Klebsiella pneumoniae carbapenemase (KPC)-encoding genes containing promoter-deletions (blaKPC-2a, blaKPC-2b, and blaKPC-2c) have disseminated in Enterobacteriaceae. The minimal inhibitory concentrations (MICs) to β-lactams in clinical KPC-producing Enterobacteriaceae range from susceptible to high-level resistant, resulting in diagnostic problems. To better understand the variability in β-lactam MICs among KPC-producing Enterobacteriaceae, three isoforms of blaKPC-2 gene were used to transform Escherichia coli W4573 and its deletion mutant of an efflux pump (AcrAB) to examine the effects on β-lactam susceptibility. MICs to β-lactams in E. coli W4573 and its acrAB mutant strain increased 1- to 500-fold (MIC from 0.125 to 64 μg mL−1 of aztreonam) in the blaKPC-2a, blaKPC-2b, and blaKPC-2c transformants compared with the cloning vector alone. However, transformants of the acrAB mutant strain remained susceptible to all β-lactams tested except for aztreonam and carbenicillin. Levels of the three promoters’ length and carbapenemase activities in the transformants harboring the blaKPC-2a, blaKPC-2b, and blaKPC-2c were correlated to the levels of β-lactam MICs in both E. coli W4573 and its mutant of an efflux pump (AcrAB). Overall, these results suggest that promoter-deletions of blaKPC-2 gene and AcrAB may be associated with the variability in β-lactam MICs in KPC-producing Enterobacteriaceae.
Enterobacteriaceae are a family of Gammaproteobacteria and are part of the normal intestinal flora. Several species of Enterobacteriaceae such as Escherichia coli and Klebsiella pneumoniae are occasional human pathogens and have been associated with a variety of diseases including urinary tract infections, pneumonia, and septicemia (Nordmann et al., 2011). Carbapenem antibiotics (imipenem and meropenem) are frequently recommended as first-line therapy for Enterobacteriaceae infections. However, the emergency of carbapenem-resistant Enterobacteriaceae associated with Klebsiella pneumoniae carbapenemase (KPC) has resulted in the carbapenem antibiotics becoming less effective (Gupta et al., 2011; da Silva et al., 2012). The first KPC-encoding gene (KPC-1) was discovered in North Carolina (Yigit et al., 2001). Since then, several variants of KPC-1 with amino acid substitutions have been reported worldwide in E. coli, K. pneumoniae, Salmonella cubana, Enterobacter cloacae, and Proteus mirabilis (Nordmann et al., 2009). KPC-1 and KPC-2 have essentially identical amino acid sequences and are the most frequent carbapenemase identified in Enterobacteriaceae (Yigit et al., 2001). Infections caused by carbapenem-resistant Enterobacteriaceae have become a major treatment challenge and have been associated with high mortality rates (Logan, 2012; Rapp & Urban, 2012).
The gene encoding KPC-2 (blaKPC-2) is located on transferable plasmids within the conserved Tn3 family transposon Tn4401. The transposon Tn4401 (c. 10-kb) is flanked by a 5-bp target site of duplication and possesses genes encoding transposase and resolvase with two novel insertion sequences (ISKpn6 and ISKpn7) and 39-bp imperfect inverted-repeat sequences. The gene blaKPC-2 is located between the two insertion sequences of Tn4401. Three isoforms of Tn4401 (Tn4401a, Tn4401b, and Tn4401c) have been reported based on a putative promoter length of the blaKPC-2 (Nordmann et al., 2009). A putative promoter of blaKPC-2 from Tn4401a (blaKPC-2a) is 100 bp shorter (a deletion) than the promoter from Tn4401b (blaKPC-2b). The same putative promoter of blaKPC-2 of Tn4401c (blaKPC-2c) is 200 bp shorter (a deletion) than that from Tn4401b (Naas et al., 2008, 2012; Gootz et al., 2009; Nordmann et al., 2009). Tn4401a has been identified from K. pneumoniae isolated from Greece and from New York City hospitals. Tn4401b has been identified from K. pneumoniae and Pseudomonas aeruginosa isolates from Colombia and is the most frequent structure encountered in the USA. Tn4401c was identified in E. coli isolated from a Parisian hospital (Naas et al., 2008; Nordmann et al., 2011).
