p oducing K. pneumoniae ha ha e po in de iciencies and exp ess ene gy- dependen e lux o luo oquinolones ha e been desc ibed (5, 12). In Esche ichia coli and Salmonella en e ica se o a Typhi- mu ium, pumps o e lux o quinolones may also expo ␤-lac- ams om he cells (19–21). Exp ession o he amA locus om K. pneumoniae in E. coli causes esis ance o no loxacin and is ela ed o bo h inc eased e lux o an imic obial agen s and loss o he OmpF po in (8). These ac i i ies a e simila o hose caused by he ansc ip ional ac i a o ma A o E. coli, which con ibu es o mul iple an imic obial esis ance by inc easing he ac i i y o he Ac AB-TolC e lux sys em and by dec easing he pe meabili y o he ou e memb ane due o inc eased an- * Co esponding au ho . Mailing add ess: Depa men o Mic obi- ology, School o Medicine, Uni e si y o Se ille, Apdo. 914, 41080 Se ille, Spain. Phone: 34 95 500 8287. Fax: 34 95 437 7413. E-mail: [email p o ec ed]. 3926 on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om sc ip ion o micF, a ansc ip ional ep esso o he po in gene ompF (3, 21). The objec i es o his s udy we e (i) o de e mine he ac i - i ies o se e al quinolones agains clinical isola es o K. pneu- moniae wi h cha ac e ized mechanisms o esis ance, (ii) o es ablish whe he po in loss and ac i e e lux o luo oquino- lones a e mo e equen ly obse ed in clinical isola es o K. pneumoniae p oducing ESBL han in hose no p oducing ESBL, (iii) o e alua e he po en ial ela ionship be ween po in exp ession and ac i e e lux, and (i ) o analyze he oles o po in loss and ac i e e lux in he quinolone esis ance le els o clinical s ains o K. pneumoniae wi h o wi hou mu a ions in opoisome ase-encoding genes. MATERIALS AND METHODS Bac e ial s ains. Fi y- h ee K. pneumoniae s ains o clinical o igin we e s udied. Th ee g oups o s ains we e selec ed o e alua ion in he p esen s udy: (i) 10 clonally un ela ed s ains o ESBL-p oducing K. pneumoniae de icien in bo h OmpK35 and OmpK36 exp ession (see below) isola ed ei he in ou labo- a o y o in ano he geog aphical zone (po in-de icien s ains), (ii) 22 clonally un ela ed ESBL-p oducing s ains exp essing po ins in hei ou e memb anes ( hese include 3 s ains clonally ela ed o hose o he p e ious g oup), and (iii) 21 clonally un ela ed non-ESBL-p oducing K. pneumoniae s ains. We lack, and a e no awa e o , ESBL-nega i e clinical s ains o K. pneumoniae de icien in he wo majo po ins o he species. The 32 ESBL-p oducing K. pneumoniae s ains included s ains C1 and C2 (14), s ain CSUB10R (1), and s ains MCQ-102, MCQ-121, and MCQ-122, kindly p o ided by P. B ad o d (2); s ains GAJ-3, GAJ-18, and GAJ-20, kindly p o ided by G. A. Jacoby (G. A. Jacoby, P. Han, M. Al a ez, and F. Teno e , Abs . 35 h In e sci. Con . An imic ob. Agen s Chemo he ., abs . C40, 1995); six s ains (LTZ205, LTZ-3273, RL-1, RL-2, RL-5, and RL-5X) om G eece, kindly p o ided by L. Tzou elekis (24); and 17 s ains om he Uni e si y Hospi al Vi gen Maca ena, Se ille, Spain, isola ed be ween 1996 and 1998. ESBL-de i- cien o ganisms we e isola ed be ween 1996 and 1998 a he Uni e si y Hospi al Vi gen Maca ena, Se ille, Spain (HUS s ains; n⫽19) o he Ciudad Sani a ia Uni e si a ia de Bell i ge, Ba celona, Spain (CSUB s ains; 2 s ains om he same pa ien ). O ganisms we e iden i ied wi h he API 20E sys em (bioMe´ ieux, La Balme Les G o es, F ance) and main ained in yp ic soy b o h (Di co, De oi , Mich.) wi h 20% glyce ol (Di co) a ⫺80°C un il use. Clonal ela ionship o s ains was de e mined by pulsed- ield gel elec opho e- sis (PFGE) o genomic DNA diges ed wi h SmaI as p e iously desc ibed (1). All ESBL-de icien s ains om Se ille ep esen ed di e en clones, while he wo CSUB s ains showed he same PFGE pa e n (da a no shown). E. coli ATCC 25922 and Pseudomonas ae uginosa ATCC 27853 we e used as con ol s ains in suscep ibili y es ing assays. Suscep ibili y es ing. MICs o an imic obial agen s we e de e mined by mi- c odilu ion acco ding o NCCLS guidelines (17). The ollowing agen s we e es ed: cip o loxacin (Baye , Le e kusen, Ge many), clina loxacin (Pa