Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, June 2008, p. 2014-2018, Vol. 52, No. 6
0066-4804/08/$08.00+0 doi:10.1128/AAC.01539-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

2nd Affiliated Hospital of Zhejiang University, Zhejiang University, 88 Jiefang Rd., Hangzhou, China 310009
Received 28 November 2007/ Returned for modification 25 January 2008/ Accepted 5 March 2008
|
|
|---|
|
|
|---|
Carbapenem-hydrolyzing KPC β-lactamases are a group of recently identified carbapenemases which belong to Bush group 2f, molecular class A. KPCs are capable of hydrolyzing carbapenems, penicillins, cephalosporins, and aztreonam, and they are inhibited by clavulanic acid and tazobactam. The initial report of one of these β-lactamases, KPC-1, was from a carbapenem-resistant Klebsiella pneumoniae strain isolated in North Carolina (27). KPC-2 was then found in isolates of K. pneumoniae (17), Salmonella enterica (16), Klebsiella oxytoca (28), and an Enterobacter sp. (11). Soon afterward, KPC-3 was found in K. pneumoniae (26) and Enterobacter cloacae isolates from New York (2). Recently, KPCs were found in France (18), South America (23), and Israel (19) outside the United States, and KPC-2 was identified for the first time in Pseudomonas aeruginosa isolates outside of the family Enterobacteriaceae (24). KPC-2 has emerged in China. A K. pneumoniae isolate from Hangzhou city producing KPC-2 has been reported (25). Almost simultaneously, we identified KPC-2 in three Serratia marcescens isolates from the same city but a different hospital (30). Most KPC enzymes occurred sporadically. However, it is worth noting that KPC-producing Klebsiella species are predominant in some facilities in New York (1, 3, 26).
In the present report, we describe the emergence of KPC-2-producing Enterobacteriaceae including S. marcescens, K. pneumoniae, and Escherichia coli in intensive care units (ICUs) in our hospital.
|
|
|---|
Antimicrobial susceptibility testing. The MICs of 15 antibiotics were determined using the agar dilution method according to CLSI recommendations (7). Antimicrobial agents were obtained from Sigma (St. Louis, MO) (ampicillin, piperacillin, cefotaxime, and gentamicin) and the pharmaceutical companies Merck & Co., Inc. (imipenem, ertapenem, and cefoxitin), Dainippon Sumitomo Pharma Co., Ltd. (meropenem), Wyeth Holdings Co. (piperacillin-tazobactam), Pfizer (cefoperazone/sulbactam), GlaxoSmithKline (ceftazidime), Roche (ceftriaxone), Bristol-Meyers Squibb (aztreonam, cefepime), and Bayer (ciprofloxacin). E. coli ATCC 25922 was used for quality control.
PFGE typing. Pulsed-field gel electrophoresis (PFGE) typing of K. pneumoniae isolates was performed according to the procedure described by PulseNet from the website of the Centers for Disease Control and Prevention (http://www.cdc.gov/pulsenet/protocols.htm) in a Rotaphor System 6.0 instrument (Whatman Biometra, Goettingen, Germany). The XbaI restriction patterns of the isolates' genomic DNAs were analyzed and interpreted according to the criteria of Tenover et al. (21).
ERIC-PCR typing. Total DNAs of S. marcescens isolates were obtained with an Axyprep bacterial genomic DNA miniprep kit (Axygen Scientific, Union City, CA) and were used as templates in enterobacterial repetitive intergenic consensus-PCR (ERIC-PCR) analysis. The sequences of the primers and PCR conditions were as described by Versalovic et al. (22).
Conjugal transfer experiment and analysis of plasmid. The conjugation experiment was carried out in mixed broth cultures. Rifampin-resistant E. coli EC600 (LacZ– Nalr Rifr) was used as the recipient strain. Overnight cultures of the donor strain (200 µl) and recipient strain (100 µl) were mixed with 600 µl of fresh Mueller-Hinton broth and were incubated for 24 h at 35°C. Then, the mixture was inoculated on Mueller-Hinton agar plates containing rifampin (Sigma; 700 mg/liter) plus imipenem (0.5 mg/liter) for 24 h at 35°C. The colonies that grew on the selecting medium were picked up and identified by the Vitek system.