KPC enzymes are Class A serine β-lactamases and can hydrolyze all classes of β-lactam antibiotics, including carbapenems. However, KPC-producing Enterobacteriaceae have been reported as having widely varied MICs to β-lactams ranging from susceptible (MIC < 1 μg mL−1) to very highly resistant (MIC > 256 μg mL−1; Anderson et al., 2007; Nordmann et al., 2009). It has been reported that impaired outer membrane proteins in KPC-producing isolates are required for high-level carbapenem resistance (Cai et al., 2008; Landman et al., 2009). Landman et al. (2009) reported 20 clinical isolates of K. pneumoniae exhibiting two clonal groups of 11 carbapenem-resistant isolates harboring carbapenemases (blaKPC-2) with variable MICs to carbapenem (8–1024 μg mL−1). They claimed that carbapenem MICs were increased (or correlated) with decreasing expression of the outer membrane protein ompK36. Although their studies clearly demonstrated that lower levels of expression of ompK36 were associated with high-level carbapenem resistance, the level of ompK36 expression in some carbapenem-resistant isolates was similar or even less than that of the carbapenem-susceptible isolates. In addition, the expression levels of blaKPC-2 in some carbapenem-resistant isolates were inconsistently related to carbapenem MICs. These observations suggest that additional mechanisms are also responsible for the variability of carbapenem MICs in KPC-producing isolates. One of such possible mechanisms is differential gene expression of the three blaKPC-2 genes (blaKPC-2a, blaKPC-2a, and blaKPC-2c), based on the different lengths of their putative promoter regions. Another possible mechanism includes the efflux pump (AcrAB), which is a well-known multidrug efflux system associated with resistance to a wide range of antibiotics in Enterobacteriaceae (Nikaido, 1996; Nikaido & Pages, 2012). Molecular details of the differential gene expression of the blaKPC-2 genes and involvement of AcrAB in relation to β-lactam susceptibility remain unclear in KPC-producing Enterobacteriaceae. This study characterized four carbapenem-resistant isolates of K. pneumoniae reported by Landman et al. (2009). Two isoforms of blaKPC-2 (blaKPC-2a and blaKPC-2b) were cloned from the carbapenem-resistant isolates and the third blaKPC-2c was constructed in vitro by a recombinant PCR method. The three isoforms of blaKPC-2 gene (blaKPC-2a,blaKPC-2b, and blaKPC-2c) were expressed in E. coli lacking AcrAB and its parental strain to better understand the mechanisms of variable β-lactam susceptibility in KPC-producing Enterobacteriaceae.
Materials and methods
Clinical isolates and culture conditions
Four carbapenem-resistant (CI516, CI839, KC850, and WO555) clinical isolates of K. pneumonia reported previously (Landman et al., 2009) were obtained from Downstate Medical Center (State University of New York, Brooklyn, NY, USA). Escherichia coli W4573 and its mutant E. coli N43 (acrA) were obtained from Yang et al. (2003). Escherichia coli DH5α was also used for general cloning experiment. All clinical isolates and E. coli strains were routinely grown in Luria–Bertani (LB) broth or LB agar plates. All antibiotics and chemicals used in this study were purchased from Sigma-Aldrich (St. Louis, MO) and added in culture media when needed.
Genotype analysis was performed by PCR-based RAPD (random amplified polymorphic DNA) fingerprinting analysis as described. Briefly, genomic DNA was extracted from each clinical isolate using Qiagen DNeasy kits (Valencia, CA) and used for the RAPD analysis. Each PCR reaction mixture (100 μL) included 100 ng of genomic DNA, arbitrary primers (primer 271: 5′-AGCGGGCCAA-3′ or primer 272: 5′-CCGGGCAAGC-3′; 0.5 μM each), 200 μM dNTP, 10 mM Tris (pH 8.0), 50 mM KCl, 3 mM MgCl2, 0.001% gelatin, 1% DMSO, and Taq polymerase (2 units; Promega). Each reaction was amplified with an MJ Research DNA Thermal Cycler model PTI-100 (Watertown, MA) using the amplification cycle: (1) four cycles, with one cycle consisting of 5 min at 94 °C, 5 min at 36 °C, and 5 min at 72 °C and (2) 30 cycles, with one cycle consisting of 1 min at 94 °C, 1 min at 36 °C, and 2 min at 72 °C, followed by a final extension step at 72 °C for 10 min.