ke-Da is, Ann A bo , Mich.), moxi loxacin (Baye ), nalidixic acid (Sigma, Mad id, Spain), no loxacin (Sigma), pe loxacin (Rhoˆne-Poulenc, S . An oine, F ance), pipe- midic acid (Sigma), and o a loxacin (P ize , G o on, Conn.). When a ailable, clinical ca ego ies we e de e mined acco ding o NCCLS guidelines (18). ESBL p oduc ion was de ec ed by mic odilu ion, and s ains we e conside ed ESBL posi i e when cla ulanic acid ( ixed concen a ion, 2 ␮g/ml) caused a ⱖ8- old educ ion o he MICs o ce azidime, ce o axime, and/o az eonam. Analysis o OMPs. Bac e ial cells g own o loga i hmic phase we e lysed by sonica ion. Ou e memb ane p o eins (OMPs) we e ob ained a e ea men o cell memb anes eco e ed by ul acen i uga ion wi h sodium lau yl-sa cosyna e (2%; Sigma) and subsequen ul acen i uga ion. OMP p o iles we e de e mined by sodium dodecyl sul a e-polyac ylamide gel elec opho esis using 11% ac yl- amide, 0.345% bisac ylamide, and 0.1% sodium dodecyl sul a e in he unning gel. A e elec opho esis, p o eins we e s ained wi h Coomassie b illian blue (Sigma) (11). Accumula ion o no loxacin. Bac e ia g own in nu ien b o h we e suspended in phospha e-bu e ed saline (PBS) (ca. 0.650 mg o bac e ia [d y weigh ]/ml), incuba ed wi h 10 ␮go no loxacin/ml a 37°C o 30 min, and cen i uged h ough a silicone oil ba ie (␳, 1.029 g/cm 3 ) o elimina e ex acellula quino- lone. Fo he 11 s ains o 53 e alua ed ha did no a e se he silicon oil laye , ex acellula no loxacin was elimina ed by a wash wi h 1 ml o PBS (a 12,800 ⫻ gand 4°C o 1 min) (16). The cell pelle ob ained by ei he me hod was placed in 2 ml o 0.1 M glycine-HCl bu e (pH 3.0), o exed, and cen i uged o 5 min a 12,800 ⫻g. The amoun o no loxacin in he supe na an (basal accumula- ion) was measu ed spec opho ome ically (12). The e ec o he ene gy inhib- i o ca bonyl cyanide m-chlo ophenylhyd azone (CCCP) (0.1 mM) was e alu- a ed in pa allel by addi ion o CCCP o bac e ial suspensions 10 min a e he addi ion o no loxacin and eincuba ion o he cells o 20 min. Expe imen s we e done in duplica e on h ee di e en days. An o ganism was conside ed o exp ess ene gy-dependen accumula ion o no loxacin when CCCP enhanced basal accumula ion by a leas 50%, a h eshold which co esponds o wice he s anda d de ia ion o he mean basal accumula ion o he s ains es ed. Analysis o mu a ions in opoisome ase-encoding genes. Thi een s ains, ep- esen a i e o all possible combina ions among le el o suscep ibili y, po in p o ile, and exp ession o ac i e e lux, we e es ed. Al e a ions in he Gy A subuni o DNA gy ase and he Pa C subuni o opoisome ase IV we e de e - mined by sequencing he quinolone esis ance-de e mining egions (QRDRs) o he g yA and pa C genes by use o p ime s p e iously desc ibed by Deguchi e al. (4). S a is ical analyses. Rela ionships be ween ESBL p oduc ion, loss o po ins, and exp ession o ac i e e lux we e de e mined by he chi-squa e es and we e conside ed s a is ically signi ican a a Po ⬍0.05. No loxacin accumula ion alues we e exp essed as means ⫾s anda d de ia ions, and di e ences among g oups we e compa ed by analysis o a iance, which was used o assess s a is ical signi icance a a Po ⬍0.05. RESULTS Suscep ibili y es ing. Twen y- wo ou o he 53 s ains s ud- ied we e suscep ible o nalidixic acid (NAL-S), and 31 we e esis an o nalidixic acid (NAL-R; MICs, in all cases, we e ⱖ64 ␮g/ml). NAL-S s ains included 13 s ains p oducing ESBL and 9 s ains no p oducing ESBL. NAL-R s ains in- cluded 19 and 12 o ganisms p oducing and no p oducing ESBL, espec i ely (Fig. 1). All NAL-S s ains we e inhibi ed by pipemidic acid a 8 ␮g/ ml, and all NAL-R s ains we e inhibi ed by ⱖ16 ␮g o pipe- midic acid/ml, excep in one case (MIC o pipemidic acid, 4␮g/ml). All NAL-S s ains we e inhibi ed by ⱕ0.06 ␮go cip o loxacin o clina loxacin/ml. MICs o bo h no loxacin and pe loxacin o NAL-S s ains we e ⱕ0.06 ␮g/ml (16 s ains) and 0.125 ␮g/ml (6 s ains), while MICs o o a loxacin and moxi loxacin o NAL-S s ains we e ⱕ0.06 ␮g/ml (20 s ains), 0.125 ␮g/ml (1 s ain), and 0.25 ␮g/ml (1 s ain). The MIC 