dissemination of escherichia coli producing ampc-type β-lactamase (cmy-11) in korea

7
International Journal of Antimicrobial Agents 24 (2004) 320–326 Dissemination of Escherichia coli producing AmpC-type -lactamase (CMY-11) in Korea Jae Young Kim b,c , Jae Seok Song a , Song Hee Bak b , Young Eun Cho b , Dae Wi Kim c , Seok Hoon Jeong e , Yeong-Min Park d , Kye Joon Lee c,1 , Sang Hee Lee a,b,a Department of Biological Science, Myongji University, San 38-2 Namdong, Yongin, Kyunggido, 449-728, Republic of Korea b Bio Technology Innovation Centre, Youngdong University, Chungbuk 370-701, Republic of Korea c School of Biological Sciences, Seoul National University, Seoul 151-742, Republic of Korea d Department of Microbiology, Pusan National University College of Medicine, Pusan 602-739, Republic of Korea e Department of Clinical Pathology, Kosin University College of Medicine, Pusan and Research Institute of Bacterial Resistance, Yonsei University College of Medicine, Seoul, Republic of Korea Received 4 December 2003; accepted 31 March 2004 Abstract Among the 51 clinical isolates collected from a university hospital in Korea, nine isolates were resistant to cephamycins. Nine isolates were shown to produce CMY-11 and these also included three isolates producing TEM-1. The results from ERIC-PCR revealed that dissemination of CMY-11 was due to outbreaks of resistant species and to the intra-species spread of resistance to cephamycins in Korea. CMY-11 -lactamase genes from nine clinical isolates that were responsible for resistance to cephamycins (cefoxitin and cefotetan), amoxicillin, cephalothin and amoxicillin-clavulanic acid, were cloned and characterised. A sequence identical to the common regions in In6, In7 and a novel integron from pSAL-1 was found upstream from bla CMY-11 gene at nucleotide 1–71. Eighteen nucleotides between position 71 and 72 were inserted into the bla CMY-11 gene. © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Escherichia coli; AmpC beta-lactamase; CMY-11; ERIC-PCR 1. Introduction The production of -lactamase is the single most-prevalent mechanism responsible for resistance to -lactams among clinical isolates of the family Enterobacteriaceae [1]. -Lactamases have been classified into classes A, B, C and D according to their amino acid homology [2]. AmpC -lactamases, mostly conferring resistance to many -lactam antibiotics (cephamycins and combina- tions of -lactams and -lactamase inhibitors), are in- cluded in class C -lactamases. Plasmid-mediated AmpC -lactamases are often expressed in large amounts and can pose therapeutic problems [3]. Plasmid-mediated Corresponding author. Tel.: +82 31 330 6195; fax: +82 31 335 8249. E-mail addresses: [email protected] (K.J. Lee), [email protected] (S.H. Lee). 1 Co-corresponding author. Tel.: +82 2 880 6705; fax: +82 2 882 9285. AmpC -lactamases have been recently reported in K. pneumoniae (CMY-1, CMY-2, CMY-8, MOX-1, MOX-2, FOX-1, FOX-5, LAT-1, LAT-2, LAT-2b, ACT-1, MIR-1 and ACC-1), K. oxytoca (CMY-5 and FOX-3), E. coli (CMY-4, CMY-6, CMY-7, CMY-9, FOX-2, FOX-4, BIL-1, LAT-3 and LAT-4), Salmonella Enteri- tidis (DHA-1), Proteus mirabilis (CMY-3) and Salmonella Seftenberg (CMY-2b) (the website of The Hall Labo- ratory of Experimental Evolution, http://www.rochester. edu/Colledge/BIO/labs/HallLab/AmpC-Phylo.html). In a previous study, we have demonstrated the involvement of a new -lactamase (CMY-11) in -lactam resistance in a clinical isolate of E. coli [4]. Because of the risk of AmpC resistance determinants among enterobacte- rial isolates, it is necessary to elucidate the AmpC re- sistance mechanism. The present study was conducted to determine the prevalence and genotypes of AmpC -lactamases from clinical isolates at a university hospital in Korea. 0924-8579/$ – see front matter © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2004.03.023

