- Email: [email protected]
Plasmid-mediated, inducible AmpC β-lactamase (DHA-1)-producing Enterobacteriaceae at a Korean hospital: wide dissemination in Klebsiella pneumoniae and Klebsiella oxytoca and emergence in Proteus mirabilis
Diagnostic Microbiology and Infectious Disease 53 (2005) 65 – 70 www.elsevier.com/locate/diagmicrobio
Plasmid-mediated, inducible AmpC h-lactamase (DHA-1)-producing Enterobacteriaceae at a Korean hospital: wide dissemination in Klebsiella pneumoniae and Klebsiella oxytoca and emergence in Proteus mirabilis Dongeun Yonga,c,d, Youngsik Limc,d, Wonkeun Songc,e, Yeoung Seon Choic,d, Doe-Young Parkc,d, Hyukmin Leea,c, Jong Hwa Yumc,d, Kyungwon Leea,c,d,T, June Myung Kimb,c,d, Yunsop Chonga,c a
Department of Laboratory Medicine, Yonsei University College of Medicine, Seoul 120-752, South Korea b Department of Internal Medicine, Yonsei University College of Medicine, Seoul 120-752, South Korea c Research Institute of Bacterial Resistance, Yonsei University College of Medicine, Seoul 120-752, South Korea d BK21 Project for Medical Sciences, Yonsei University College of Medicine, Seoul 120-752, South Korea e Department of Laboratory Medicine, University of Hallym College of Medicine, Seoul 150-072, South Korea Received 9 December 2004; accepted 25 March 2005
Abstract The aim of the study was to investigate the phenotypic and genetic characteristics of recently emerging cefoxitin-resistant and inductionpositive isolates of Escherichia coli, Klebsiella species, and Proteus mirabilis. Strains of Enterobacteriaceae were isolated at a Korean tertiary care hospital between June and December 2002. Induction was tested using cefoxitin and aztreonam disks, the bla DHA allele was detected by PCR, and pulsed-field gel electrophoresis (PFGE) patterns were also analyzed. Among the cefoxitin-resistant isolates, 2.7% of E. coli, 21.1% of Klebsiella pneumoniae, 32.0% of Klebsiella oxytoca, and 8.3% of P. mirabilis isolates showed induction, and were bla DHA-1 allele positive. To the best of our knowledge, this is the first report of bla DHA-1 in P. mirabilis. The MICs of ceftazidime, cefotaxime, and aztreonam increased significantly by higher inoculum, suggesting that their clinical usefulness is limited. Presence of multiple PFGE patterns and identical patterns in some isolates suggest that the widely disseminated bla DHA-1 in Klebsiella species was because of both horizontal and clonal spread. D 2005 Elsevier Inc. All rights reserved. Keywords: DHA-1; AmpC h-lactamase; Proteus mirabilis; Korea
1. Introduction h-Lactamase production is the predominant mechanism for resistance to h-lactam antibiotics in Gram-negative bacilli. Resistance due to extended-spectrum h-lactamase (ESBL) appeared in Klebsiella pneumoniae and in other enterobacterial species, but they remained susceptible to cephamycins (Philippon et al., 2002). Plasmid-mediated AmpC h-lactamases were first detected in 1988 (Bauernfeind et al., 1989, 1998; Papanicolaou
T Corresponding author. Department of Laboratory Medicine, Yonsei Universtiy College of Medicine, Seoul 120-752, South Korea. Tel.: +82-2228-2446; fax: +82-2-313-0908. E-mail address: [email protected] (K. Lee). 0732-8893/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2005.03.008
et al., 1990), and subsequently various AmpC enzymes have been found, particularly in Klebsiella spp., Escherichia coli, and Proteus mirabilis (Paterson et al., 2003; Philippon et al., 2002). The majority of plasmid-mediated AmpC enzymes are expressed constitutively, with the exceptions of DHA, ACT-1, and CFE-1. DHA-1 was first reported in 1992 in Salmonella enteritidis in Saudi Arabia and France (Gaillot et al., 1997), DHA-2 in K. pneumoniae in France (Fortineau et al., 2001), ACT-1 in K. pneumoniae in the United States (Reisbig and Hanson, 2002), and CFE-1 in Citrobacter freundii in Japan (Nakano et al., 2004). In Korea, cefoxitin-resistant E. coli and K. pneumoniae has been increasingly noted. A recent investigation detected 15 isolates with undetermined bla DHA subtype (Song et al.,
66
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
2002). Recently, we have increasingly noted presence of induction-positive isolates of K. pneumoniae at a tertiary care hospital laboratory. Some of the plasmid-mediated AmpC enzyme-producing isolates are susceptible to third-generation cephalosporins and aztreonam in vitro (Philippon et al., 2002), but their clinical utility is uncertain. As for DHA-producing organisms, even in vitro susceptibility data are limited, because only a few studies have been reported (Alvarez et al., 2004; Yan et al., 2002). In addition to the treatment problem, accurate differentiating methods for plasmid-mediated AmpC h-lactamase-producing isolates are required to improve surveillance and infection control (Perez-Perez and Hanson, 2002), but the current guidelines from the National Committee for Clinical Laboratory Standards (NCCLS, 2004) do not include any recommendation. Recently, Perez-Perez and Hanson (2002) recommended the use of multiplex PCR to identify family-specific AmpC h-lactamase genes. The aim of this study was to determine the phenotypic and genetic characteristics of induction-positive strains of E. coli, Klebsiella spp., and P. mirabilis isolated at a Korean hospital. We also investigated the susceptibility, inoculum effect on the minimum inhibitory concentrations of antimicrobial agents, and the molecular epidemiological features of bla DHA-1 allele-positive strains. 2. Materials and methods 2.1. Screening of inducible AmpC b-lactamase– and ESBL-producing isolates Bacterial strains were isolated from patients at a Korean tertiary care hospital between June and December 2002. Species were identified by conventional methods or by using the Vitek GNI system (bioMe´rieux, Marcy l’Etoile, France). Cefoxitin-resistant E. coli, Klebsiella spp., and P. mirabilis isolates, detected by the disk diffusion method (NCCLS, 2002) using commercial disks and MuellerHinton agar (Becton Dickinson, Cockeysville, MD), were tested for induction by applying cefoxitin and aztreonam disks 15 mm apart from edge to edge (Yan et al., 2002). A presence of truncated inhibition zone of aztreonam disk was interpreted as induction positive. ESBL producers were detected the double-disk synergy test using amoxicillin-clavulanic acid, cefotaxime, ceftazidime, and cefepime disks at a distance of 15 mm from disk edge to edge. Induction-positive isolates were stored in skim milk at 708C until used for further tests. 2.2. PCR protocol and nucleotide sequencing bla DHA alleles were detected by PCR using 1 AL (20 pmol) of each of the following primers: DHA-F (5V-ATG AAA AAA TCG TTA TCT GCA A-3V) and DHA-R (5V-CTA TTT TGA GTG CAC TGG AAT AA-3V), which amplify 1140 bp of the bla DHA gene, 1 AL of heat-