90 s (MICs a which 90% o isola es we e inhibi ed) o no loxacin and pe loxacin o NAL-S s ains we e 0.125 ␮g/ml, and MIC 90 s o he emaining ou luo oquinolones o his g oup o s ains we e ⱕ0.06 ␮g/ml. Dis ibu ions o MICs o luo oquinolones agains NAL-R s ains a e shown in Table 1. I can be obse ed ha MICs o no loxacin and pe loxacin o NAL-R s ains we e always ⱖ0.5 ␮g/ml, and hose o cip o loxacin we e always ⱖ0.125 ␮g/ ml. On he o he hand, MICs o clina loxacin we e ⱕ2␮g/ml o all 32 NAL-R s ains. As shown in Table 1, MIC 90 so all luo oquinolones o which he NCCLS es ablishes b eak- poin s all wi hin he esis an ca ego y. Analysis o OMPs. All 53 s ains es ed p oduced an OmpA- like p o ein o abou 32 kDa. All 21 ESBL-nega i e s ains (ei he NAL-S o NAL-R) p oduced OMPs in he ange o 35 o 37 kDa, compa ible wi h he hypo hesis ha hey a e po ins. Among ESBL-p oducing s ains, po in exp ession was ob- se ed in 22 s ains (68.8%), and no in he emaining 10 s ains (31.2%) (Fig. 1). These da a indica e ha po in loss is signi ican ly (P⬍0.01) mo e equen among ESBL-p oducing VOL. 46, 2002 EFFLUX, PORINS, AND ESBL IN K. PNEUMONIAE 3927 on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om s ains. Rep esen a i e esul s o OMP analysis a e p esen ed in Fig. 2. Accumula ion o no loxacin. In a se o independen expe - imen s in which he absolu e alue o no loxacin accumula ion was de e mined o 10 s ains ( ep esen a i e o he di e en accumula ion models obse ed) by bo h he silicon oil me hod and he PBS washing me hod, lowe absolu e alues we e ob- ained wi h he la e me hod, bu he pe cen ages o inc ease in no loxacin accumula ion caused by CCCP de e mined by he wo me hods we e no signi ican ly di e en (da a no shown). Fo his eason we ha e conside ed he pe cen age o inc ease in no loxacin accumula ion caused by CCCP o be an indica o o e lux, ega dless o whe he his was de e mined by he silicon oil me hod (42 s ains) o he PBS washing me hod (11 s ains). Only o 11 o 53 s ains did CCCP enhance he concen a- ion o accumula ed no loxacin mo e han 50% o e he co - esponding basal accumula ion alue. Ac i e e lux was sig- ni ican ly mo e equen (P⬍0.001) among po in-de icien s ains (7 o 10 [70.0%]) han among po in-p oducing s ains (4 o 43 [9.3%]). A highe pe cen age o s ains showing ac i e e lux o no loxacin was no ed among ESBL-posi i e s ains (8 o 32 [25.0%]) han among ESBL-nega i e s ains (3 o 21 [14.3%]), bu his di e ence was no s a is ically signi ican (P⬎0.05). Fo se en s ains he inc ease in no loxacin accu- mula ion due o CCCP was 210 o 275%, and o ou s ains he inc ease was 56 o 75%. Addi ion o CCCP caused a ⬍50% inc ease in he accumula ion o no loxacin in 19 s ains: 1 o 20% in 6 s ains, 11 o 20% in 8 s ains, and 31 o 49% in 5 s ains. In e es ingly, CCCP caused dec eased no loxacin ac- cumula ion in he emaining 23 s ains: he pe cen ages o dec ease we e 62% (n⫽1), 36% (n⫽1), 20 o 26% (n⫽3), 10 o 19% (n⫽10), and 1 o 9% (n⫽8). Among he 42 s ains o which accumula ion o no loxacin was assayed by he silicon oil me hod, basal accumula ion al- FIG. 1. Rela ionship be ween suscep ibili y o nalidixic acid, ESBL-p oduc ion, po in exp ession, and ac i e e lux o no loxacin (AEN) in 53 clinical isola es o K. pneumoniae. TABLE 1. Dis ibu ion o MICs, MIC 50 s, and MIC 90 so six luo oquinolones agains 31 NAL-R s ains o K. pneumoniae D ug No. o s ains inhibi ed a he ollowing MIC (␮g/ml): MIC 50 (␮g/ml) MIC 90 (␮g/ml) ⬍0.06 0.125 0.25 0.5 12481632ⱖ64 Cip o loxacin 1 1 5 1564 4 2 2 4 32 Clina loxacin 10 2 4 9 4 2 0.25 1 Moxi loxacin 4 3 6682 1 1 2 8 No loxacin 4 4225 8 2 4 8 64 Pe loxacin 4 2336 8 2 3 8 32 T o a loxacin 2 3 7484 2 1 2 8 3928 MARTI ´NEZ-MARTI ´NEZ ET AL. ANTIMICROB.AGENTS CHEMOTHER. on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om ues we e highe in s ains lacking ac i e e lux o no loxacin han in s ains showing ac i e e lux o no loxacin (P⬍0.05). No such compa ison was a emp ed o s ains assayed by he cen i uga ion me hod because none o hem exp essed ac i e e lux as p e iously de ined. Accumula ion o no loxacin