Upload: jae-young-kim

Post on 27-Nov-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

International Journal of Antimicrobial Agents 24 (2004) 320–326

Dissemination ofEscherichia coliproducing AmpC-type�-lactamase (CMY-11) in Korea

Jae Young Kimb,c, Jae Seok Songa, Song Hee Bakb, Young Eun Chob, Dae Wi Kimc,Seok Hoon Jeonge, Yeong-Min Parkd, Kye Joon Leec,1, Sang Hee Leea,b,∗

a Department of Biological Science, Myongji University, San 38-2 Namdong, Yongin, Kyunggido, 449-728, Republic of Koreab Bio Technology Innovation Centre, Youngdong University, Chungbuk 370-701, Republic of Korea

c School of Biological Sciences, Seoul National University, Seoul 151-742, Republic of Koread Department of Microbiology, Pusan National University College of Medicine, Pusan 602-739, Republic of Korea

e Department of Clinical Pathology, Kosin University College of Medicine, Pusan and Research Institute of Bacterial Resistance,Yonsei University College of Medicine, Seoul, Republic of Korea

Received 4 December 2003; accepted 31 March 2004

Abstract

Among the 51 clinical isolates collected from a university hospital in Korea, nine isolates were resistant to cephamycins. Nine isolates wereshown to produce CMY-11 and these also included three isolates producing TEM-1. The results from ERIC-PCR revealed that dissemination ofCMY-11 was due to outbreaks of resistant species and to the intra-species spread of resistance to cephamycins in Korea. CMY-11�-lactamasegenes from nine clinical isolates that were responsible for resistance to cephamycins (cefoxitin and cefotetan), amoxicillin, cephalothin andamoxicillin-clavulanic acid, were cloned and characterised. A sequence identical to the common regions in In6, In7 and a novel integron frompSAL-1 was found upstream fromblaCMY-11 gene at nucleotide 1–71. Eighteen nucleotides between position 71 and 72 were inserted into theblaCMY-11 gene.© 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

Keywords: Escherichia coli; AmpC beta-lactamase; CMY-11; ERIC-PCR

1. Introduction

The production of�-lactamase is the single most-prevalentmechanism responsible for resistance to�-lactams amongclinical isolates of the family Enterobacteriaceae[1].�-Lactamases have been classified into classes A, B,C and D according to their amino acid homology[2].AmpC �-lactamases, mostly conferring resistance tomany �-lactam antibiotics (cephamycins and combina-tions of �-lactams and�-lactamase inhibitors), are in-cluded in class C�-lactamases. Plasmid-mediated AmpC�-lactamases are often expressed in large amounts andcan pose therapeutic problems[3]. Plasmid-mediated

∗ Corresponding author. Tel.:+82 31 330 6195; fax:+82 31 335 8249.E-mail addresses:[email protected] (K.J. Lee),

[email protected] (S.H. Lee).1 Co-corresponding author. Tel.:+82 2 880 6705;

fax: +82 2 882 9285.

AmpC �-lactamases have been recently reported inK. pneumoniae (CMY-1, CMY-2, CMY-8, MOX-1,MOX-2, FOX-1, FOX-5, LAT-1, LAT-2, LAT-2b, ACT-1,MIR-1 and ACC-1), K. oxytoca (CMY-5 and FOX-3),E. coli (CMY-4, CMY-6, CMY-7, CMY-9, FOX-2,FOX-4, BIL-1, LAT-3 and LAT-4), Salmonella Enteri-tidis (DHA-1), Proteus mirabilis(CMY-3) and SalmonellaSeftenberg (CMY-2b) (the website of The Hall Labo-ratory of Experimental Evolution,http://www.rochester.edu/Colledge/BIO/labs/HallLab/AmpC-Phylo.html). In aprevious study, we have demonstrated the involvementof a new �-lactamase (CMY-11) in�-lactam resistancein a clinical isolate ofE. coli [4]. Because of the riskof AmpC resistance determinants among enterobacte-rial isolates, it is necessary to elucidate the AmpC re-sistance mechanism. The present study was conductedto determine the prevalence and genotypes of AmpC�-lactamases from clinical isolates at a university hospital inKorea.