MLST (multilocus sequence typing) analysis
MLST was performed for carbapenem-resistant K. pneumoniae as described on the K. pneumoniae MLST website (http://www.pasteur.fr/recherche/genopole/PF8/mlst/Kpneumoniae.html). Seven genes (rpoB, gapA, mdh, pgi, phoE, infB, and tonB) were amplified as suggested by the website and PCR fragments were used to determine nucleotide sequences (GENEWIZ, Inc., South Plainfield, NJ). Nucleotide sequences for the seven genes were submitted to the MLST sequence type database.
Antibiotic susceptibility testing
Antibiotic susceptibility was determined as minimal inhibitory concentrations (MICs) using the broth dilution method following the guidelines of the Clinical and Laboratory Standards Institute (CLSI, 2010) and Kwon & Lu (2006a, b).
Detection of genes encoding carbapenem-hydrolyzing β-lactamase and cloning carbapenemase-encoding genes
The genes encoding carbapenemase (blaKPC-2) were detected by PCR as described (Landman et al., 2009). Genomic DNA extracted from carbapenem-resistant isolates (CI516, CI839, KC850, and WO555) was used to amplify blaKPC-2 genes encompassing 500 bp upstream from a start codon and 50 bp downstream from a stop codon of KPC-2 (kpc1f: 5′-accctgacagccgcgatgctggat-3′, kpc1r: 5′-gcggtggtgggccaatagat-3′ from GenBank accession number FJ223605). The amplified PCR fragments (1400-bp) were used to determine the nucleotide sequence and inserted into a cloning vector pQF50bla::Km, which was a modified vector from pQF50 (Kwon & Lu, 2006a, b) and inactivated a bla gene by a kanamycin = resistant cassette (Aranda et al., 2010). To construct the gene blaKPC-2c, two PCR fragments were amplified by primer pairs of kpc1f/kpc2cr (fragment-A: 5′atattccttgtttgaaggtggagttacggacggcctcaggaag-3′) and kpc1r/kpc2cf (fragment-B: 5′-cttcctgaggccgtccgtaactccaccttcaaacaaggaatat-3′). The PCR primers of kpc2cf and kpc2cr were designed from GenBank accession number FJ223606 and nucleotide sequences of the primers were overlapped (22-bp) for fragment-A and fragment-B. Both fragment-A and fragment-B were used as DNA templates for a second PCR amplification by priming kpc1f/kpc1r. The PCR fragment from the second amplification was used to determine nucleotide sequence and inserted into the cloning vector pQF50bla::Km and named pKP26. PCR reaction and amplification conditions were followed as described (Kwon & Lu, 2007).
Construction of blaKPC-2 promoter::lacZ transcriptional fusions and β-galactosidase assay
A putative promoter for the blaKPC-2 genes was amplified by a PCR primer pair (forward: 5′-accctgacagccgcgatgct-3′/reverse: 5′-atacagtgacatcaacgata-3′) with genomic DNA of clinical isolates (CI516 for blaKPC-2a and CI839 for blaKPC-2b) and plasmid DNA carrying blaKPC-2c (pKP26). The PCR primer pair was designed to amplify the DNA fragment encompassing upstream 500 bp from a start codon of the blaKPC-2b (GenBank accession number EU176011.1). Each of PCR fragments was purified by QIAGEN spin columns (Chatsworth, CA) and inserted into the SmaI site of a broad-host-range transcriptional lacZ fusion vector, pQF50 (Kwon & Lu, 2006a,b). The orientation and the DNA sequence identity of each insert were confirmed by nucleotide sequence determination (GENEWIZ, Inc., South Plainfield, NJ). The resulting recombinant plasmids and a cloning vector were introduced into E. coli W4573 and N43. The cells harboring each plasmid were grown in LB until the optical density at 600 nm reached 0.6–0.7. Then, the cells were used to determine β-galactosidase activity by Miller's method (Miller, 1972).