extracted template, and PreMix (Bioneer, Cheongwon, Korea) in a total volume of 20 AL. bla DHA allele-positive isolates were further tested: for bla DHA-1, using DHA-1/2 F (5V-TCT GCC GCT GAT AAT GTC G-3V), which is common to both bla DHA-1 and bla DHA-2, and DHA-1 R (5V-GCC GCC GGA TCA TTC AGC GC-3V); for bla DHA2, using DHA-1/2 F and DHA-2 R (5V-TCT GCC GGG TCA TTC AAC AT-3V), respectively. For nucleotide sequencing, PCR products were amplified in a mixture of 1 AL (20 pmol) of each of DHA-1 X-F (5V-AAC GTC TGA CCA TAA TCC AC-3V) and DHA-1 X-R (5V-GCC CAT AAA GCA AAT TAT TG-3V), and 0.5 AL of 2.5 U ExTaq DNA polymerase (Takara, Otsu, Japan) in total volume of 100 AL. DNAs were extracted from the PCR products using a commercial kit (Qiagen, Hilden, Germany), and were sequenced using the primers DHA-1 X-F, DHA-1 XR, DHA-1 Lt-F (5V-GCT GGG GTT ATC TCA CAC CT-3V) and DHA-1 Rt-R (5V-CCG TAC GCA TAC TGG CTT TGC-3V) by the dideoxy-chain termination method using an ABI 3700 DNA sequencer (Perkin-Elmer, Foster City, CA). The PCR conditions used to detect and to sequence bla DHA, bla DHA-1, and bla DHA-2 consisted of 30 cycles denaturation at 948C for 30 s, annealing at 508C (to detect bla DHA and sequence), at 588C (to detect bla DHA-1), or at 558C (to detect bla DHA-2) for 30 s, and extension at 728C for 50 s followed by a final extension at 728C for 7 min. Class 1 integron was detected by PCR as described previously (Riccio et al., 2000). E. coli strain (YMC 02/8/U310), which harbors bla DHA-1, and Pseudomonas aeruginosa (YMC 95/1/704), which harbors class 1 integron, were used as a positive control strains for DHA and class 1 integron PCR, respectively. 2.3. MIC determination The NCCLS agar dilution method (NCCLS, 2002) was used. The antimicrobial agents used were amoxicillin (Kun Wha Pharmaceuticals, Seoul, Korea), ceftazidime and clavulanic acid (GlaxoSmithKline, Greenford, UK), piperacillin (Samsung Pharmaceuticals, Seoul, Korea), tazobactam (Wyeth, Pearl River, NY), cephalothin (Sigma, St. Louis, MO), cefoxitin and imipenem (Merck Sharp & Dohme, West Point, PA), cefotaxime (Handok Pharmaceuticals, Seoul, Korea), and aztreonam and cefepime (Bristol-Myers Squibb, Princeton, NJ). Constant concentrations (4 Ag/mL) of tazobactam or clavulanic acid were added to piperacillin, ceftazidime, cefotaxime, aztreonam, or cefepime. The ratio of amoxicillin to clavulanic acid used was 2:1. To determine the inoculum effect, approximate inocula of 104 and 106 CFU per spot were applied using a Steers replicator (Craft Machine, Chester, PA). To exclude the effect of coexistent ESBL, we did not include the ESBL-producing isolates. E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used for quality control.
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70 Table 1 bla DHA allele–positive isolates among E. coli, Klebsiella spp., and P. mirabilis Species (no. of isolates tested)
No. (%) of isolates with Cefoxitin resistance
Induction positivea
bla DHA positive by PCRb
E. coli (1937) K. pneumoniae (899) K. oxytoca (149) P. mirabilis (124) Total (3109)
291 (15.0) 261 (29.0)
8 (2.7) 55 (21.1)
8 55
2 6
25 (16.8) 12 (9.7) 589 (18.9)
8 (32.0) 1 (8.3) 72 (12.2)
8 1 72
4 1 13
bla DHA-1 sequencedc
a
Cefoxitin-resistant isolates only were tested using cefoxitin and aztreonam disks. b Antagonism-positive isolates only were tested. c Randomly selected bla DHA allele-positive isolates were tested.
2.4. Transfer of resistance, plasmid analysis, and Southern hybridization The transferability of inducible cefoxitin resistance was tested by the agar mating method using randomly selected isolates and azide-resistant E. coli J53 recipient. MacConkey agar supplemented with 100 Ag/mL of sodium azide, and 16 Ag/mL of cefoxitin was used to select transconjugants. Plasmids were isolated from clinical isolates and their transconjugants using the alkaline lysis method (Sambrook and Russell, 2001), and the size was estimated by comparing their electrophoretic mobilities with those of E. coli V517. Southern hybridization was used to determine the location of bla DHA-1 on the plasmid, chromosome, or integron. A bla DHA-1 probe was prepared using a DIG DNA Labeling and Detection Kit (Roche Diagnostics, Mannheim, Germany), and testing was performed according to the manufacturer’s instructions.
67
2.5. Pulsed-field gel electrophoresis XbaI- and I-CeuI-digested genomic DNA was prepared according to the instruction of BioRad (Hercules, CA), and fragments were separated for 20 h at 6 V/cm at 118C using a CHEF-DR II System (BioRad), with initial and final pulse times of 0.5 and 30 s, respectively. Pulsedfield gel electrophoresis (PFGE) banding type similarities were determined using the Dice coefficient and unweighted pair group method using UVIBand/Map software (UVItech, Cambridge, UK). Possibly, related isolates with 6 band differences were identified by using 80% similarity coefficients. 2.6. Nucleotide sequence accession number The nucleotide sequences of bla DHA-1 in 4 clinical isolates have been assigned to the following GenBank accession numbers: E. coli YMC 02/08/U310 (AY205598), K. pneumoniae YMC 02/09/R340 (AY205600), Klebsiella oxytoca YMC 02/09/P206 (AY205599), and P. mirabilis YMC 02/08/U1255 (AY205601).
3. Results 3.1. Detection of DHA-producing isolates Of the 3,109 unduplicated isolates of E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis tested, 15.0%, 29.0%, 16.8%, and 9.7%, respectively, were resistant to cefoxitin (Table 1). Of the cefoxitin-resistant isolates, induction was observed in 8 of 291 (2.7%) E. coli, 55 of 261 (21.1%) K. pneumoniae, 8 of 25 (32.0%) K. oxytoca, and 1 of 12 (8.3%) P. mirabilis isolates. bla DHA alleles were detected by PCR in all of the 72 induction-positive isolates. All of these isolates showed
Table 2 Agar dilution antimicrobial susceptibilities of bla DHA-1 allele-positive isolates of E. coli, Klebsiella spp., and P. mirabilis depending on inoculum size Antimicrobial agents
MIC range (Ag/mL) with approximate inoculum (CFU/spot) E. coli (n = 6)
Amoxicillin Amoxicillin-clavulanic acid Piperacillin Piperacillin-tazobactam Cephalothin Cefoxitin Ceftazidime Ceftazidime-clavulanic acid Cefotaxime Cefotaxime-clavulanic acid Aztreonam Aztreonam-clavulanic acid Cefepime Cefepime-clavulanic acid Imipenem
K. pneumoniae (n = 18)
K. oxytoca (n = 5)
P. mirabilis (n = 1)
104
106
104
106
104
106
104
106
N128 32 64 – N256 1–2 N128 32– N128 1–16 4 –32 1– 4 1– 4 0.5 – 4 1– 8 0.03 – 0.06 V 0.008 – 0.06 0.12 – 0.25
N128 N128 N 256 N128 N128 128 – N128 16 – N128 NT 64 – N128 NT 128 – N128 NT NT NT NT
N128 32 – 64 64 – N 256 2– N 256 N128 128 – N128 4 – N128 16 – N128 1– 64 4 – 64 0.5 –128 4 – 64 0.03 – 0.5 0.06 – 0.5 0.12 –1
N128 N128 N 256 N128 N128 128 – N128 64 – N128 NT N128 NT N128 NT NT NT NT
N128 32 32– N 256 1– N 256 N128 64 – N128 1–32 4 – N128 1– 4 2–16 0.5 –16 1–16 0.03 – 0.06 0.015 – 0.12 0.25 – 0.5
N128 N128 N 256 N128 N128 128– N128 64 – N128 NT 64 – N128 NT N128 NT NT NT NT
N128 128 2 V 0.5 N128 64 4 32 0.5 32 0.015 1 0.06 0.25 2
N128 N128 N 256 N128 N128 N128 N128 NT N128 NT N128 NT NT NT NT
The MICs of ESBL-producing isolates were not included.