0924-8579/$ – see front matter © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.doi:10.1016/j.ijantimicag.2004.03.023

J.Y. Kim et al. / International Journal of Antimicrobial Agents 24 (2004) 320–326 321

2. Materials and methods

2.1. Clinical isolates and bacterial strains

Nine isolates ofE. coli, producing CMY-11, were col-lected at Kosin Medical Centre (Pusan, Korea) during a12-month period from June 1998 to May 1999, and theirprofiles are presented inTable 1. These isolates were iden-tified by conventional techniques[5] and/or Vitek GNI card(bioMérieux Vitek Inc., Hazelwood, MO, USA).E. coli J53AzideR was used as a recipient strain for transfer exper-iments by transconjugation[6]. E. coli ATCC 25922 wasused as a MIC reference strain.E. coli BL21 (DE3) pLysSwas used as an expression host strain.

2.2. Susceptibility toβ-lactams and isoelectric focusing(IEF) analysis

Susceptibility was determined on Mueller-Hinton agarplates (Difco Laboratories, Detroit, MI, USA) containingserially two-fold-diluted�-lactams as previously described[4]. These results were interpreted by using the NationalCommittee for Clinical Laboratory Standards (NCCLS) cri-teria [7].

IEF was performed in Ready Gel Precast IEF Polyacry-lamide Gel (Bio-Rad, Hercules, CA, USA) as previouslydescribed[4]. Gels were developed with 0.5 mM nitrocefin(Merck, Whitehouse Station, USA). Inducibility of AmpC�-lactamases was inferred from the intensity of IEF patternsfor induced and uninduced�-lactamase extracts by cefoxitin(50 mg/l).

2.3. Plasmid preparation and Southern blot analysis

Isolation of plasmid DNA from nine clinical isolates wasperformed as described by Sambrook and Russel[8]. Theprepared plasmids were separated in 1.0% agarose using aField Inversion Gel Electrophoresis (FIGE) Mapper Elec-trophoresis System (Bio-Rad, Hercules, CA, USA). Theplasmids were purified with a Gel Extraction Kit (Genomid,Research Triangle Park, NC, USA). The purified plasmidswere used as the source of template DNA for PCR ampli-fication.

Transfer of DNA to nylon membrane (Hybond-N; Amer-sham International, Buckinghamshire, England) was per-formed essentially as described by Sambrook and Russel[8].Labelling of DNA (PCR product between C1 primer and C2primer) probe was performed with digoxigenin as describedby the manufacturer (Boehringer Mannheim Biochemicals,Indianapolis, USA). Hybridisation was performed at 68◦Cwith buffers recommended in the instructions included inthe digoxigenin kit from Boehringer.

2.4. Transconjugation experiments

Transconjugation experiments were performed as de-scribed previously[4] with sodium azide-resistantE. coli

J53 AzideR as a recipient. Transconjugants were selectedon Muller-Hinton agar supplemented with sodium azide(Sigma, Louis, MO, USA) (150 mg/l) to inhibit the growthof the donor strain and cefoxitin (20 mg/l) to inhibit thegrowth of the recipient strain.

2.5. PCR amplification and DNA sequencing

The primers for PCR amplification were designed byselecting consensus sequences in multiple nucleotide align-ment of 60 TEM-type�-lactamase genes (blaTEM), 27SHV-type �-lactamase genes (blaSHV), and 5 CMY-type�-lactamase genes (blaCMY) using the Primer Calculatorprogram (Williamstone Enterprises, Waltham, MA, USA).The primers have been described previously[9,10]: T1and T2 for blaTEM; C1, C2, CMYF1 and CMYR1 forplasmid-encodedblaCMY. PCR amplifications, DNA se-quencing, and DNA sequence analysis were carried out asdescribed previously[9,10].

Enterobacterial repetitive consensus (ERIC)-PCRs wereperformed in 50�l volumes containing 10 ng of genomicDNA from six clinical isolates, 4 mM MgCl2, 50 pM of eachprimer: ERIC1R (5′-ATGTAAGCTCCTGGGGATTCAC-3′)and ERIC2 (5′-AAGTAAGTGACTGGGGTGAGCG-3′)[11], 1.25 U of TaKaRa ExTaqpolymerase (TaKaRa, Otsa,Shiga, Japan), 0.2 mM of each dATP, dCTP, dGTP, and dTTPin 25 mM TAPS (N-Tris(hydroxy)methyl-3-amino-propanesulfonic acid, pH 9.3), 50 mM KCl, and 1 mM 2-mercapto-ethanol. Amplification was carried out using the follow-ing programme: 95◦C for 5 min followed by 35 cyclesof 1 min at 52◦C, 5 min at 70◦C, and 1 min 92◦C. Thefinal extension step was performed at 70◦C for 10 min.The analysis of amplified products (10�l aliquots) wasperformed in 2% Seakem LE agarose (BMA, Rockland,ME, USA). Macrorestriction analysis ofXbaI-digestedchromosomal DNA was done by pulsed-field gel elec-trophoresis (PFGE) by the instruction of Bio-Rad andfragments were separated for 20 h at 6 v/cm at 11◦C us-ing a CHEF-DR II System (Bio-Rad), with initial andfinal pulse times of 0.5 and 30 s, respectively. DNA finger-prints were interpreted as recommended by Tenover et al.[12].