in s ains lacking ac i e e lux a e cells we e ea ed wi h CCCP anged om 219 o 589 ng/mg o bac e ia (d y weigh ), while he co esponding ange in ac i e-e lux-posi i e s ains was 273 o 505 ng/mg o bac e ia (d y weigh ) (P⬎0.05). No loxacin accumula ion in 11 s ains wi h ac i e e lux o no loxacin is p esen ed acco ding o po in exp ession pa e n in Table 2. No loxacin accumula ion in h ee s ains de icien in po in exp ession and lacking ac i e e lux o no loxacin (Table 3) was no signi ican ly di e en (P⬎0.05) om ha obse ed o s ains ha do no exp ess ac i e e lux bu ex- p ess po ins. Ta ge modi ica ions. As shown in Table 4, he QRDRs o bo h Gy A and Pa C o he ou NAL-S s ains e alua ed we e iden ical o he sequences epo ed o K. pneumoniae s ain ATCC 13833: codons 83 and 87 o gy A coded o Se and Asp, espec i ely, while codons 78, 80, and 84 o pa C coded o Gly, Se , and Glu, espec i ely. The same esul s we e ob ained o wo s ains wi h low-le el esis ance o nalidixic acid: MICs o nalidixic acid we e 64 and 256 ␮g/ml o s ains HUS 33-99 and HUS 298-98, espec i ely. All s ains wi h high-le el esis ance o nalidixic acid (MIC ⱖ1,024 ␮g/ml) had mu a ions a gy A, ei he alone o combined wi h mu a ions in pa C. No cases o mu a ion in pa C we e obse ed in he absence o mu a ions in gy A. Mu a ions in codon 83 o gy A caused a change om Se o ei he Phe o Ty , while mu a ions in codon 87 caused a change om Asp o ei he Ty o Asn. Mu a ions in pa C caused changes a posi ion 78 (Gly o Cys), 80 (Se o A g), o 84 (Glu o Gly). The e we e no ob ious ela ionships be ween he ypes o mu a ions in opoisome ase genes and he le el o esis ance o luo oquinolones. No loxacin accumula ion and po in exp ession in pai s o clonally ela ed s ains (p esen ing he same PFGE pa e n) a e p esen ed in Table 5. DISCUSSION Resis ance o luo oquinolones in K. pneumoniae is mo e commonly ound among ESBL-p oducing s ains han among non-ESBL p oduce s (22). In his s udy he possible associa- ions be ween al e ed pe meabili y, ac i e e lux o no loxacin, and ESBL p oduc ion in clinical isola es o K. pneumoniae we e e alua ed. Addi ionally, opoisome ase changes we e also de e mined o some s ains in o de o e alua e in mo e de ail he ele ance o he abo e-men ioned mechanisms. The pos- FIG. 2. OMP p o iles o ESBL-p oducing (A) and non-ESBL-p oducing (B) K. pneumoniae s ains. (A) Lane 2, s ain HUS 76-96; lane 3, s ain HUS 56-96; lane 4, s ain HUS 33-96; lane 5, s ain HUS 71-96; lane 6, s ain HUS 20-97; lane 7, s ain HUS 91-96. (B) Lane 9, s ain RL-1; lane 10, s ain RL-2; lane 11, s ain RL-5X; lane 12, s ain RL-5b; lane 13, s ain GAJ-3; lane 14, s ain HUS 8-98. Lanes 1 and 8, molecula size ma ke s (in kilodal ons). Symbols: open ci cle, LamB-like p o ein; a owhead, po in; solid ci cle, OmpA-like p o ein. TABLE 2. Accumula ion o no loxacin wi hou and wi h 100 ␮M CCCP in 11 s ains o K. pneumoniae showing ene gy-dependen accumula ion o no loxacin and exp essing po ins o no S ain Accumula ion o no loxacin a (ng/mg o bac e ia [d y w ]) Basal b Wi h CCCP c Po in de icien RL-1 103 ⫾25 273 ⫾53 RL-5b 128 ⫾25 352 ⫾34 LTZ-3273 253 ⫾41 437 ⫾59 C1 247 ⫾61 505 ⫾55 C2 220 ⫾35 410 ⫾81 CSUB10R 188 ⫾19 429 ⫾50 MCQ-122 269 ⫾39 430 ⫾32 Po in exp essing CSUB-61A 298 ⫾22 445 ⫾13 HUS 71-96 245 ⫾59 436 ⫾73 HUS 298-98 182 ⫾80 417 ⫾56 GAJ-18 161 ⫾63 331 ⫾32 a Da a a e means ⫾s anda d de ia ions om h ee independen duplica e expe imen s. b Tes ed in he absence o CCCP. c Tes ed in he p esence o 100 ␮M CCCP. TABLE 3. Accumula ion o no loxacin wi hou and wi h CCCP in h ee s ains o K. pneumoniae de icien in po ins and lacking ene gy-dependen accumula ion o no loxacin S ain Accumula ion o no loxacin a (ng/mg o bac e ia [d y w ]) Basal Wi h 100 mM CCCP RL-2 441 ⫾80 387 ⫾90 GAJ-3 386 ⫾33 386 ⫾33 MCQ-102 454 ⫾119 372 ⫾88 a Da a a e exp essed as means ⫾s anda d de ia ions om h ee independen duplica e expe imen s. VOL. 46, 2002 EFFLUX, PORINS, AND ESBL IN K. PNEUMONIAE 3929 on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om sible p o ec ion o DNA gy ase by a Qn -like p o ein (14, 23) has been uled ou in a pa allel s udy (da a no shown). Figu e 1 shows he dis ibu ion o he 53 s ains es ed in his s udy acco ding o quinolone suscep ibili y, ESBL p oduc ion, po in exp ession, and ac i e e lux exp ession. Ou esul s indica e ha in clinical isola es o K. pneumo- niae, po in loss is mo e equen among ESBL-posi i e han among ESBL-nega i e s ains and ha in his o ganism ac i e e lux o no loxacin occu s mo e commonly among po in-de- icien s ains. Al hough he pe cen age o s ains showing ac- i e e lux was highe among ESBL-posi i e han among ESBL-nega i e isola es, his end was no s a is ically signi i- can . I is possible ha he me hod we used o de ec ing ac i e e lux is no sensi i e enough and/o ha he b eakpoin o de ining i (a 50% inc ease in accumula ion in he p esence o CCCP) is oo s ic (no e ha o i e s ains, CCCP enhanced accumula ion by 31 o 49%). I would be in e es ing o e alua e a la ge numbe o s ains, o measu e accumula ion o o he subs a es, o o use o he me hodological app oaches (an as- say wi h adiolabeled luo oquinolone, o immunological de- TABLE 4. Mu a ions a he QRDRs o gy A and pa C in 13 clinical s ains o K. pneumoniae S ain P esence o absence a o : MIC Mu a ion d ESBL Po in E lux NAL b CIP c Gy A Pa C Se 83 Asp87 Gly78 Se 80 Glu84 CSUB-61A ⫺⫹⫺ 0.5 0.004 ————— CSUB-61B ⫺⫹⫹ 0.5 0.008 ————— HUS 76-96 ⫺⫹⫺ 2ⱕ0.06 ————— LTZ-3273 ⫹⫺⫹ 16 ⱕ0.06 ————— HUS 33-96 ⫺⫹⫺ 64 0.125 ————— HUS 298-98 ⫺⫹⫹ 256 0.5 ————— C2 ⫹⫺⫹⬎1,024 0.5 Phe ———— MCQ-102 ⫹ e ⫺⫺⬎1,024 0.5 Ty ———— C1 ⫹⫺⫹⬎1,024 4 Ty ——— HUS 22-97 ⫹⫹⫺⬎1,024 4 Phe Ty ——Gly HUS 20-97 ⫺⫹⫺⬎1,024 8 Phe Ty ——Gly MCQ-121 ⫹ e ⫹⫺⬎1,024 64 Phe Asn Cys A g — MCQ-122 ⫹ e ⫺⫹⬎1,024 128 Phe Asn Cys A g — a ⫹, p esence; ⫺, absence. b NAL, nalidixic acid. c CIP, cip o loxacin. d —, no change wi h espec o wild- ype sequence. e AmpC ype ␤-lac amase p oduce . TABLE 5. Accumula ion o no loxacin wi hou and wi h CCCP in i e pai s o clonally ela ed K. pneumoniae s ains wi h di e en mechanisms o esis ance o luo oquinolones S ain PFGE pa e n P esence o absence a o : No loxacin accumu- la ions b (ng/mg o bac e ia [d y w ]) Mu a ion(s) c MIC (␮g/ml) ESBL Po in Basal Wi h CCCP Gy A Pa C NAL d CIP e NOR CLIN g CUSB-61A 1 ⫺⫹570 ⫾27 219 ⫾78 w w 0.5 0.004 0.03 ⱕ0.06 CUSB-61B 1 ⫺⫹298 ⫾22 445 ⫾13 w w 0.5 0.008 0.06 ⱕ0.06 RL-5X 2 ⫹⫹470 ⫾25 409 ⫾79 w w 8 ⱕ0.06 0.25 ⱕ0.06 RL-2 2 ⫹⫺441 ⫾80 387 ⫾90 w w 4 ⱕ0.06 0.5 ⱕ0.06 LB2 h 3⫹⫹286 ⫾24 385 ⫾65 Se 83Ty w ⬎1,024 0.5 2 ⱕ0.06 LB4 h 3⫹⫺116 ⫾21 263 ⫾38 Se 83Ty w ⬎1,024 1 4 ⱕ0.06 CSUB10S 4 ⫹⫹313 ⫾6 ND Se 83Phe w ⬎1,024 0.5 1 ⱕ0.06 CSUB10R 4 ⫹⫺188 ⫾19 429 ⫾50 Se 83Phe w ⬎1,024 4 8 ⱕ0.06 MCQ-121 5 ⫹ i ⫹434 ⫾53 405 ⫾79 Se 83Phe, Asp87Asn Gly78Cys, Se 80A g ⬎1,024 64 256 1 MCQ-122 5 ⫹ i ⫺298 ⫾22 445 ⫾13 Se 83Phe, Asp87Asn Gly78Cys, Se 80A g ⬎1,024 128 ⬎256 2 a ⫹, p esence; ⫺, absence. b Basal, wi hou CCCP; wi h CCCP, in he p esence o 100 ␮M CCCP. Da a a e means ⫾s anda d de ia ions om h ee independen duplica e expe imen s. ND, no de e mined. c w , wild ype. d NAL, nalidixic acid. e CIP, cip o loxacin. NOR, no loxacin. g CLIN, clina loxacin. h Da a o LB2 and LB4 ha e been epo ed p e iously (11) and a e included o compa ison. i Also p oducing he AmpC ype ACT-1 ␤-lac amase. 3930 MARTI ´NEZ-MARTI ´NEZ ET AL. ANTIMICROB.AGENTS CHEMOTHER. on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om ec ion o he e lux sys em p o eins) o con i m ha ac i e e lux is no ela ed o ESBL p oduc ion. The da a in Table 1 indica e he exis ence o c oss- esis ance o nalidixic acid and luo oquinolones in K. pneumoniae, al- hough he e ec s o he mechanisms o esis ance a e di e en o di e en compounds. Clina loxacin was, in e ms o MIC 90 s, he mos ac i e luo oquinolone o hose es ed he e, ollowed by moxi loxacin, o a loxacin, cip o loxacin, pe loxa- cin, and no loxacin. K. pneumoniae ATCC 13833 and many o he quinolone- suscep ible clinical isola es con ain se ine ins ead o h eonine a posi ion 83, as p e iously epo ed o K. pneumonaie M5a1 (7). A ecen epo has shown ha Klebsiella oxy oca 13182 con ains h eonine (25). Ou da a wi h ou