68
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
Table 3 Agar dilution antimicrobial susceptibilities of bla DHA-1-positive clinical isolates and their transconjugants Antimicrobial agents
Amoxicillin-clavulanic acid Piperacillin Piperacillin-tazobactam Ceftazidime Ceftazidime-clavulanic acid Cefotaxime Cefotaxime-clavulanic acid Aztreonam Aztreonam-clavulanic acid
MIC (Ag/mL)a E. coli YMC 02/8/U310
K. pneumoniae YMC 02/10/R340
K. oxytoca YMC 02/9/P206
Wild
Transconjugant
Wild
Transconjugant
Wild
Transconjugant
32 N256 2 8 16 4 4 1 4
64 32 1 2 16 0.5 2 0.5 4
32 64 4 64 N128 4 16 8 16
64 16 1 4 16 1 2 1 4
32 128 2 2 4 1 2 1 4
64 32 1 4 16 1 2 2 4
a The MICs (Ag/mL) of other antimicrobial agents for all 3 isolates and their transconjugants were the following: amoxicillin, N128 and N128; cephalothin, N128 and z32; cefoxitin, z128 and z64; cefepime, 0.03 – 0.06 and 0.015 – 0.03; cefepime-clavulanic acid, 0.015 – 0.06 and V 0.008 – 0.12; and imipenem, 0.25 – 0.5 and 0.12 – 0.25, respectively.
positive PCR reaction with bla DHA-1 allele-specific primers, but not with bla DHA-2 allele-specific primers. PCR products from 13 randomly selected isolates were used to determine nucleotide sequences, and all were found to be identical to that of bla DHA-1.
3.2. Resistance transfer and Southern hybridization Inducible cefoxitin resistance was transferred by conjugation from all of the randomly selected 2 isolates of E. coli, 1 K. pneumoniae, and 1 K. oxytoca. Repeated attempts failed to transfer resistance from 1 P. mirabilis isolate.
Fig. 1. A dendrogram of PFGE bands of XbaI-digested genomic DNA of K. pneumoniae based on the Dice coefficient. Twenty-seven types and five subtypes were recognized. Thirteen, eight, and five isolates of K. pneumoniae belonged to types B, G, and X, respectively. The numbers of isolates with the same band type are shown in parentheses.
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
The 4 clinical isolates and 4 transconjugants contained plasmids of N150 kb, which hybridized with the bla DHA-1 probe, but the I-CeuI-digested genomic DNA bands of 3 isolates and their transconjugants did not hybridize. Class 1 integrons were detected by PCR in 28 of 34 isolates (2 isolates of E. coli, 21 of K. pneumoniae, 4 of K. oxytoca, and 1 of P. mirabilis), but none of the amplicons hybridized with the bla DHA-1 probe. 3.3. Antimicrobial susceptibility of isolates and transconjugants The antimicrobial susceptibilities of the bla DHA-1positive clinical isolates excluding 2 ESBL-producing ones showed that the MICs of amoxicillin and cephalothin were N128 Ag/mL and those of cefepime and imipenem were V 2 Ag/mL for all isolates (Table 2). The MICs of ceftazidime, cefotaxime, and aztreonam were highly variable depending on isolates, and the levels increased significantly in the presence of clavulanic acid for most isolates, but those of piperacillin tended to decrease in the presence of tazobactam. When the inoculum was increased to approximately 106 CFU, all the MICs of piperacillin, ceftazidime, cefotaxime, and aztreonam increased significantly. Comparison showed that the MICs of piperacillin for 3 clinical isolates varied depending on species (i.e., N256 Ag/mL for E. coli, 64 Ag/mL for K. pneumoniae, and 128 Ag/mL for K. oxytoca), but those for their transconjugants were similar, 32, 16, and 32 Ag/mL, respectively (Table 3). 3.4. Epidemiology of DHA-1–producing isolates The majority of bla DHA-1 allele-positive strains were isolated from patients in general wards (56.9%), followed by patients in intensive care units (34.7%), and the sources were mostly sputum (45.8%) or urine (38.9%). PFGE patterns of the XbaI-digested genomic DNAs of the 54 isolates of bla DHA-1 allele-positive K. pneumoniae showed 27 different types, indicating multiclonality. However, 13, 8, and 5 isolates belonged to types B, G, and X, respectively (Fig. 1). All 8 isolates of E. coli and 7 of 8 K. oxytoca isolates showed different PFGE band patterns (data not shown).
4. Discussion Plasmid-mediated AmpC h-lactamase-producing gramnegative bacilli are being found increasingly in many parts of the world (Philippon et al., 2002), but reports on DHAproducing organisms are relatively rare. CMY-1-producing K. pneumoniae and E. coli have been prevalent and single amino acid-substituted CMY-1b (CMY-10) has emerged in Korea (Bauernfeind et al., 1998). Therefore, most of the plasmid-mediated AmpC h-lactamases in Korea was presumed to be these enzymes until other types were recently
69
reported (Song et al., 2002). It is interesting that bla DHA-1 allele-positive isolates of E. coli, K. pneumoniae, and K. oxytoca are widely disseminated at a Korean hospital. All of bla DHA allele-positive isolates in this study showed positive reaction only with bla DHA-1 allele-specific primers, and the bla DHA alleles in 13 randomly selected isolates showed a sequence identical to that of bla DHA-1 (Gaillot et al., 1997). We presume from this result that all bla DHA allele-positive isolates in our study carried the bla DHA-1 gene. As far as we are aware, this detection of bla DHA-1 in P. mirabilis is the first worldwide and indicates this species can also acquire the resistance. Plasmids carrying bla DHA-1 in K. pneumoniae and S. enteritidis (Gaillot et al., 1997; Yan et al., 2002) and those carrying bla DHA-2 in K. pneumoniae (Fortineau et al., 2001) have been reported to be ca. 70 –200 kb, and some of these were transferable by conjugation. In the present study, resistance was transferred by agar mating, and the isolates and transconjugants carried plasmids of variable size of N150 kb. It has been reported that the bla DHA-1 gene is located on a class 1 integron (Verdet et al., 2000); thus, the failure of our PCR product of the class 1 integron to hybridize with the bla DHA probe requires further investigation. The MICs for ESBL-producing isolates were not included in Table 2. We tested analytical isoelectric focusing and found none of them had h-lactamase band of pI N8.0, suggesting no coexistence of other AmpC enzymes in these strains. In another our study using multiplex PCR, it was shown that none of the isolates had more than one plasmid-mediated AmpC h-lactamase genes (unpublished data). Some bla DHA-1 allele-positive isolates were inhibited by low concentrations of ceftazidime, cefotaxime, or aztreonam (Table 2), but the clinical efficacies of these drugs are unknown. Controversy continues concerning the use of third-generation cephalosporins for the treatment of infections due to inducible chromosomal AmpC h-lactamaseproducing Enterobacteriaceae (Goldstein, 2002; Livermore et al., 2004). However, in view of the significant increase in MICs with higher inoculum or by the addition of clavulanic acid, third-generation cephalosporins with or without clavulanic acid may be inappropriate for the treatment of severe infections by DHA-1 producing isolates. As in other studies (Barnaud et al., 1998; Fortineau et al., 2001; Gaillot et al., 1997; Yan et al., 2002), all of the bla DHA-1 allelepositive isolates remained highly susceptible to cefepime and imipenem. Presence of multiple PFGE patterns of the XbaI-digested genomic DNAs of bla DHA-positive isolates of E. coli, K. pneumoniae, and K. oxytoca (Fig. 1) suggests that horizontal transfer of the gene-carrying plasmids occurs readily to genetically heterogeneous organisms. Some isolates with identical patterns indicated that clonal spread also played a role in the dissemination. As is in other nosocomial infections, the most frequent sources of the