2.6. Cloning of blaCMY-11 genes

Plasmid DNAs prepared from six clinical isolates wereused as template DNA in polymerase chain reaction (PCR).Oligonucleotides N-NdeI (5′-GTAGACCATATGCAACAA-CGACAATCC-3′) and C-XhoI (5′-GAATGTCTCGAG-CTCTTTCTTTCAACC-3′) were used as forward andbackward primers, respectively. Both primers containeda tail with NdeI (N-NdeI) and XhoI (C-XhoI) recognitionsequences (in bold type). All restriction enzymes werepurchased from Roche Applied Science (Mannheim, Ger-many). Since N-NdeI and C-XhoI primer were present in

322J.Y

.K

ime

ta

l./Inte

rna

tion

al

Jou

rna

lo

fA

ntim

icrob

ial

Age

nts

24

(20

04

)3

20

–3

26

Table 1Profiles of nineE. coli isolates, their transformants (trf) and host (E. coli BL21 [DE3] pLysS) for transformation

Strain Age of patient(year)/sexa

Type ofspecimen

Wardb Type ofinfectionc

MICs (mg/l) of �-lactamsd pI

AMX AUG LOT TAX TAZ ATM FEP IMP MER FOX TTN

K98678.3 40/F Pus OS Wound 128 128 >256 16 16 8 0.25 0.5<0.06 >256 256 8.0trf K98678.3 64 64 256 8 8 4 0.25 0.25 <0.06 128 128 8.0K981976.1 40/F Pus ENT Wound 128 128 >256 8 8 8 0.25 0.5<0.06 >256 256 8.0trf K981976.1 64 64 256 4 4 4 0.25 0.5 <0.06 128 128 8.0K982516.4 69/M Pus GS Wound 256 256 >256 8 16 16 0.25 0.5<0.06 >256 256 5.4, 8.0trf K982516.4 128 128 >256 8 16 8 0.125 0.25<0.06 256 256 8.0K983179.2 40/F Bed sore ENT Bed sore 256 256 >256 8 16 16 0.25 0.5<0.06 >256 256 8.0trf K983179.2 128 128 256 8 16 8 0.125 0.25<0.06 256 256 8.0K983806 71/F Sputum Neurology Pneumonia 256 256 >256 16 16 16 0.25 0.25 0.125 >256 256 8.0trf K983806 128 128 256 8 16 8 0.25 0.25<0.06 256 256 8.0K983533 57/F Urine Neurology UTI 128 128 >256 8 16 8 0.25 0.25<0.06 >256 256 8.0trf K983533 128 128 256 4 8 4 0.125 0.25<0.06 128 128 8.0K986110 60/M Sputum NS Pneumonia 256 256 >256 16 16 8 0.25 0.5 0.125 >256 256 8.0trf K986110 128 128 256 8 8 4 0.25 0.25 <0.06 128 128 8.0K9815491 70/M Urine Neurology UTI 128 128 >256 16 16 16 0.5 0.25<0.06 >256 256 5.4, 8.0trf K9815491 128 128 256 8 16 8 0.25 0.25<0.06 256 128 8.0K9816129 70/M Urine Neurology UTI 128 128 >256 16 16 16 0.25 0.25 0.125 >256 256 5.4, 8.0trf K9816129 64 64 256 8 16 8 0.25 0.25 <0.06 256 128 8.0BL21 (DE3) pLysS 1 1 1 <0.06 <0.06 <0.06 0.06 0.25 <0.06 <0.06 <0.06

a Abbreviation: M, male; F, female.b Abbreviation: OS, orthopaedic surgery; ENT, otolaryngology; GS, general surgery; NS, neonate surgery.c Abbreviation: UTI, urine tract infection.d Abbreviation: AMX, amoxicillin; AUG, amoxicillin-clavulanic acid; LOT, cephalothin; TAX, cefotaxime; TAZ, ceftazidime; ATM, aztreonam; FEP, cefepime; IMP, imipenem; MER, meropenem;

FOX, cefoxitin; TTN, cefotetan.

