|Home | About | Journals | Submit | Contact Us | Français|
All of the carbapenem-resistant Escherichia coli (CREC) isolates identified in our hospital from 2005 to 2008 (n = 10) were studied. CREC isolates were multidrug resistant, all carried blaKPC-2, and six of them were also extended-spectrum beta-lactamase producers. Pulsed-field gel electrophoresis indicated six genetic clones; within the same clone, similar transferable blaKPC-2-containing plasmids were found whereas plasmids differed between clones. Tn4401 elements were identified in all of these plasmids.
Carbapenem resistance in Escherichia coli is usually attributed to the acquisition of β-lactamases such as AmpC (14, 23, 24, 27, 31), metallo β-lactamases (4, 17, 25, 33), or KPC-type carbapenemases (2, 3, 26, 32). In 2005, KPC-2-mediated carbapenem-resistant E. coli (CREC) clinical strains were first identified in our hospital (21).
The increasing prevalence of carbapenem-resistant Enterobacteriaceae in Israel (30), along with concerns regarding the emergence of highly epidemic clones, led to the study of carbapenem resistance in E. coli in our hospital. We determined CREC prevalence, elucidated the molecular mechanisms contributing to carbapenem resistance, and explored the molecular epidemiology and plasmids associated with this resistance.
All of the CREC isolates identified in our hospital from February 2005 to October 2008 were included in this study. Strains were identified as resistant to at least one carbapenem using the Vitek-2 and agar dilution (MIC of imipenem or meropenem, >4 μg/ml; MIC of ertapenem, >2 μg/ml). Antibiotic susceptibilities were determined by Vitek-2 (bioMérieux Inc., Marcy l'Etoile, France), and MICs of carbapenems were determined by agar dilution according to the Clinical and Laboratory Standards Institute (CLSI) protocols (8). MICs of tigecycline and colistin and MICs of imipenem, meropenem, and ertapenem lower than 0.5 μg/ml were determined by Etest (AB Biodisk, Solna, Sweden). The criteria used for the interpretation of carbapenem MICs were based on the CLSI 2010 guidelines (9). The interpretive criterion used for tigecycline was based on FDA breakpoint values for Enterobacteriaceae that define a MIC of ≤2 as susceptible.
β-Lactamases were analyzed by analytical isoelectric focusing (IEF) (16) on crude enzyme preparations from sonicated cell cultures as described elsewhere (21). The following β-lactamases were used as controls: TEM-1, pI = 5.4; TEM-26, pI = 5.6; K1, pI = 6.5; SHV-1, pI = 7.6; P99, pI = 7.8; ACT-1, pI = 9.
The genetic relatedness between isolates was determined using pulsed-field gel electrophoresis (PFGE) as previously described (7). DNA macrorestriction patterns were compared according to the Dice similarity index (1.5% tolerance interval) (9a) using GelCompar II version 2.5 (Applied Maths, Kortrijk, Belgium). A PFGE clone was defined as a group of strains showing >85% banding pattern similarity (19).
Multilocus sequence typing (MLST) was performed on two representative E. coli clones according to the protocol at the E. coli MLST website (http://www.pasteur.fr/recherche/genopole/PF8/mlst/EColi.html).
Epidemiological links and potential contact between patients were analyzed using data on room location, consulting physicians, and other procedures performed during their hospitalization.
PCR molecular screening of β-lactamase genes and Tn4401 elements (18) was performed using the primers listed in Table Table1.1. PCR products were sized on an agarose gel and sequenced using an ABI PRISM 3100 genetic analyzer (PE Biosystems). Nucleotide and deduced protein sequences were identified using the BLAST algorithm (www.ncbi.nlm.nih.gov/).