70
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
bla DHA-1 allele-positive isolates were sputum and urine samples from hospitalized patients. The types of plasmid-mediated AmpC h-lactamase in E. coli and Klebsiella spp. cannot be differentiated by phenotype. AmpC multiplex PCR differentiated the 6 plasmid-mediated AmpC-specific families (Perez-Perez and Hanson, 2002). It has been reported that all DHAproducing isolates were induction test positive (Yan et al., 2002). In the present study, bla DHA alleles were detected in all 72 isolates showing induction. However, the sensitivity of the aztreonam and cefoxitin disk method is unknown, although it was more sensitive than tests using ceftazidime or cefotaxime (data not shown). Therefore, further studies are needed to determine the prevalence of bla DHA-1 allelepositive isolates in Korea. In conclusion, bla DHA-1 allele-positive isolates are common, especially in K. pneumoniae and K. oxytoca in Korea. bla DHA-1 was detected for the first time in P. mirabilis, and bla DHA-1 was found transferable by conjugation. A simple and accurate detection method for DHA-1–producing isolates is needed to control the spread of resistance. It is a concern that the DHA-1 h-lactamase gene may spread to various species of Enterobacteriaceae, against which all h-lactams, with the exception of carbapenem and fourth-generation cephalosporins, may be inactive. Acknowledgment This study was supported by the Yonsei University Research Fund (2003-35). References Alvarez M, Tran JH, Chow N, Jacoby GA (2004) Epidemiology of conjugative plasmid-mediated AmpC h-lactamases in the United States. Antimicrob Agents Chemother 48:533 – 537. Barnaud G, Arlet G, Verdet C, Gaillot O, Lagrange PH, Philippon A (1998) Salmonella enteritidis: AmpC plasmid-mediated inducible h-lactamase (DHA-1) with an ampR gene from Morganella morganii. Antimicrob Agents Chemother 42:2352 – 2358. Bauernfeind A, Chong Y, Schweighart S (1989) Extended broad spectrum h-lactamase in Klebsiella pneumoniae including resistance to cephamycins. Infection 17:316 – 321. Bauernfeind A, Chong Y, Lee K (1998) Plasmid-mediated AmpC h-lactamases: How far have we gone 10 years after the discovery? Yonsei Med J 39:520 – 525. Fortineau N, Poirel L, Nordmann P (2001) Plasmid-mediated and inducible cephalosporinase DHA-2 from Klebsiella pneumoniae. J Antimicrob Chemother 47:207 – 210.
Gaillot O, Clement C, Simonet M, Philippon A (1997) Novel transferable h-lactam resistance with cephalosporinase characteristics in Salmonella enteritidis. J Antimicrob Chemother 39:85 – 87. Goldstein FW (2002) Cephalosporinase induction and cephalosporin resistance: A longstanding misinterpretation. Clin Microbiol Infect 8: 823 – 825. Livermore DM, Brown DF, Quinn JP, Carmeli Y, Paterson DL, Yu VL (2004) Should third-generation cephalosporins be avoided against AmpC-inducible Enterobacteriaceae? Clin Microbiol Infect 10:84 – 85. Nakano R, Okamoto R, Nakano Y, Kaneko K, Okitsu N, Hosaka Y, Inoue M (2004) CFE-1, a novel plasmid-encoded AmpC h-lactamase with an ampR gene originating from Citrobacter freundii. Antimicrob Agents Chemother 48:1151 – 1158. National Committee for Clinical Laboratory Standards (2002) Performance standards for antimicrobial susceptibility testing: Twelfth informational supplement. Document M100-S12. Wayne, PA7 NCCLS. National Committee for Clinical Laboratory Standards (2004) Performance standards for antimicrobial susceptibility testing: fourteenth informational supplement. Document M100-S14. Wayne, PA7 NCCLS. Papanicolaou GA, Medeiros AA, Jacoby GA (1990) Novel plasmidmediated h-lactamase (MIR-1) conferring resistance to oxyimino- and a-methoxy h-lactams in clinical isolates of Klebsiella pneumoniae. Antimicrob Agents Chemother 34:2200 – 2209. Paterson DL, Hujer KM, Hujer AM, Yeiser B, Bonomo MD, Rice LB, Bonomo RA (2003) Extended-spectrum h-lactamases in Klebsiella pneumoniae bloodstream isolates from seven countries: Dominance and widespread prevalence of SHV- and CTX-M-type h-lactamases. Antimicrob Agents Chemother 47:3554 – 3560. Perez-Perez FJ, Hanson ND (2002) Detection of plasmid-mediated AmpC h-lactamase genes in clinical isolates by using multiplex PCR. J Clin Microbiol 40:2153 – 2162. Philippon A, Arlet G, Jacoby GA (2002) Plasmid-determined AmpC-type h-lactamases. Antimicrob Agents Chemother 46:1 – 11. Reisbig MD, Hanson ND (2002) The ACT-1 plasmid-encoded AmpC h-lactamase is inducible: Detection in a complex h-lactamase background. J Antimicrob Chemother 49:557 – 560. Riccio ML, Franceschini N, Boschi L, Caravelli B, Cornaglia G, Fontana R, Amicosante G, Rossolini GM (2000) Characterization of the metalloh-lactamase determinant of Acinetobacter baumannii AC-54/97 reveals the existence of bla IMP allelic variants carried by gene cassettes of different phylogeny. Antimicrob Agents Chemother 44:1229 – 1235. Sambrook J, Russell DW (2001) Molecular cloning, a laboratory manual. 3rd ed. Cold Spring Harbor (NY)7 Cold Spring Harbor Laboratory Press, pp 1.32. Song W, Kim JS, Kim MN, Kim EC, Park YJ, Yong D, Lee K, Lee WG, Jeong SH, Lee KM (2002) Occurrence and genotypic distributions of plasmid-mediated AmpC h-lactamase–producing Escherichia coli and Klebsiella pneumoniae in Korea. Kor J Lab Med 22:410 – 416. Verdet C, Arlet G, Barnaud G, Lagrange PH, Philippon A (2000) A novel integron in Salmonella enterica serovar Enteritidis, carrying the bla DHA-1 gene and its regulator gene ampR, originated from Morganella morganii. Antimicrob Agents Chemother 44:222 – 225. Yan JJ, Ko WC, Jung YC, Chuang CL, Wu JJ (2002) Emergence of Klebsiella pneumoniae isolates producing inducible DHA-1 h-lactamase in a university hospital in Taiwan. J Clin Microbiol 40: 3121 – 3126.