J.Y. Kim et al. / International Journal of Antimicrobial Agents 24 (2004) 320–326 323

the origin and in the end of structural genes, respec-tively, all regulatory signals from the original genes wereeliminated by the cloning strategy. PCR amplificationswere carried out on a DNA thermal cycler (mod. 2400;Perkin-Elmer Cetus, Norwalk, CT, USA) as previously de-scribed[4]. The expected PCR product of 1179 bp was veri-fied by agarose gel electrophoresis. TheNdeI/XhoI-digestedPCR products were ligated with theNdeI/XhoI-digestedvector pET-30a(+) (Novagen, Wisconsin, WI, USA).The ligation mixture was introduced by transformationinto competentE. coli BL21 (DE3) pLysS cells (No-vagen). Transformants were selected onto Luria-Bertani(Difco Laboratories) agar plates containing chlorampheni-col (25 mg/l) and kanamycin (50 mg/l). The presence ofthe desired hybrid plasmid was purified by WizardPlusMinipreps DNA Purification System (Promega, Madison,WI, USA) and verified by restriction analysis. Sequenc-ing of the whole clonedblaCMY-11 gene was performedby direct sequencing method with an automatic sequencer(mod. 373A; Applied Biosystems, Weiterstadt, Germany),as previously described[4]. Expression of blaCMY-11gene was induced by isopropyl-�-d-thiogalactopyranoside(IPTG, 0.5 mM).

2.7. Nucleotide sequence accession number

The blaCMY-11 gene nucleotide sequence data appear inthe GenBank nucleotide sequence database under the acces-sion number AF357599 forE. coli K98678.3, AF381619for E. coli K981976.1, AF381620 forE. coli K982516.4,AF381621 forE. coli K983179.2, AF381622 forE. coliK983533, AF381623 forE. coli K983806, AF381624 forE.coli K986110, AF381625 forE. coli K9815491, AF381626for E. coli K9816129.

3. Results

3.1. Properties of nine CMY-11-producing isolates

We previously obtained 51 clinical isolates from patientswith various infections[13]. Fifty-one isolates resistant tothird-generation cephalosporins (cefotaxime, ceftazidimeand ceftriaxone), monobactam (aztreonam) or cephamycins(cefoxitin and cefotetan) were selected from 710E. coliand 237K. pneumoniaeat Kosin Medical Centre. Nine offifty-one clinical isolates showed high levels of resistanceto amoxicillin, amoxicillin-clavulanic acid, cephalothin,cefoxitin and cefotetan (Table 1). The plasmid contentsof nine clinical isolates were analysed by FIGE and alarge plasmid (130 kb) was detected in all isolates (datanot shown). Nine isolates were mated withE. coli J53AzideR. Transconjugants of nine isolates were obtainedat a frequency of5× 10−5. All transconjugants containedonly a large plasmid (130 kb), designated pYMG-2. The130 kb plasmids from the nine isolates and their transcon-

jugants hybridised with labelled 847-bp PCR product be-tween C1 and C2 primer set specific for genes of CMY-1,FOX-1, FOX-2, FOX-3 and MOX-1. The�-lactamasesof nine isolates were not induced by cefoxitin. Isoelec-tric points (pIs) for nine isolates are shown inTable 1.Nine clinical isolates and their transformants producedthe �-lactamase with an apparent pI of 8.0. Three clinicalisolates (K982516.4, K9815491 and K9816129) also pro-duced another�-lactamase with an apparent pI of 5.4. PCRamplification of specific regions of plasmid pYMG-2 withprimer pairs was successful. 838-bp PCR product (betweenT1 and T2 primer) forblaTEM, 847-bp (between C1 and C2primer) and 1475-bp PCR product (between CMYF1 andCMYR1 primer) for blaCMY were obtained. Taking intoaccount the resistance phenotypes of nine clinical isolates,the resistance genotypes of these isolates were analysed bydirect sequencing of the PCR-amplified fragments specificfor blaTEM and blaCMY genes. On the basis of DNA se-quencing, nine clinical isolates and their transformants har-bouredblaCMY-11 gene. Three clinical isolates (K982516.4,K9815491 and K9816129) contained also anotherblaTEM-1gene.