Plasmid DNAs were obtained with an Axyprep plasmid miniprep kit (Axygen Scientific) and were digested by various endonucleases, including EcoRI, HindIII, and BcuI (MBI Fermentas, Lithuania). The plasmid DNA and digestion products were separated by electrophoresis at 0.6% agarose gel at a constant voltage of 100 V for 4 h. The PCR-based replicon typing of the blaKPC-2-encoding plasmids was conducted following the method described by Carattoli et al. (5).
IEF of β-lactamases. The crude β-lactamase extracts of original isolates and their E. coli transconjugants were prepared by ultrasonic treatment of bacterial cells. Isoelectric focusing (IEF) was performed on PhastGel polyacrylamide gel (pH, 3 to 9; Amersham Biosciences, Uppsala, Sweden) using a PhastSystem instrument (Pharmacia Biotech, Uppsala, Sweden) according to the method of Mathew et al. (15). β-Lactamase activity was visualized by staining the gel with Nitrocefin (Oxoid, Basingstoke, Hampshire, England). The isoelectric points (pIs) were determined after comparison to known β-lactamases TEM-1 (pI of 5.4), TEM-28 (pI of 6.1), SHV-7 (pI of 7.6), and ACT-1 (pI of 9.0).
PCR amplification and DNA sequence analysis of bla genes.
Plasmid DNAs from original isolates and their E. coli transconjugants were used as templates in PCR amplification. The primers used to amplify blaKPC, blaTEM, blaSHV, and blaCTX-M were as described previously (27, 29). The reaction was conducted in a GeneAmp PCR System 9600 thermal cycler (Applied Biosystems, Foster City, CA). The PCR products were cloned into a pGEM-T Easy vector (Promega), and recombinant plasmids were transformed into E. coli DH5
as described previously (6). Inserts were sequenced using an ABI3730 sequencer (Applied Biosystems), and the sequences were compared with the reported sequences from GenBank.
Analysis of OMPs. Outer membrane proteins (OMPs) were isolated as described by Hernandez-Alles et al. (9). Strains were grown in Mueller-Hinton broth overnight at 37°C with shaking. Electrophoretic analysis of OMPs by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed in 11.6% acrylamide-0.4% bisacrylamide-0.1% SDS gels. The 0.75-mm thick minigel was run at a constant current of 20 mA for 85 min with a Mini Protein 3 slab electrophoresis cell (Bio-Rad, Richmond, CA).
The ompK35 and ompK36 genes of K. pneumoniae K1 and K10 were amplified by using PCR (13). The products were sequenced, and the sequences were compared with the reported sequences from GenBank.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Antimicrobial susceptibility patterns of S. marcescens, K. pneumoniae, and E. coli isolates and their E. coli transconjugants
|
![]() View larger version (94K): [in a new window] |
FIG. 1. PFGE patterns of chromosomal DNA restriction fragments from K. pneumoniae isolates. Lanes 1 to 10, K. pneumoniae K1 to K10; lanes 11 and 12, K. pneumoniae control strains that did not produce KPC-2.
|
Transfer of carbapenem resistance and plasmid analysis.
Transfer of β-lactam resistance from S. marcescens, K. pneumoniae, and E. coli isolates to E. coli EC600 by conjugation was successful. All the E. coli transconjugants exhibited significantly reduced carbapenem susceptibility with imipenem and meropenem MICs of
1 to 2 µg/ml and an ertapenem MIC of 4 µg/ml. They were also resistant to penicillins and aztreonam and were resistant or intermediately resistant to cephalosporins and cefoxitin, but they were susceptible to quinolones and aminoglycosides (Table 1).