Plasmid-mediated, inducible AmpC h-lactamase (DHA-1)-producing Enterobacteriaceae at a Korean hospital: wide dissemination in Klebsiella pneumoniae and Klebsiella oxytoca and emergence in Proteus mirabilis Dongeun Yonga,c,d, Youngsik Limc,d, Wonkeun Songc,e, Yeoung Seon Choic,d, Doe-Young Parkc,d, Hyukmin Leea,c, Jong Hwa Yumc,d, Kyungwon Leea,c,d,T, June Myung Kimb,c,d, Yunsop Chonga,c a
Department of Laboratory Medicine, Yonsei University College of Medicine, Seoul 120-752, South Korea b Department of Internal Medicine, Yonsei University College of Medicine, Seoul 120-752, South Korea c Research Institute of Bacterial Resistance, Yonsei University College of Medicine, Seoul 120-752, South Korea d BK21 Project for Medical Sciences, Yonsei University College of Medicine, Seoul 120-752, South Korea e Department of Laboratory Medicine, University of Hallym College of Medicine, Seoul 150-072, South Korea Received 9 December 2004; accepted 25 March 2005
Abstract The aim of the study was to investigate the phenotypic and genetic characteristics of recently emerging cefoxitin-resistant and inductionpositive isolates of Escherichia coli, Klebsiella species, and Proteus mirabilis. Strains of Enterobacteriaceae were isolated at a Korean tertiary care hospital between June and December 2002. Induction was tested using cefoxitin and aztreonam disks, the bla DHA allele was detected by PCR, and pulsed-field gel electrophoresis (PFGE) patterns were also analyzed. Among the cefoxitin-resistant isolates, 2.7% of E. coli, 21.1% of Klebsiella pneumoniae, 32.0% of Klebsiella oxytoca, and 8.3% of P. mirabilis isolates showed induction, and were bla DHA-1 allele positive. To the best of our knowledge, this is the first report of bla DHA-1 in P. mirabilis. The MICs of ceftazidime, cefotaxime, and aztreonam increased significantly by higher inoculum, suggesting that their clinical usefulness is limited. Presence of multiple PFGE patterns and identical patterns in some isolates suggest that the widely disseminated bla DHA-1 in Klebsiella species was because of both horizontal and clonal spread. D 2005 Elsevier Inc. All rights reserved. Keywords: DHA-1; AmpC h-lactamase; Proteus mirabilis; Korea
1. Introduction h-Lactamase production is the predominant mechanism for resistance to h-lactam antibiotics in Gram-negative bacilli. Resistance due to extended-spectrum h-lactamase (ESBL) appeared in Klebsiella pneumoniae and in other enterobacterial species, but they remained susceptible to cephamycins (Philippon et al., 2002). Plasmid-mediated AmpC h-lactamases were first detected in 1988 (Bauernfeind et al., 1989, 1998; Papanicolaou
T Corresponding author. Department of Laboratory Medicine, Yonsei Universtiy College of Medicine, Seoul 120-752, South Korea. Tel.: +82-2228-2446; fax: +82-2-313-0908. E-mail address: [email protected] (K. Lee). 0732-8893/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2005.03.008
et al., 1990), and subsequently various AmpC enzymes have been found, particularly in Klebsiella spp., Escherichia coli, and Proteus mirabilis (Paterson et al., 2003; Philippon et al., 2002). The majority of plasmid-mediated AmpC enzymes are expressed constitutively, with the exceptions of DHA, ACT-1, and CFE-1. DHA-1 was first reported in 1992 in Salmonella enteritidis in Saudi Arabia and France (Gaillot et al., 1997), DHA-2 in K. pneumoniae in France (Fortineau et al., 2001), ACT-1 in K. pneumoniae in the United States (Reisbig and Hanson, 2002), and CFE-1 in Citrobacter freundii in Japan (Nakano et al., 2004). In Korea, cefoxitin-resistant E. coli and K. pneumoniae has been increasingly noted. A recent investigation detected 15 isolates with undetermined bla DHA subtype (Song et al.,
66
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
2002). Recently, we have increasingly noted presence of induction-positive isolates of K. pneumoniae at a tertiary care hospital laboratory. Some of the plasmid-mediated AmpC enzyme-producing isolates are susceptible to third-generation cephalosporins and aztreonam in vitro (Philippon et al., 2002), but their clinical utility is uncertain. As for DHA-producing organisms, even in vitro susceptibility data are limited, because only a few studies have been reported (Alvarez et al., 2004; Yan et al., 2002). In addition to the treatment problem, accurate differentiating methods for plasmid-mediated AmpC h-lactamase-producing isolates are required to improve surveillance and infection control (Perez-Perez and Hanson, 2002), but the current guidelines from the National Committee for Clinical Laboratory Standards (NCCLS, 2004) do not include any recommendation. Recently, Perez-Perez and Hanson (2002) recommended the use of multiplex PCR to identify family-specific AmpC h-lactamase genes. The aim of this study was to determine the phenotypic and genetic characteristics of induction-positive strains of E. coli, Klebsiella spp., and P. mirabilis isolated at a Korean hospital. We also investigated the susceptibility, inoculum effect on the minimum inhibitory concentrations of antimicrobial agents, and the molecular epidemiological features of bla DHA-1 allele-positive strains. 2. Materials and methods 2.1. Screening of inducible AmpC b-lactamase– and ESBL-producing isolates Bacterial strains were isolated from patients at a Korean tertiary care hospital between June and December 2002. Species were identified by conventional methods or by using the Vitek GNI system (bioMe´rieux, Marcy l’Etoile, France). Cefoxitin-resistant E. coli, Klebsiella spp., and P. mirabilis isolates, detected by the disk diffusion method (NCCLS, 2002) using commercial disks and MuellerHinton agar (Becton Dickinson, Cockeysville, MD), were tested for induction by applying cefoxitin and aztreonam disks 15 mm apart from edge to edge (Yan et al., 2002). A presence of truncated inhibition zone of aztreonam disk was interpreted as induction positive. ESBL producers were detected the double-disk synergy test using amoxicillin-clavulanic acid, cefotaxime, ceftazidime, and cefepime disks at a distance of 15 mm from disk edge to edge. Induction-positive isolates were stored in skim milk at 708C until used for further tests. 2.2. PCR protocol and nucleotide sequencing bla DHA alleles were detected by PCR using 1 AL (20 pmol) of each of the following primers: DHA-F (5V-ATG AAA AAA TCG TTA TCT GCA A-3V) and DHA-R (5V-CTA TTT TGA GTG CAC TGG AAT AA-3V), which amplify 1140 bp of the bla DHA gene, 1 AL of heat-