NAL-S s ains also indica e ha se ine is no mally ound a posi ion 83 o Gy A in K. pneumoniae, while Phe o Ty is p esen in K. pneumoniae s ains wi h high-le el esis ance o NAL. Changes in Gy A Asp87 o Ty o o Asp ha e been obse ed in ou NAL-R s ains. These ha e also been he mos equen subs i u ions ound in o he s udies (4, 5, 6, 25). Mu a ions in pa C (coding o Se 80A g and Glu84Gly) we e obse ed in ou ou o i e NAL-R and cip o loxacin- esis an s ains. In he wo clonally ela ed s ains MCQ-121 and MCQ-122, a mu a ion in pa C ha is esponsible o a Gly78Cys change and ha has no been pe iously epo ed o K. pneumoniae was no ed. The ele- ance o his change awai s u he s udy. This s udy and p e ious s udies (4, 5, 6, 12) indica e ha he le els o esis ance o nalidixic acid and luo oquinolones in clinical isola es o K. pneumoniae a e no comple ely explained by changes in he QRDRs o gy A and pa C (Table 4; compa e s ain HUS76-96 wi h s ain LTZ-3273 and s ain MCQ-102 wi h s ain C1). S ains CSUB10R and LB4 a e de icien in po ins, con ain a Gy A change, and in addi ion exhibi ene gy-dependen accu- mula ion o luo oquinolones (11, 12). On he o he hand, he clonally ela ed s ains CSUB10S and LB2 exp ess OmpK36 and do no show ene gy-dependen accumula ion o luo o- quinolones. This sugges s a link be ween he loss o he wo po ins and exp ession o ene gy-dependen accumula ion o luo oquinolones in clinical isola es o K. pneumoniae. To e al- ua e his hypo hesis, no loxacin accumula ion was de e mined in a collec ion o un ela ed po in-de icien s ains and, o compa ison, in a ep esen a i e numbe o po in-exp essing s ains (ei he p oducing o no p oducing ESBL). We measu ed no loxacin accumula ion by using a luo o- me ic assay, elimina ing ex acellula luo oquinolone by cen- i uga ion h ough a silicon oil ba ie o by washing cells wi h PBS. In his s udy, lowe no loxacin accumula ion was ob- ained when he PBS washing me hod ins ead o he silicon oil me hod was used, in ag eemen wi h p e ious epo s (16; unpublished da a). Al hough he ac ual easons o hese di - e ences ha e no ye been sa is ac o ily explained, he ap- p oach we used o de ining ac i e e lux as a a io o accumu- la ion in he p esence o CCCP o accumula ion in he absence o CCCP allows compa ison o he esul s ob ained by hese wo me hods. In some o he s ains we s udied, CCCP caused dec eased ( a he han inc eased) accumula ion o no loxacin. I can be hypo hesized ha , in hese s ains, CCCP induces he exp es- sion o one o mo e e lux pumps also in ol ed in luo oquin- olone elimina ion, bu u he wo k is needed o cla i y hese esul s. Se en o he 11 s ains exp essing ac i e e lux o no loxacin we e po in de icien , while among he 10 po in-de icien s ains e alua ed, 7 exp essed ene gy-dependen accumula ion o luo- oquinolones (Fig. 1). Accumula ions (ei he in he absence o in he p esence o CCCP) in he se en s ains exp essing e lux o luo oquino- lones and lacking po ins we e no s a is ically di e en (P⬍ 0.05) om hose in ou s ains exp essing ene gy-dependen accumula ion o luo oquinolones and p oducing po ins (Table 2), while accumula ions in s ains lacking ac i e e lux we e no di e en (P⬎0.05) o s ains exp essing o lacking po ins (Table 3). This sugges s ha when he majo po ins o K. pneumoniae a e los , no loxacin (and pe haps o he luo o- quinolones) may s ill accumula e in he cell by o he means and ha dec eased accumula ion depends mo e on elimina ion by ac i e e lux han on dec eased pe meabili y by po in loss. Fu he s udies a e equi ed o deduce he independen oles o ac i e e lux and o loss o po ins in he inal amoun o no loxacin accumula ed in s ains exp essing bo h mecha- nisms. In a p e ious s udy i was shown ha when o he mecha- nisms o esis ance a e absen , loss o po ins in K. pneumoniae C3 does no con ibu e signi ican ly o quinolone esis ance (10). In his s udy we ha e obse ed, o s ains CSUB-61A and CSUB-61B, ha exp ession o ene gy-dependen accumula ion o luo oquinolones alone does no signi ican ly inc ease he le el o esis ance (Table 5). When bo h po in loss and ac i e