K. pneumoniae isolates had identical plasmid profiles, except for that of strain K10. S. marcescens isolates and all E. coli transconjugants acquired a plasmid with a size of approximately 50 kb similar to the largest plasmid of E. coli V517 (data not shown). Plasmid DNA restriction analysis using EcoRI, HindIII, and BcuI showed identical restriction patterns among the representative E. coli transconjugants (Fig. 2). PCR-based replicon typing indicated that all blaKPC-2-encoding plasmids were negative for all the replicons tested, while the HI1, A/C, and FIIAs replicons were detected in total plasmid from K. pneumoniae K1 (data not shown). These results suggested that the carbapenem resistance-related plasmids existing in clinical isolates of S. marcescens, K. pneumoniae, and E. coli were closely related.
![]() View larger version (47K): [in a new window] |
FIG. 2. Restriction patterns of plasmid DNA from E. coli transconjugants of E. coli, partial K. pneumoniae and S. marcescens isolates, and partial original S. marcescens isolates. Lane 1, E. coli transconjugant of E. coli E1; lanes 2 to 4, E. coli transconjugants of K. pneumoniae K1, K8, and K10; lanes 5 to 7, E. coli transconjugants of S. marcescens S1, S10, and S20; lanes 8 and 9, original S. marcescens S1 and S10; lane 10, E. coli EC600 as a negative control; M, 1-kb DNA ladder (MBI Fermentas). Plasmid DNA was digested with EcoRI, HindIII, and BcuI endonucleases.
|
![]() View larger version (64K): [in a new window] |
FIG. 3. IEF patterns of crude β-lactamase extracts from partial S. marcescens, K. pneumoniae, and E. coli isolates and their E. coli transconjugants. (A) Lane 1, E. coli E1; lanes 2 to 4, K. pneumoniae K1, K8, and K10; lane 5, E. coli transconjugant of E. coli E1; lanes 6 and 7, E. coli transconjugants of K. pneumoniae K1 and K10; M, strains producing TEM-1 (pI of 5.4), TEM-28 (pI of 6.1), SHV-7 (pI of 7.6), and ACT-1 (pI of 9.0). (B) Lanes 1 to 4, S. marcescens S1, S5, S10, and S20; lanes 5 to 7, E. coli transconjugants of S. marcescens S1, S10, and S20; M, strains producing TEM-1 (pI of 5.4), TEM-28 (pI of 6.1), SHV-7 (pI of 7.6), and ACT-1 (pI of 9.0).
|
Analysis of OMPs. The MICs of carbapenems for K. pneumoniae K10 were significantly higher than those of other K. pneumoniae isolates and E. coli transconjugants. This may be due to alterations in outer membrane permeability. Therefore, SDS-PAGE analysis of OMPs was performed, and the result revealed the lack of two OMPs of approximately 39 and 32 kDa in K. pneumoniae K10, compared to results for strains K1 and ATCC 13883 (Fig. 4). The 32-kDa OMP corresponds to OmpA, which is a specific porin and does not contribute to antimicrobial resistance phenotypes. It was uncertain whether the 39-kDa OMP was OmpK36 or OmpK35, since in some strains the OmpK36 migrates faster than OmpK35 (9). Amplification and sequencing of the ompK35 gene indicated that both K. pneumoniae K1 and K10 contained identical gene sequences for ompK35. For the ompK36 gene, the size of the amplification product from K. pneumoniae K10 was about 1,600 bp larger than that of strain K1. An insertion sequence ISEcp1 was identified between nucleotides 4 and 5 upstream of the ompK36 start codon in strain K10. We inferred that ISEcp1 interfered with the upstream regulatory region and resulted in disruption of the ompK36 gene. These results suggested that OmpK36 was not expressed in K. pneumoniae K10.