extracted template, and PreMix (Bioneer, Cheongwon, Korea) in a total volume of 20 AL. bla DHA allele-positive isolates were further tested: for bla DHA-1, using DHA-1/2 F (5V-TCT GCC GCT GAT AAT GTC G-3V), which is common to both bla DHA-1 and bla DHA-2, and DHA-1 R (5V-GCC GCC GGA TCA TTC AGC GC-3V); for bla DHA2, using DHA-1/2 F and DHA-2 R (5V-TCT GCC GGG TCA TTC AAC AT-3V), respectively. For nucleotide sequencing, PCR products were amplified in a mixture of 1 AL (20 pmol) of each of DHA-1 X-F (5V-AAC GTC TGA CCA TAA TCC AC-3V) and DHA-1 X-R (5V-GCC CAT AAA GCA AAT TAT TG-3V), and 0.5 AL of 2.5 U ExTaq DNA polymerase (Takara, Otsu, Japan) in total volume of 100 AL. DNAs were extracted from the PCR products using a commercial kit (Qiagen, Hilden, Germany), and were sequenced using the primers DHA-1 X-F, DHA-1 XR, DHA-1 Lt-F (5V-GCT GGG GTT ATC TCA CAC CT-3V) and DHA-1 Rt-R (5V-CCG TAC GCA TAC TGG CTT TGC-3V) by the dideoxy-chain termination method using an ABI 3700 DNA sequencer (Perkin-Elmer, Foster City, CA). The PCR conditions used to detect and to sequence bla DHA, bla DHA-1, and bla DHA-2 consisted of 30 cycles denaturation at 948C for 30 s, annealing at 508C (to detect bla DHA and sequence), at 588C (to detect bla DHA-1), or at 558C (to detect bla DHA-2) for 30 s, and extension at 728C for 50 s followed by a final extension at 728C for 7 min. Class 1 integron was detected by PCR as described previously (Riccio et al., 2000). E. coli strain (YMC 02/8/U310), which harbors bla DHA-1, and Pseudomonas aeruginosa (YMC 95/1/704), which harbors class 1 integron, were used as a positive control strains for DHA and class 1 integron PCR, respectively. 2.3. MIC determination The NCCLS agar dilution method (NCCLS, 2002) was used. The antimicrobial agents used were amoxicillin (Kun Wha Pharmaceuticals, Seoul, Korea), ceftazidime and clavulanic acid (GlaxoSmithKline, Greenford, UK), piperacillin (Samsung Pharmaceuticals, Seoul, Korea), tazobactam (Wyeth, Pearl River, NY), cephalothin (Sigma, St. Louis, MO), cefoxitin and imipenem (Merck Sharp & Dohme, West Point, PA), cefotaxime (Handok Pharmaceuticals, Seoul, Korea), and aztreonam and cefepime (Bristol-Myers Squibb, Princeton, NJ). Constant concentrations (4 Ag/mL) of tazobactam or clavulanic acid were added to piperacillin, ceftazidime, cefotaxime, aztreonam, or cefepime. The ratio of amoxicillin to clavulanic acid used was 2:1. To determine the inoculum effect, approximate inocula of 104 and 106 CFU per spot were applied using a Steers replicator (Craft Machine, Chester, PA). To exclude the effect of coexistent ESBL, we did not include the ESBL-producing isolates. E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used for quality control.
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70 Table 1 bla DHA allele–positive isolates among E. coli, Klebsiella spp., and P. mirabilis Species (no. of isolates tested)
No. (%) of isolates with Cefoxitin resistance
Induction positivea
bla DHA positive by PCRb
E. coli (1937) K. pneumoniae (899) K. oxytoca (149) P. mirabilis (124) Total (3109)
291 (15.0) 261 (29.0)
8 (2.7) 55 (21.1)
8 55
2 6
25 (16.8) 12 (9.7) 589 (18.9)
8 (32.0) 1 (8.3) 72 (12.2)
8 1 72
4 1 13
bla DHA-1 sequencedc
a
Cefoxitin-resistant isolates only were tested using cefoxitin and aztreonam disks. b Antagonism-positive isolates only were tested. c Randomly selected bla DHA allele-positive isolates were tested.
2.4. Transfer of resistance, plasmid analysis, and Southern hybridization The transferability of inducible cefoxitin resistance was tested by the agar mating method using randomly selected isolates and azide-resistant E. coli J53 recipient. MacConkey agar supplemented with 100 Ag/mL of sodium azide, and 16 Ag/mL of cefoxitin was used to select transconjugants. Plasmids were isolated from clinical isolates and their transconjugants using the alkaline lysis method (Sambrook and Russell, 2001), and the size was estimated by comparing their electrophoretic mobilities with those of E. coli V517. Southern hybridization was used to determine the location of bla DHA-1 on the plasmid, chromosome, or integron. A bla DHA-1 probe was prepared using a DIG DNA Labeling and Detection Kit (Roche Diagnostics, Mannheim, Germany), and testing was performed according to the manufacturer’s instructions.
67
2.5. Pulsed-field gel electrophoresis XbaI- and I-CeuI-digested genomic DNA was prepared according to the instruction of BioRad (Hercules, CA), and fragments were separated for 20 h at 6 V/cm at 118C using a CHEF-DR II System (BioRad), with initial and final pulse times of 0.5 and 30 s, respectively. Pulsedfield gel electrophoresis (PFGE) banding type similarities were determined using the Dice coefficient and unweighted pair group method using UVIBand/Map software (UVItech, Cambridge, UK). Possibly, related isolates with 6 band differences were identified by using 80% similarity coefficients. 2.6. Nucleotide sequence accession number The nucleotide sequences of bla DHA-1 in 4 clinical isolates have been assigned to the following GenBank accession numbers: E. coli YMC 02/08/U310 (AY205598), K. pneumoniae YMC 02/09/R340 (AY205600), Klebsiella oxytoca YMC 02/09/P206 (AY205599), and P. mirabilis YMC 02/08/U1255 (AY205601).
3. Results 3.1. Detection of DHA-producing isolates Of the 3,109 unduplicated isolates of E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis tested, 15.0%, 29.0%, 16.8%, and 9.7%, respectively, were resistant to cefoxitin (Table 1). Of the cefoxitin-resistant isolates, induction was observed in 8 of 291 (2.7%) E. coli, 55 of 261 (21.1%) K. pneumoniae, 8 of 25 (32.0%) K. oxytoca, and 1 of 12 (8.3%) P. mirabilis isolates. bla DHA alleles were detected by PCR in all of the 72 induction-positive isolates. All of these isolates showed
Table 2 Agar dilution antimicrobial susceptibilities of bla DHA-1 allele-positive isolates of E. coli, Klebsiella spp., and P. mirabilis depending on inoculum size Antimicrobial agents
MIC range (Ag/mL) with approximate inoculum (CFU/spot) E. coli (n = 6)
Amoxicillin Amoxicillin-clavulanic acid Piperacillin Piperacillin-tazobactam Cephalothin Cefoxitin Ceftazidime Ceftazidime-clavulanic acid Cefotaxime Cefotaxime-clavulanic acid Aztreonam Aztreonam-clavulanic acid Cefepime Cefepime-clavulanic acid Imipenem
K. pneumoniae (n = 18)
K. oxytoca (n = 5)
P. mirabilis (n = 1)
104
106
104
106
104
106
104
106
N128 32 64 – N256 1–2 N128 32– N128 1–16 4 –32 1– 4 1– 4 0.5 – 4 1– 8 0.03 – 0.06 V 0.008 – 0.06 0.12 – 0.25
N128 N128 N 256 N128 N128 128 – N128 16 – N128 NT 64 – N128 NT 128 – N128 NT NT NT NT
N128 32 – 64 64 – N 256 2– N 256 N128 128 – N128 4 – N128 16 – N128 1– 64 4 – 64 0.5 –128 4 – 64 0.03 – 0.5 0.06 – 0.5 0.12 –1
N128 N128 N 256 N128 N128 128 – N128 64 – N128 NT N128 NT N128 NT NT NT NT
N128 32 32– N 256 1– N 256 N128 64 – N128 1–32 4 – N128 1– 4 2–16 0.5 –16 1–16 0.03 – 0.06 0.015 – 0.12 0.25 – 0.5
N128 N128 N 256 N128 N128 128– N128 64 – N128 NT 64 – N128 NT N128 NT NT NT NT
N128 128 2 V 0.5 N128 64 4 32 0.5 32 0.015 1 0.06 0.25 2
N128 N128 N 256 N128 N128 N128 N128 NT N128 NT N128 NT NT NT NT
The MICs of ESBL-producing isolates were not included.