Carbapenemase activity was measured as reported previously (Kwon & Lu, 2006a, b) with minor modifications for cell-free crude extract. Briefly, the cells were harvested, washed, and resuspended in 5 mL of 50 mM sodium phosphate buffer (pH 7.0). The cells were broken by sonication (Misonix Sonicators, Newtown, CT; 37% amplitude [15 W], 10 s, 10 times with 30-s cooling between bursts, until the cell extract was clear), and the cell-free crude extract was collected after centrifugation at 15 000 g for 5 min. The protein concentration of the crude extracts was measured by the Bradford method (Bradford, 1976) using bovine serum albumin as the standard. The enzymatic activity of carbapenemase was determined using nitrocefin (Oxoid) as the substrate. The reaction mixture (1 mL) contained 100 μg mL−1 of nitrocefin in 0.1 M sodium phosphate buffer (pH 7.0). The activity of carbapenemase was monitored at 30 °C by absorbance changes at 486 nm in a DU®730 UV/Visible spectrophotometer (Beckman Coulter). One unit of activity was defined as the amount of carbapenemase that digests 1 μmol of nitrocefin per minute at 37 °C. A molar extinction coefficient of 20 500 M−1 cm−1 was used for nitrocefin.
Results and discussion
Antibiotic susceptibility, genotype analysis, and MLST
The four clinical isolates obtained from Downstate Medical Center (Brooklyn, NY) were used to confirm their antibiotic susceptibility and genotypes as shown in the previous study (Landman et al., 2009). The isolates were resistant to aztreonam, carbenicillin, ceftazidime, imipenem, meropenem, gentamicin, and ciprofloxacin (Table 1). The isolates exhibited two different genotype patterns (clonal group A for CI516 and KC805; clonal group B for CI839 and WO555; Fig. 1) consistent with the previous observations (Landman et al., 2009). To further characterize the isolates, MLST analysis was performed and revealed that sequence types of clonal groups A and B were 42 and 258, respectively. Sequence type 258 has been suggested to be a dominant carrier of KPC-producing genes and has disseminated worldwide (Kitchel et al., 2009; Samuelsen et al., 2009). The sequence type 42 was identified in isolates from Turkey and UK (Diancourt et al., 2005), suggesting that KPC-encoding genes may be disseminating as multiple sequence types (or clonal groups).
Table 1. Antibiotic susceptibility of clinical isolates of Klebsiella pneumonia
Detection and cloning of carbapenemase-encoding genes
The presence of the blaKPC-2 genes in the four carbapenem-resistant isolates was confirmed by PCR methods as described previously (Landman et al., 2009). The full length of the genes including a putative promoter region was cloned into pQF50-bla::Km as described in 'Materials and methods'. Nucleotide sequence analysis of the inserts revealed that the deduced amino acid sequences of all four genes were identical to that reported for KPC-2 (Naas et al., 2008); however, the putative promoter sequences were of two different types. The putative promoter sequence of the blaKPC-2 genes from CI516, KC805, and WO555 was 100% identical to that of GenBank accession number EU176014.1 whereas the sequence from CI839 was 100% identical to GenBank accession number EU176011.1. These results imply that the isolates of CI516, KC805, and WO555 harbor blaKPC-2a genes which have a 100-bp deletion in their promoter regions compared with that of the isolate CI839, which harbors blaKPC-2b. The gene blaKPC-2a is found in both clonal groups, whereas the gene blaKPC-2b was only found in clonal group B, suggesting that the KPC-encoding genes are not associated with a specific clonal group or sequence type, possibly because the KPC-encoding genes reside in a transposable genetic element (transposon Tn4401) and can be nonspecifically transferred to other isolates (Kitchel et al., 2009).