3.2. ERIC-PCR patterns of E. coli isolates

Recently, ERIC-PCR generates a characteristic genomicfingerprinting which can be used to reveal intra- andinter-species genotypic variations among Enterobacteri-aceae[14]. All clinical isolates generated distinct bandsranging in size 0.1–4.2 kb and varied from 9 to 13 bands byERIC-PCR (Fig. 1). ERIC-PCR analysis of genomic DNAfrom the clinical isolates revealed five different types: A,B, C, D and E. In type A and E, each strain (K98678.3or K981976.1; K9815491 or K9816129) was isolated fromthe same patient (Table 1). Clinical isolates from differentpatients, except for K982516.4, K983179.2, K983802.1 andK983806, showed different fingerprint types. These resultswere confirmed by pulsed-field gel electrophoresis (PFGE)analysis withXbaI, showing that the differences amongeach clinical isolates in regard to number of bands weremore than 10 (data not shown).

3.3. DNA sequence analysis

The nucleotide sequence of thisblaCMY-11 gene inpYMG-2 differed from that ofblaCMY-1 in pMVP-1 [15]by two point mutations, T→ G at position 1197 and A→ T at position 1347, leading to amino acid substitutionIle → Ser at position 315 and Asn→ Ile at position 366,respectively. Compared with the upstream sequence ofthe blaCMY-1 gene, 18 nucleotides (TTTATACTTCCTAT-ACCC, I-18) between position 71 and 72 were inserted intothe blaCMY-11 gene in pYMG-2. I-18 was identical to 18nucleotides between position 54 and 71, indicating a directduplication of 18 nucleotides found upstream ofblaCMY-11gene. The initial 71-bp sequence before I-18 sequence of

324 J.Y. Kim et al. / International Journal of Antimicrobial Agents 24 (2004) 320–326

Fig. 1. ERIC-PCR patterns of genomic DNA fromE. coli isolates (K98678.3, lane 2; K981976.1, lane 3; K982516.4, lane 4; K983179.2, lane 5; K983806,lane 6; K983802.1[7], lane 7; K983533, lane 8; K986110, lane 9; K9815491, Lane 10; K9816129, lane 11). Lane 1 (HindIII/ EcoRI-digested phageλ)and 12 (100 bp stepwise ladder) show band patterns of DNA marker fragments (sizes in bp are indicated on the edge of the gel). Lanes 2, 3: type A;lanes 4–7: type B; lane 8: type C; lane 9: type D; lane 10, 11: type E. ERIC-PCR was performed with primers ERIC2 and ERIC1R.

blaCMY-11 was identical as a conserved integron sequence(2.1-kb so-called common region) previously identified inthe In7 integron from pDGO100 (GenBank accession num-ber L06822), In6 integron from pSa (GenBank accessionnumber L06418) and a novel integron from pSAL-1 (Gen-Bank accession number AJ237702). The common regionwas always located downstream of the truncated 3′-CS(conserved sequence) of the integrons and ended at thesame site approximately 0.2-kb downstream of the 3′ endof orf513 (orf341 in pSAL-1), which encodes a putativetransposase[16] (Fig. 2).

Fig. 2. Comparison of the regions between the common regions of In6, In7, pSAL-1 (DHA-1), pRMOX-1 (MOX-1), pMVP-1 (CMY-1), pYMG-1(CMY-10), and pYMG-2 (CMY-11) and the second 3′-CS (conserved sequence) of In6, In7, and pSAL-1 (DHA-1). Asterisks indicate sequence identity.Horizontal arrows indicate a direct duplication of 18 bp found in pYMG-2. Vertical arrow indicates 3′ end of the common region. GenBank (or EMBL)accession numbers are as follows: pSa, L06418; pDGO100, L06822; pSAL-1, AJ237702; pRMOX-1, D13304; pMVP-1, X92508; pYMG-1, AF357598;pYMG-2, AF381619.