68
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
Table 3 Agar dilution antimicrobial susceptibilities of bla DHA-1-positive clinical isolates and their transconjugants Antimicrobial agents
Amoxicillin-clavulanic acid Piperacillin Piperacillin-tazobactam Ceftazidime Ceftazidime-clavulanic acid Cefotaxime Cefotaxime-clavulanic acid Aztreonam Aztreonam-clavulanic acid
MIC (Ag/mL)a E. coli YMC 02/8/U310
K. pneumoniae YMC 02/10/R340
K. oxytoca YMC 02/9/P206
Wild
Transconjugant
Wild
Transconjugant
Wild
Transconjugant
32 N256 2 8 16 4 4 1 4
64 32 1 2 16 0.5 2 0.5 4
32 64 4 64 N128 4 16 8 16
64 16 1 4 16 1 2 1 4
32 128 2 2 4 1 2 1 4
64 32 1 4 16 1 2 2 4
a The MICs (Ag/mL) of other antimicrobial agents for all 3 isolates and their transconjugants were the following: amoxicillin, N128 and N128; cephalothin, N128 and z32; cefoxitin, z128 and z64; cefepime, 0.03 – 0.06 and 0.015 – 0.03; cefepime-clavulanic acid, 0.015 – 0.06 and V 0.008 – 0.12; and imipenem, 0.25 – 0.5 and 0.12 – 0.25, respectively.
positive PCR reaction with bla DHA-1 allele-specific primers, but not with bla DHA-2 allele-specific primers. PCR products from 13 randomly selected isolates were used to determine nucleotide sequences, and all were found to be identical to that of bla DHA-1.
3.2. Resistance transfer and Southern hybridization Inducible cefoxitin resistance was transferred by conjugation from all of the randomly selected 2 isolates of E. coli, 1 K. pneumoniae, and 1 K. oxytoca. Repeated attempts failed to transfer resistance from 1 P. mirabilis isolate.
Fig. 1. A dendrogram of PFGE bands of XbaI-digested genomic DNA of K. pneumoniae based on the Dice coefficient. Twenty-seven types and five subtypes were recognized. Thirteen, eight, and five isolates of K. pneumoniae belonged to types B, G, and X, respectively. The numbers of isolates with the same band type are shown in parentheses.
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
The 4 clinical isolates and 4 transconjugants contained plasmids of N150 kb, which hybridized with the bla DHA-1 probe, but the I-CeuI-digested genomic DNA bands of 3 isolates and their transconjugants did not hybridize. Class 1 integrons were detected by PCR in 28 of 34 isolates (2 isolates of E. coli, 21 of K. pneumoniae, 4 of K. oxytoca, and 1 of P. mirabilis), but none of the amplicons hybridized with the bla DHA-1 probe. 3.3. Antimicrobial susceptibility of isolates and transconjugants The antimicrobial susceptibilities of the bla DHA-1positive clinical isolates excluding 2 ESBL-producing ones showed that the MICs of amoxicillin and cephalothin were N128 Ag/mL and those of cefepime and imipenem were V 2 Ag/mL for all isolates (Table 2). The MICs of ceftazidime, cefotaxime, and aztreonam were highly variable depending on isolates, and the levels increased significantly in the presence of clavulanic acid for most isolates, but those of piperacillin tended to decrease in the presence of tazobactam. When the inoculum was increased to approximately 106 CFU, all the MICs of piperacillin, ceftazidime, cefotaxime, and aztreonam increased significantly. Comparison showed that the MICs of piperacillin for 3 clinical isolates varied depending on species (i.e., N256 Ag/mL for E. coli, 64 Ag/mL for K. pneumoniae, and 128 Ag/mL for K. oxytoca), but those for their transconjugants were similar, 32, 16, and 32 Ag/mL, respectively (Table 3). 3.4. Epidemiology of DHA-1–producing isolates The majority of bla DHA-1 allele-positive strains were isolated from patients in general wards (56.9%), followed by patients in intensive care units (34.7%), and the sources were mostly sputum (45.8%) or urine (38.9%). PFGE patterns of the XbaI-digested genomic DNAs of the 54 isolates of bla DHA-1 allele-positive K. pneumoniae showed 27 different types, indicating multiclonality. However, 13, 8, and 5 isolates belonged to types B, G, and X, respectively (Fig. 1). All 8 isolates of E. coli and 7 of 8 K. oxytoca isolates showed different PFGE band patterns (data not shown).
4. Discussion Plasmid-mediated AmpC h-lactamase-producing gramnegative bacilli are being found increasingly in many parts of the world (Philippon et al., 2002), but reports on DHAproducing organisms are relatively rare. CMY-1-producing K. pneumoniae and E. coli have been prevalent and single amino acid-substituted CMY-1b (CMY-10) has emerged in Korea (Bauernfeind et al., 1998). Therefore, most of the plasmid-mediated AmpC h-lactamases in Korea was presumed to be these enzymes until other types were recently
69
reported (Song et al., 2002). It is interesting that bla DHA-1 allele-positive isolates of E. coli, K. pneumoniae, and K. oxytoca are widely disseminated at a Korean hospital. All of bla DHA allele-positive isolates in this study showed positive reaction only with bla DHA-1 allele-specific primers, and the bla DHA alleles in 13 randomly selected isolates showed a sequence identical to that of bla DHA-1 (Gaillot et al., 1997). We presume from this result that all bla DHA allele-positive isolates in our study carried the bla DHA-1 gene. As far as we are aware, this detection of bla DHA-1 in P. mirabilis is the first worldwide and indicates this species can also acquire the resistance. Plasmids carrying bla DHA-1 in K. pneumoniae and S. enteritidis (Gaillot et al., 1997; Yan et al., 2002) and those carrying bla DHA-2 in K. pneumoniae (Fortineau et al., 2001) have been reported to be ca. 70 –200 kb, and some of these were transferable by conjugation. In the present study, resistance was transferred by agar mating, and the isolates and transconjugants carried plasmids of variable size of N150 kb. It has been reported that the bla DHA-1 gene is located on a class 1 integron (Verdet et al., 2000); thus, the failure of our PCR product of the class 1 integron to hybridize with the bla DHA probe requires further investigation. The MICs for ESBL-producing isolates were not included in Table 2. We tested analytical isoelectric focusing and found none of them had h-lactamase band of pI N8.0, suggesting no coexistence of other AmpC enzymes in these strains. In another our study using multiplex PCR, it was shown that none of the isolates had more than one plasmid-mediated AmpC h-lactamase genes (unpublished data). Some bla DHA-1 allele-positive isolates were inhibited by low concentrations of ceftazidime, cefotaxime, or aztreonam (Table 2), but the clinical efficacies of these drugs are unknown. Controversy continues concerning the use of third-generation cephalosporins for the treatment of infections due to inducible chromosomal AmpC h-lactamaseproducing Enterobacteriaceae (Goldstein, 2002; Livermore et al., 2004). However, in view of the significant increase in MICs with higher inoculum or by the addition of clavulanic acid, third-generation cephalosporins with or without clavulanic acid may be inappropriate for the treatment of severe infections by DHA-1 producing isolates. As in other studies (Barnaud et al., 1998; Fortineau et al., 2001; Gaillot et al., 1997; Yan et al., 2002), all of the bla DHA-1 allelepositive isolates remained highly susceptible to cefepime and imipenem. Presence of multiple PFGE patterns of the XbaI-digested genomic DNAs of bla DHA-positive isolates of E. coli, K. pneumoniae, and K. oxytoca (Fig. 1) suggests that horizontal transfer of the gene-carrying plasmids occurs readily to genetically heterogeneous organisms. Some isolates with identical patterns indicated that clonal spread also played a role in the dissemination. As is in other nosocomial infections, the most frequent sources of the