e lux a e p esen in he same o ganism ha has a opoisom- e ase change(s), a mode a e inc ease in he le el o luo o- quinolone esis ance is obse ed. This has been desc ibed p e- iously o s ains LB2 and LB4 (11) and o s ains CSUB10S and CSUB10R (1) and has been obse ed he e o s ains MCQ-121 and MCQ-122 (wi h double mu a ions in bo h gy A and pa C). This indica es ha mul iple a ge mu a ions a e much mo e impo an han nonspeci ic mechanisms o quin- olone esis ance, bu he la e may be a c i ical s ep a o ing he eme gence o mu an s wi h al e ed a ge s. K. pneumoniae may exp ess a basal e lux o luo oquinolo- nes, as al eady obse ed o E. coli and o he en e obac e ia. The wide ange o no loxacin accumula ion in he s ains we s udied may e lec a ia ion in unde ined ac o s ha de e - mine binding o he d ug o he cell bu may also ep esen a iable exp ession o ac i e e lux in di e en o ganisms. In E. coli, Ac AB-TolC is he mo e clinically ele an e lux sys em, since i is in ol ed in he elimina ion o luo oquinolones, ␤-lac ams, and o he an imic obial agen s (15, 20). P elimina y esul s (da a no shown) indica e ha K. pneumoniae con ains an Ac AB homologue. New s udies a e in p og ess o e alua e he impo ance o Ac AB-TolC and o he e lux sys ems in he s ains included in his epo . ACKNOWLEDGMENTS We hank C. A danuy, P. B ad o d, G. A. Jacoby, and L. Tzou ele- kis o supplying bac e ial s ains. This wo k was suppo ed by he Fondo de In es igacio´n Sani a ia, Minis e io de Sanidad y Consumo o Spain, g an s 00/0242 ( o L.M.-M.) and 01/0034 ( o V.J.B.). A.D.-S. was suppo ed by a p edoc- o al ellowship om he Minis e io de Educacio´n y Cul u a o he Spanish go e nmen . VOL. 46, 2002 EFFLUX, PORINS, AND ESBL IN K. PNEUMONIAE 3931 on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om REFERENCES 1. A danuy, C., J. Lin˜a es, M. A. Domínguez, S. He na´ndez-Alle´s, V. J. Benedí, and L. Ma ínez-Ma ínez. 1998. Ou e memb ane p o iles o clonally e- la ed Klebsiella pneumoniae isola es om clinical samples and ac i i ies o cephalospo ins and ca bapenems. An imic ob. Agen s Chemo he . 42:1636– 1640. 2. B ad o d, P. A., C. U ban, N. Ma iano, S. J. P ojan, J. J. Rahal, and K. Bush. 1997. Imipenem esis ance in Klebsiella pneumoniae is associa ed wi h he combina ion o ACT-1, a plasmid-media ed AmpC ␤-lac amase, and he loss o an ou e memb ane p o ein. An imic ob. Agen s Chemo he . 41:563– 569. 3. Cohen, S. P., L. M. McMu y, and S. B. Le y. 1989. C oss- esis ance o luo oquinolones in mul iple-an ibio ic- esis an (Ma ) Esche ichia coli se- lec ed by e acycline o chlo amphenicol: dec eased d ug accumula ion associa ed wi h memb ane changes in addi ion o OmpF educ ion. An imi- c ob. Agen s Chemo he . 33:1318–1325. 4. Deguchi, T., A. Fukuoka, M. Yasuda, M. Nakano, S. Ozeki, E. Kanema su, Y. Nishino, S. Ishiha a, Y. Ban, and Y. Kawada. 1997. Al e a ions in he Gy A subuni o DNA gy ase and he Pa C subuni o opoisome ase IV in quinolone- esis an clinical isola es o Klebsiella pneumoniae. An imic ob. Agen s Chemo he . 41:699–701. 5. Deguchi, T., T. Kawamu a, M. Yasuda, M. Nakano, H. Fukuda, H. Ka o, N. Ka o, Y. Okano, and Y. Kawada. 1997. In i o selec ion o Klebsiella pneu- moniae s ains wi h enhanced quinolone esis ance du ing luo oquinolone ea men o u ina y ac in ec ions. An imic ob. Agen s Chemo he . 41: 1609–1611. 6. Deguchi, T., M. Yasuda, T. Kawamu a, M. Nakano, S. Ozeki, E. Kanema su, Y. Nishino, and Y. Kawada. 1997. Imp o ed an imic obial ac i i y o DU- 6859a, a new luo oquinolone, agains quinolone- esis an Klebsiella pneu- moniae and En e obac e cloacae isola es wi h al e a ions in Gy A and Pa C p o eins. An imic ob. Agen s Chemo he . 41:2544–2546. 7. Dim i, G. P., and H. K. Das. 1990. Cloning and sequence analysis o gy A gene o Klebsiella pneumoniae. Nucleic Acids Res. 18:151–156. 8. Geo ge, A. M., R. M. Hall, and H. W. S okes. 1995. Mul id ug esis ance in Klebsiella pneumoniae: a no el gene, amA, con e s a mul id ug esis ance pheno ype in Esche ichia coli. Mic obiology 141:1909–1920. 