![]() View larger version (53K): [in a new window] |
FIG. 4. SDS-PAGE analysis of OMPs extracted from K. pneumoniae ATCC 13883 and K. pneumoniae K1 and K10. Lane 1, K. pneumoniae ATCC 13883; lane 2, K. pneumoniae K1; lane 3, K. pneumoniae K10; M, protein molecular mass standard (MBI Fermentas).
|
|
|
|---|
The KPCs are most frequently detected in K. pneumoniae, and they are being increasingly detected in E. coli (4, 8, 10, 19). In this study, we described the first discovery of KPC-2 in E. coli in China. We identified KPC-2 in three S. marcescens isolates from the NICU in our hospital in February 2006 (30). From April 2006 to February 2007, 21 KPC-2-producing S. marcescens isolates were recovered in the same ward. The similar antimicrobial susceptibility patterns and identical ERIC-PCR patterns, plasmid profiles, restriction patterns, and IEF results suggested that they belonged to a clonal strain. During this period, 10 KPC-2-producing K. pneumoniae isolates that were indistinguishable or closely related to each other were recovered from the NICU and SICU. These results indicated that KPC-2-producing organisms were widespread in ICUs, especially in the NICU. To investigate the molecular epidemiology of the KPC-2-encoding plasmids existing in E. coli, S. marcescens, and K. pneumoniae isolates, restriction analysis using various endonucleases was performed. The identical restriction patterns suggested that the same transmissible KPC-2-encoding plasmid had spread among the three different genera and resulted in the spread of carbapenem resistance.
The pI 6.5 and 7.3 β-lactamases detected by IEF were not identified by PCR and DNA sequencing. In a previous study, we had demonstrated that the pI 6.5 β-lactamase was probably either the degradation product or the precursor of KPC-2 (30). Given the uncommon isoelectric point and low activity, we presumed that the pI 7.3 β-lactamase was either the degradation product or the precursor of CTX-M-14 or CTX-M-15.
Ertapenem, a relatively new carbapenem in clinical use, was strongly active against ESBL-producing Klebsiella isolates, with an MIC90 of 0.06 µg/ml, compared with 0.5 µg/ml for imipenem. Nevertheless, the MICs of ertapenem for ESBL-producing Klebsiella isolates were about two- to fourfold above those for nonproducers (14). Paterson et al. (20) reported that 10.9% of ESBL-producing K. pneumoniae organisms isolated from intra-abdominal infections worldwide were resistant to ertapenem, while only 4 to 5% were resistant to imipenem and meropenem. Jacoby et al. (12) reported that most K. pneumoniae derived from porin-deficient strains which produce plasmid-mediated β-lactamases were resistant to ertapenem but showed only slightly reduced susceptibility to imipenem and meropenem. In the current study, ertapenem showed slightly less activity against KPC-2 producers than did imipenem and meropenem. The MICs of ertapenem for major KPC-2-producing original isolates and all E. coli transconjugants were about twofold above those of imipenem and meropenem.
The E. coli transconjugant that only produced KPC-2 exhibited significantly reduced susceptibility to imipenem and meropenem (MICs of 1 to 2 µg/ml) and ertapenem (MICs of 4 µg/ml), suggesting that reduced carbapenem susceptibility in isolates of S. marcescens, K. pneumoniae, and E. coli was mainly due to production of KPC-2. However, it was not sufficient to explain the high-level carbapenem resistance in K. pneumoniae K10 (MICs of 128, 256, and >256 µg/ml), since the E. coli transconjugant remained susceptible to carbapenems at CLSI breakpoints (MICs of
4 µg/ml). In addition, the plasmid profiles and PFGE patterns of K. pneumoniae K10 were slightly different from those of other K. pneumoniae isolates. These results suggested that other mechanisms may contribute to carbapenem resistance in strain K10. The loss of a 39-kDa porin was observed in this isolate by SDS-PAGE of OMPs. The lost porin was subsequently proven to be OmpK36 and not the OmpK35 that was described by Yigit et al. (27) and Woodford et al. (26).
KPC-producing organisms are being increasingly detected. The host range of these KPCs is no longer limited to K. pneumoniae, and the geographical distribution is no longer limited to the northeastern United States. Sporadic occurrences of KPC-2 were once reported in China. They are now rapidly spreading in ICUs in our hospital. Effective measures for early identification and control should be adopted to prevent the potential continuous dissemination of these carbapenem-resistant pathogens.
Published ahead of print on 10 March 2008. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»