70
D. Yong et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 65–70
bla DHA-1 allele-positive isolates were sputum and urine samples from hospitalized patients. The types of plasmid-mediated AmpC h-lactamase in E. coli and Klebsiella spp. cannot be differentiated by phenotype. AmpC multiplex PCR differentiated the 6 plasmid-mediated AmpC-specific families (Perez-Perez and Hanson, 2002). It has been reported that all DHAproducing isolates were induction test positive (Yan et al., 2002). In the present study, bla DHA alleles were detected in all 72 isolates showing induction. However, the sensitivity of the aztreonam and cefoxitin disk method is unknown, although it was more sensitive than tests using ceftazidime or cefotaxime (data not shown). Therefore, further studies are needed to determine the prevalence of bla DHA-1 allelepositive isolates in Korea. In conclusion, bla DHA-1 allele-positive isolates are common, especially in K. pneumoniae and K. oxytoca in Korea. bla DHA-1 was detected for the first time in P. mirabilis, and bla DHA-1 was found transferable by conjugation. A simple and accurate detection method for DHA-1–producing isolates is needed to control the spread of resistance. It is a concern that the DHA-1 h-lactamase gene may spread to various species of Enterobacteriaceae, against which all h-lactams, with the exception of carbapenem and fourth-generation cephalosporins, may be inactive. Acknowledgment This study was supported by the Yonsei University Research Fund (2003-35). References Alvarez M, Tran JH, Chow N, Jacoby GA (2004) Epidemiology of conjugative plasmid-mediated AmpC h-lactamases in the United States. Antimicrob Agents Chemother 48:533 – 537. Barnaud G, Arlet G, Verdet C, Gaillot O, Lagrange PH, Philippon A (1998) Salmonella enteritidis: AmpC plasmid-mediated inducible h-lactamase (DHA-1) with an ampR gene from Morganella morganii. Antimicrob Agents Chemother 42:2352 – 2358. Bauernfeind A, Chong Y, Schweighart S (1989) Extended broad spectrum h-lactamase in Klebsiella pneumoniae including resistance to cephamycins. Infection 17:316 – 321. Bauernfeind A, Chong Y, Lee K (1998) Plasmid-mediated AmpC h-lactamases: How far have we gone 10 years after the discovery? Yonsei Med J 39:520 – 525. Fortineau N, Poirel L, Nordmann P (2001) Plasmid-mediated and inducible cephalosporinase DHA-2 from Klebsiella pneumoniae. J Antimicrob Chemother 47:207 – 210.
Gaillot O, Clement C, Simonet M, Philippon A (1997) Novel transferable h-lactam resistance with cephalosporinase characteristics in Salmonella enteritidis. J Antimicrob Chemother 39:85 – 87. Goldstein FW (2002) Cephalosporinase induction and cephalosporin resistance: A longstanding misinterpretation. Clin Microbiol Infect 8: 823 – 825. Livermore DM, Brown DF, Quinn JP, Carmeli Y, Paterson DL, Yu VL (2004) Should third-generation cephalosporins be avoided against AmpC-inducible Enterobacteriaceae? Clin Microbiol Infect 10:84 – 85. Nakano R, Okamoto R, Nakano Y, Kaneko K, Okitsu N, Hosaka Y, Inoue M (2004) CFE-1, a novel plasmid-encoded AmpC h-lactamase with an ampR gene originating from Citrobacter freundii. Antimicrob Agents Chemother 48:1151 – 1158. National Committee for Clinical Laboratory Standards (2002) Performance standards for antimicrobial susceptibility testing: Twelfth informational supplement. Document M100-S12. Wayne, PA7 NCCLS. National Committee for Clinical Laboratory Standards (2004) Performance standards for antimicrobial susceptibility testing: fourteenth informational supplement. Document M100-S14. Wayne, PA7 NCCLS. Papanicolaou GA, Medeiros AA, Jacoby GA (1990) Novel plasmidmediated h-lactamase (MIR-1) conferring resistance to oxyimino- and a-methoxy h-lactams in clinical isolates of Klebsiella pneumoniae. Antimicrob Agents Chemother 34:2200 – 2209. Paterson DL, Hujer KM, Hujer AM, Yeiser B, Bonomo MD, Rice LB, Bonomo RA (2003) Extended-spectrum h-lactamases in Klebsiella pneumoniae bloodstream isolates from seven countries: Dominance and widespread prevalence of SHV- and CTX-M-type h-lactamases. Antimicrob Agents Chemother 47:3554 – 3560. Perez-Perez FJ, Hanson ND (2002) Detection of plasmid-mediated AmpC h-lactamase genes in clinical isolates by using multiplex PCR. J Clin Microbiol 40:2153 – 2162. Philippon A, Arlet G, Jacoby GA (2002) Plasmid-determined AmpC-type h-lactamases. Antimicrob Agents Chemother 46:1 – 11. Reisbig MD, Hanson ND (2002) The ACT-1 plasmid-encoded AmpC h-lactamase is inducible: Detection in a complex h-lactamase background. J Antimicrob Chemother 49:557 – 560. Riccio ML, Franceschini N, Boschi L, Caravelli B, Cornaglia G, Fontana R, Amicosante G, Rossolini GM (2000) Characterization of the metalloh-lactamase determinant of Acinetobacter baumannii AC-54/97 reveals the existence of bla IMP allelic variants carried by gene cassettes of different phylogeny. Antimicrob Agents Chemother 44:1229 – 1235. Sambrook J, Russell DW (2001) Molecular cloning, a laboratory manual. 3rd ed. Cold Spring Harbor (NY)7 Cold Spring Harbor Laboratory Press, pp 1.32. Song W, Kim JS, Kim MN, Kim EC, Park YJ, Yong D, Lee K, Lee WG, Jeong SH, Lee KM (2002) Occurrence and genotypic distributions of plasmid-mediated AmpC h-lactamase–producing Escherichia coli and Klebsiella pneumoniae in Korea. Kor J Lab Med 22:410 – 416. Verdet C, Arlet G, Barnaud G, Lagrange PH, Philippon A (2000) A novel integron in Salmonella enterica serovar Enteritidis, carrying the bla DHA-1 gene and its regulator gene ampR, originated from Morganella morganii. Antimicrob Agents Chemother 44:222 – 225. Yan JJ, Ko WC, Jung YC, Chuang CL, Wu JJ (2002) Emergence of Klebsiella pneumoniae isolates producing inducible DHA-1 h-lactamase in a university hospital in Taiwan. J Clin Microbiol 40: 3121 – 3126.