4. Discussion

Nine clinical isolates and their transformants were resis-tant to cefoxitin (MIC ≥ 32 mg/l) and cefotetan (MIC≥32 mg/l), and produced a transferable�-lactamase with pIof 8.0. The MIC of amoxicillin was not reduced by clavu-lanic acid in the nine clinical isolates, which probably cor-responded to MIC patterns of strains producing Bush group1 (class C) AmpC�-lactamases. On the basis of the pIsof �-lactamases, their MIC patterns, induction by cefox-itin and DNA sequencing, the�-lactamase with a pI of 8.0

J.Y. Kim et al. / International Journal of Antimicrobial Agents 24 (2004) 320–326 325

was CMY-11. The�-lactamase with pIs of 5.4 was TEM-1.These�-lactamases had the same pIs as those previouslyreported (TEM-1[17] and CMY-11[4]).

ERIC-PCR patterns suggested that isolates from the samepatient were genetically identical, except for K983179.2,K983802.1, and K983533, indicating resistance againstcephamycins spread due to outbreaks of resistant species.Clinical isolates from different patients, except forK982516.4, K983179.2, K983802.1 and K983806, were ge-netically unrelated, indicating resistance spread also due tothe dissemination of�-lactamases (TEM-1 and CMY-11).The nine clinical isolates were from urine, pus and sputa(Table 1), indicating that these were important potentialsources of spread.

The sequence (nucleotides 1–253) precedingblaCMY-10[9] andblaCMY-11, less a 15 nucleotide (I-15) and an 18 nu-cleotide duplication (I-18), respectively, is identical to thatupstream fromblaCMY-1 [15]. All putative gene expressionsignals are preserved. The truncated sequence preceding theI-15 or I-18 (nucleotides 1–71) has been identified as part ofthe unusual class 1 integrons In6 and In7[18] and the novelintegron on pSAL-1[19]. In all cases the homology is toa sequence in the modified 3′-conserved sequence of theseintegrons and is lost abruptly after the I-15 or I-18 sequence[19], duplicated or not. This finding indicates insertion atthis point (I-15 or I-18) of different sequences, encoding adihydrofolate reductase (dfrA10, In7) [18], a chlorampheni-col acetyl transferase (catA2, In6) [18], and variousampCgenes (blaDHA-1, pSAL-1; blaCMY-1, pMVP-1; blaMOX-1,pRMOX-1; blaCMY-10, pYMG-1 [9]; blaCMY-11, pYMG-2)[19], suggesting that this is a secondary locus for gene cap-ture by integrons such as In6 and In7. Although the varietyof sequences that have been inserted, apparently at the samepoint, is reminiscent of the different gene cassettes found inthe standard variable insert region of integrons[20,21], thereis no evidence that 59 base elements are needed for inser-tion at the secondary site. This would suggest involvementof another site-specific, integron associated recombinationsystem that is able to capture antibiotic resistance genes onto mobile DNA structures and so promote their dissemina-tion among bacteria of clinical importance. Compared withCMY-1 [15], CMY-10 [9] has a single amino acid substi-tution, Asn366Ile, while CMY-11 has two (Ile315Ser andAsn366Ile). CMY-1, CMY-10 and CMY-11 were all foundin Korean clinical isolates. A likely evolutionary sequenceis CMY-1 → CMY-10 → CMY-11. TheblaCMY-10 geneharboured I-15 andblaCMY-11 gene contained I-18. Theseresults indicate that the insertion ofblaCMY gene into asul1-type integron (In6, In7, or pSAL-1) by I-15 may evolveinto that ofblaCMY gene by I-18.

Acknowledgements

This work was supported by a grant from BioGreen 21Program, Rural Development Administration, Republic ofKorea.

References

[1] Sanders CC, Sanders Jr WE. Beta-lactamase in gram-negative bac-teria: global trends and clinical impact. Clin Infect Dis 1992;15:824–39.

[2] Bush K, Jacoby GA, Medeiros AA. A functional classificationscheme for beta-lactamases and its correlation with molecular struc-ture. Antimicrob Agents Chemother 1995;39:1211–33.

[3] Marchese A, Arlet G, Schito GC, Lagrange PH, Philippon A. Charac-terization of FOX-3, an AmpC-type plasmid-mediated beta-lactamasefrom an Italian isolate ofKlebsiella oxytoca. Antimicrob AgentsChemother 1998;42:464–7.