Plasmids were purified as described previously (12) and transformed into E. coli DH10B by electroporation (Electroporator 2510; Eppendorf, Hamburg, Germany). Transformants were selected on LB agar plates containing 100 μg/ml ampicillin, and selected colonies were screened by PCR for the presence of blaKPC. Plasmid size estimation was performed by digestion of plasmid DNA prepared as described previously (7, 22), followed by S1 nuclease (190 U; Promega, Madison, WI) (1) and PFGE. Electrophoresis was carried out as described previously (7), using the Lambda ladder marker (New England Biolabs, Boston, MA).
Comparison of KPC-encoding plasmids was performed using restriction length polymorphism (RFLP) following digestion with the BglII, EcoRV, SmaI, and KpnI endonucleases (New England Biolabs, Boston, MA). Southern analysis was performed as described previously (21), using a radioactively labeled blaKPC-2 probe (892 bp) obtained with blaKPC primers (5).
Ten CREC isolates were studied. They originated from various isolation sites of 10 patients with no apparent epidemiological connection. The overall prevalence of carbapenem resistance in E. coli during this study period was 0.063% (10 cases out of 15,918 E. coli isolates). All 10 isolates were multidrug resistant (Table (Table2).2). The MIC50s and MIC90s of imipenem and meropenem were 4 and 8 μg/ml, those of ertapenem were 16 and 32 μg/ml, and those of doripenem were 1 and 4 μg/ml, respectively. All of the isolates were susceptible to tigecycline (MIC50 and MIC90 of 0.19 and 0.75 μg/ml) and to colistin (MIC50 and MIC90 of 0.125 and 0.19 μg/ml) (Table (Table22).
PFGE of the 10 CREC isolates revealed six distinct genetic clones (Fig. (Fig.1):1): four clones from 2005 (21), a new clone consisting of four isolates in 2006, and a different clone in 2007. Isolates belonging to the 2006 clone, although genetically identical, originated from four patients hospitalized in different wards during a 1-month period with no apparent epidemiological relatedness. MLST of CREC isolate 386 from 2005 identified this strain as being of sequence type 471 (ST471) reported before in France (11). CREC isolate 547, which belonged to the 2006 clone (Fig. (Fig.1),1), possessed a novel sequence type, ST39.
IEF demonstrated the production of more than one β-lactamase by each isolate (data not shown). A β-lactamase with an apparent pI of 6.7 was observed in 9 out of 10 isolates, consistent with the pI of KPC-type carbapenemases. PCR screening for β-lactamase genes, followed by sequencing, revealed the presence of blaKPC-2 in all of the strains. Six of the 10 isolates were also extended-spectrum beta-lactamase (ESBL) producers (Table (Table2).The2).The strains carrying CTX-M enzymes showed higher MICs of ceftazidime and cefepime than the non-ESBL producers (Table (Table22).
Transformation experiments were performed with eight CREC isolates (Table (Table2).2). Plasmid DNA derived from an E. coli 1679 transformant showed a plasmid size different from that of the donor, suggesting rearrangements of plasmid DNA within this strain; therefore, it was excluded from further analysis. Plasmid DNA analysis of transformants indicated that each has acquired a single plasmid (Fig. (Fig.2).2). PCR screening results of plasmid DNA confirmed the presence of blaKPC in all of them, while not all β-lactamases were transferred (Table (Table2).2). Acquisition of the blaKPC-2-containing plasmids usually elevated the MICs of cephalosporins, aztreonam, aminoglycosides, and carbapenems, yet none of the transformants presented the same level of carbapenem resistance as the respective donor strain (Table (Table22).
CREC isolates possessed two or three plasmids which differed in size (Fig. (Fig.2A,2A, lanes D). Isolates from the same year and belonging to the same clone carried highly similar-sized plasmids. Southern blot analysis of plasmid DNA from clinical isolates and their transformants showed that each clinical isolate carried a single plasmid encoding blaKPC-2 and that these plasmids varied in size, ranging from ~45 kb (Fig. (Fig.2A,2A, lanes 4 and 6) to ~100 kb (carried by E. coli strain 157 isolated in 2005) (lane 2). The plasmid DNA RFLP patterns of seven transformants, obtained by using several endonucleases, revealed different restriction patterns but a shared common region, especially between strains isolated in the same year. Southern analysis of the resulting fragments with a labeled-blaKPC-2 probe revealed the same hybridization signal, suggesting that these plasmids share a large fragment harboring blaKPC-2 (Fig. (Fig.3),3), similar to what we found previously in the 2005 isolates (21). Southern blot analysis following restriction with the SmaI endonuclease, which digests blaKPC-2 at nucleotide 790, led to two hybridization signals, suggesting the presence of a single copy of blaKPC-2 in all of the transferred plasmids (Fig. (Fig.3B3B).