9. He na´ndez-Alle´s, S., S. Albe í, D. Al a ez, A. Dome´nech-Sa´nchez, L. Ma - ínez-Ma ínez, J. Gil, J. M. Toma´s, and V. J. Benedí.1999. Po in exp ession in clinical isola es o Klebsiella pneumoniae. Mic obiology 145:673–679. 10. He na´ndez-Alle´s, S., M. C. Conejo, A. Pascual, J. M. Toma´s, V. J. Benedí, and L. Ma ínez-Ma ínez. 2000. Rela ionship be ween ou e memb ane al e a ions and suscep ibili y o an imic obial agen s in isogenic s ains o Klebsiella pneumoniae. J. An imic ob. Chemo he . 46:273–277. 11. Ma ínez-Ma ínez, L., S. He na´ndez-Alle´s, S. Albe í, J. M. Toma´s, V. J. Benedí, and G. A. Jacoby. 1996. In i o selec ion o po in-de icien mu an s o Klebsiella pneumoniae wi h inc eased esis ance o ce oxi in and expand- ed-spec um cephalospo ins. An imic ob. Agen s Chemo he . 40:342–348. 12. Ma ínez-Ma ínez, L., I. Ga cía, S. Balles a, V. J. Benedí, S. He na´ndez- Alle´s, and A. Pascual. 1998. Ene gy-dependen accumula ion o luo oquino- lones in quinolone- esis an Klebsiella pneumoniae s ains. An imic ob. Agen s Chemo he . 42:1850–1852. 13. Ma ínez-Ma ínez, L., A. Pascual, S. He na´ndez-Alle´s, D. Al a ez-Díaz, A. I. Sua´ ez, J. T an, V. J. Benedí, and G. A. Jacoby. 1999. Roles o ␤-lac- amases and po ins in ac i i ies o ca bapenems and cephalospo ins agains Klebsiella pneumoniae. An imic ob. Agen s Chemo he . 43:1669–1673. 14. Ma ínez-Ma ínez, L., A. Pascual, and G. A. Jacoby. 1998. Quinolone e- sis ance om a ans e able plasmid. Lance 351:797–799. 15. Mazza iol, A., Y. Tokue, T. M. Kanegawa, G. Co naglia, and H. Nikaido. 2000. High-le el luo oquinolone- esis an clinical isola es o Esche ichia coli o e p oduce mul id ug e lux p o ein Ac A. An imic ob. Agen s Chemo he . 44:3441–3443. 16. Mo ime , P. G., and L. J. Piddock. 1991. A compa ison o me hods used o measu ing he accumula ion o quinolones by En e obac e iaceae,Pseudomo- nas ae uginosa and S aphylococcus au eus. J. An imic ob. Chemo he . 28: 639–653. 17. Na ional Commi ee o Clinical Labo a o y S anda ds. 1997. Me hods o dilu ion suscep ibili y es s o bac e ia ha g ow ae obically, 4 h ed. Ap- p o ed s anda d M7-A4. Na ional Commi ee o Clinical Labo a o y S an- da ds, Wayne, Pa. 18. Na ional Commi ee o Clinical Labo a o y S anda ds. 2001. Pe o mance s anda ds o an imic obial suscep ibili y es ing, 11 h in o ma ional supple- men . NCCLS documen M100-S11. Na ional Commi ee o Clinical Lab- o a o y S anda ds, Wayne, Pa. 19. Nikaido, H., M. Basina, V. Nguyen, and E. Y. Rosenbe g. 1998. Mul id ug e lux pump Ac AB o Salmonella yphimu ium exc e es only hose ␤-lac am an ibio ics con aining lipophilic side chains. J. Bac e iol. 180:4686–4692. 20. Nikaido, H., and H. I. Zgu skaya. 2001. Ac AB and ela ed mul id ug e lux pumps o Esche ichia coli. J. Mol. Mic obiol. Bio echnol. 3:215–218. 21. Okusu, H., D. Ma, and H. Nikaido. 1996. Ac AB e lux pump plays a majo ole in he an ibio ic esis ance pheno ype o Esche ichia coli mul iple-an i- bio ic- esis ance (Ma ) mu an s. J. Bac e iol. 178:306–308. 22. Pa e son, D. L., L. Mulazimoglu, J. M. Casellas, W. C. Ko, H. Goossens, A. on Go be g, S. Mohapa a, G. M. T enholme, K. P. Klugman, J. G. Mc- Co mack, and V. L. Yu. 2000. Epidemiology o cip o loxacin esis ance and i s ela ionship o ex ended-spec um ␤-lac amase p oduc ion in Klebsiella pneumoniae isola es causing bac e emia. Clin. In ec . Dis. 30:473–478. 23. T an, J. H., and G. A Jacoby. 2002. Mechanism o plasmid-media ed quin- olone esis ance. P oc. Na l. Acad. Sci. USA 99:5638–5642. 24. Tzou elekis, L., E. Tzelepi, E. P ina akis, M. Gazouli, A. Ka ahou a, P. Giakkoupi, O. Pania a, and N. J. Legakis. 1998. Spo adic eme gence o Klebsiella pneumoniae s ains esis an o ce epime and ce pi ome in G eek hospi als. J. Clin. Mic obiol. 36:266–268. 25. Weigel, L. M., C. D. S ewa d, and F. C. Teno e . 1998. gy A mu a ions associa ed wi h luo oquinolone esis ance in eigh species o En e obac e i- aceae. An imic ob. Agen s Chemo he . 42:2661–2667. 3932 MARTI ´NEZ-MARTI ´NEZ ET AL. ANTIMICROB.AGENTS CHEMOTHER. on July 31, 2017 by USE/BCTA.GEN UNIVERSITARIAh p://aac.asm.o g/Downloaded om