[4] Lee SH, Jeong SH, Park Y-M. Characterization ofblaCMY-10, anovel, plasmid-encoded AmpC-type beta-lactamase gene in a clinicalisolate of Enterobacter aerogenes. J Appl Microbiol 2003;95:744–52.

[5] Farmer III JJ. Enterobacteriaceae: introduction and identification. In:Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, editors.Manual of clinical microbiology. Washington, DC: American Societyfor Microbiology; 1995.

[6] Kim J, Kwon Y, Pai H, Kim JW, Cho DT. Survey ofKlebsiellapneumoniaestrains producing extended-spectrum beta-lactamases:prevalence of SHV-12 and SHV-12a in Korea. J Clin Microbiol1998;36:1446–9.

[7] National Committee for Clinical Laboratory Standards. Performancestandards for antimicrobial susceptibility testing. M100-S9. 9th ed.Wayne, PA: National Committee for Clinical Laboratory Standards;1999.

[8] Sambrook J, Russel DW. Extraction and purification of plasmid;screening of bacterial colonies by hybridization. In: Argentine J,Irwin N, Jassen KA, Zierler SCM, Mclnerny N, Brown D, SchaeferS, editors. Molecular cloning: a laboratory manual. New York: ColdSpring Harbor Laboratory Press; 2001.

[9] Lee SH, Kim JY, Lee GS, et al. Characterization ofblaCMY-11, anAmpC-type plasmid-mediated beta-lactamase gene in a Korean clini-cal isolate ofEscherichia coli. J Antimicrob Chemother 2002;49:269–73.

[10] Lee SH, Kim JY, Lee SK, Jin W, Lee KJ. Discriminatory detec-tion of extended-spectrum beta-lactamases by restriction fragmentlength dimorphism-polymerase chain reaction. Lett Appl Microbiol2000;31:307–12.

[11] Versalovic J, Koeuth T, Lupski JR. Distribution of repetitive DNAsequences in eubacteria and application to fingerprinting of bacterialgenomes. Nucleic Acids Res 1991;19:6823–31.

[12] Tenover FC, Arbeit RD, Goering RV, et al. Interpreting chromosomalDNA restriction patterns produced by pulsed-field gel electrophore-sis criteria for bacterial strain typing. J Clin Microbiol 1995;33:2233–9.

[13] Lee SH, Jeong SH. Antibiotic susceptibility of bacterial strains iso-lated from patients with various infections. Lett Appl Microbiol2002;34:215–21.

[14] Della-Costa LM, Irino K, Rodrigues J, Rivera ING, TrabulsiLR. Characterization of diarrhoeagenicEscherichia coli clonesby ribotyping and ERIC-PCR. J Med Microbiol 1998;47:227–34.

[15] Bauernfeind A, Stemplinger I, Jungwirth R, Wilhelm R, Chong Y.Comparative characterization of the cephamycinaseblaCMY-1 geneand its relationship with other beta-lactamase genes. AntimicrobAgents Chemother 1996;40:1926–30.

[16] Villa L, Visca P, Tosini F, Pezzella C, Carattoli A. Composite integronarray generated by insertion of an ORF341-type integron within aTn21-like element. Microb Drug Resist 2002;8:1–8.

[17] Pai H, Lyu S, Lee JH, et al. Survey of extended-spectrumbeta-lactamases in clinical isolates ofEscherichia coliand Kleb-siella pneumoniae: prevalence of TEM-52 in Korea. J Clin Microbiol1999;37:1758–63.

326 J.Y. Kim et al. / International Journal of Antimicrobial Agents 24 (2004) 320–326

[18] Strokes HW, Tomaras C, Parsons Y, Hall RM. The partial3′-conserved segment duplications in the integrons In6 from pSa andIn7 from pDGO100 have a common origin. Plasmid 1993;30:39–50.

[19] Verdet C, Arlet G, Barnaud G, Lagrange PH, Philippon A. Anovel integron in Salmonella entericaserover Enteritidis, carry-ing the bla(DHA-1) gene and its regulator gene ampR, originated

from Morganella morganii. Antimicrob Agents Chemother 2000;44:222–5.

[20] Hall RM, Collis CM. Mobile gene cassettes and integrons: captureand spread of genes by site-specific recombination. Mol Microbiol1995;15:593–600.

[21] Recchia GD, Hall RM. Gene cassettes: a new class of mobile element.Microbiology 1995;141:3015–27.