PCR screening and sequencing of all Tn4401 elements (18) revealed the presence of tnpA transposase, tnpR, and the insertion sequences ISKpn6 and ISKpn7 in all of the isolates and transformants. Based on the sequence recognition of SmaI, blaKPC-2 should be digested, as well as the two genes surrounding it—IstB (part of ISKpn7) and ISKpn6—in a single site, resulting in two DNA fragments of ~1 and 1.7 kb. These two fragments were visualized by Southern hybridization (Fig. (Fig.3B),3B), which may indicate that the structure of Tn4401 in the close vicinity surrounding blaKPC-2 in our strains is conserved.
While carbapenem resistance in Enterobacteriaceae is increasing worldwide, CREC isolates are still rare. However, carbapenem resistance in E. coli is considered to be a great public health threat due to its potential to spread in hospital and community settings (29). This is the first study focusing on the molecular epidemiology and nature of carbapenem resistance in a collection of E. coli isolates within a hospital setting. This paper presents an extension of our previous study in which we first described CREC isolates residing outside the United States (21).
Isolates showed a multidrug resistance phenotype, like all KPC-producing Enterobacteriaceae; however, they possessed lower carbapenem MICs (4- to 8-fold lower) compared to the MICs of carbapenem-resistant Klebsiella pneumoniae ST258 (12). Genotyping of the isolates revealed that resistance to carbapenems in E. coli from 2005 to 2008 was not clonally related, except for four cases in 2006 that were genetically identical, but epidemiological data did not prove an apparent linkage among them. Sixty percent of the KPC-2-producing strains were also ESBL producers but apparently belonged to clones different from those described before (7).
Carbapenem resistance in E. coli during the years studied was rendered by a KPC-2 carbapenemase encoded on various-sized plasmids, which differed between clones but had regions in common. This is the first report describing the presence of Tn4401 elements in the vicinity of blaKPC-2 in E. coli previously described (18). The exact source of the blaKPC-2 gene from E. coli identified in our hospital is still uncertain. Originally, KPC-2 was detected from Enterobacter cloacae in our hospital in 2004 (6, 15), suggesting that they may have acted as a reservoir for the blaKPC-2 gene.
In contrast to epidemic K. pneumoniae clone ST258 (20), carbapenem-resistant E. coli clones did not spread significantly during the last 4 years since their emergence in our hospital or worldwide. However, the potential transfer of blaKPC-2 genes into highly fit, rapidly spreading E. coli strains is disturbing. Strict infection control policies, together with joint efforts, will aid in limiting the further dissemination of blaKPC into E. coli, the most common clinical pathogen.
This work was supported in part by European Commission FP7: SATURN—Impact of Specific Antibiotic Therapies on the Prevalence of Human Host Resistant Bacteria research grant 241796.
We thank George Jacoby, Lahey Clinic, Burlington, MA, for providing the strains used as positive controls in the IEF and PCR experiments. Doripenem was kindly provided by Karen Bush and Anne Marie Queenan, R. W. Johnson Pharmaceutical Research Institute, Raritan, NJ. We thank Tali Kotlovsky for her helpful computerized analysis of epidemiological links and potential contact between patients.
This work was performed in partial fulfillment of the requirements for an M.S. degree of Moran G. Goren, Department of Microbiology and Clinical Immunology, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel.
Published ahead of print on 15 March 2010.