Effects of putative promoter length of the blaKPC-2 genes on β-lactam susceptibility
To understand the effect of the three different promoter lengths on β-lactam susceptibility, the third gene of blaKPC-2 (blaKPC-2c) was constructed as described in 'Materials and methods' and the three genes (blaKPC-2a, blaKPC-2b, and blaKPC-2c) were used to transform E. coli W4573. MIC levels of the transformants harboring the blaKPC-2b were over 4-fold higher than those of the blaKPC-2a, which in turn were much higher (up to 500-fold; from 0.25 to 128 μg mL−1) to all β-lactam antibiotics tested compared with the same cells harboring only the cloning vector (Table 2). MIC levels of the transformant harboring the blaKPC-2c were also higher (2- to 16-fold) than those of the same cell harboring a cloning vector alone, but the MIC levels were much lower than those of cells containing blaKPC-2a or blaKPC-2b (Table 2). These results suggest that the putative promoter length in Enterobacteriaceae is related to the MIC levels to β-lactam antibiotics.
KPC-producing Enterobacteriaceae have spread worldwide but are difficult to detect based on β-lactam susceptibility testing (i.e. c. 25% to 87% of KPC-producing isolates are susceptible to β-lactams; Marchaim et al., 2008; Endimiani et al., 2009; Nordmann et al., 2009; Landman et al., 2010). A previous study showed that decreased expression of ompK36 was one of the factors contributing to β-lactam susceptibility in KPC-producing K. pneumoniae (Landman et al., 2009). This study suggests that promoter deletions of the blaKPC-2 genes are also associated with the variability of β-lactam susceptibility. This finding is consistent with the notion that clinical isolates harboring promoter deletions of the KPC-encoding genes exhibit lower MICs to β-lactams (Gootz et al., 2009; Nordmann et al., 2011). However, the isolates harboring blaKPC-2a (KC805 and WO555) showed much higher carbapenem MICs than that of the isolate harboring blaKPC-2b (CI839), which is likely due to the impaired outer membrane proteins, as suggested by Landman et al. (2009).
Efflux pump (AcrAB) and β-lactam susceptibility produced by the blaKPC-2 genes
The mutant E. coli N43 lacking acrAB was transformed by the three blaKPC-2 genes and used to determine MIC levels to β-lactam antibiotics. MIC levels of the transformants harboring blaKPC-2a and blaKPC-2b were higher (up to 500-fold; from 0.125 to 64 μg mL−1) to all β-lactam antibiotics tested compared with the same mutant strain harboring the cloning vector alone. MIC levels of the transformant harboring blaKPC-2c were similar to those of the same cells harboring only the cloning vector. However, all transformants harboring one of the three genes remained susceptible to β-lactam antibiotics (MIC levels < 8 μg mL−1), with the exception of the transformant harboring blaKPC-2a (MIC 128 μg mL−1 to carbenicillin) and the transformants harboring blaKPC-2b (MIC 64 or 128 μg mL−1 to aztreonam or carbenicillin, respectively). An intact acrAB clone restored MIC levels of the mutant strain to those those similar to its parental strain. MIC levels to tetracycline of the transformants harboring the three blaKPC-2 genes were unchanged (Table 2). These results suggest that the efflux pump (AcrAB) is required for the full range of resistance as well as for the variation in β-lactam MICs in KPC-producing Enterobacteriaceae.
Transcriptional levels and carbapenemase activity produced by the blaKPC-2 genes
Transcriptional levels of the three blaKPC-2 genes were examined using promoter::lacZ transcriptional fusion assays as described in 'Materials and methods'. Results revealed that the β-galactosidase activity of the blaKPC-2b promoter::lacZ fusion was c. 2.2-fold higher than that of the blaKPC-2a promoter::lacZ fusion and c. 6-fold higher than that of the blaKPC-2c promoter::lacZ fusion in E. coli W4573 (Fig. 2). Similar results were observed in E. coli N43 (data not shown). These results suggest that the transcriptional levels of the three genes correlated with the deletions in the putative promoter.
The β-lactam MICs of the acrA mutant strain harboring the blaKPC-2 genes were significantly lower than those of its parental strain harboring the same blaKPC-2 gene, suggesting that the lower β-lactam MICs might be associated with lower carbapenemase activity. To test this possibility, carbapenemase activity was measured from the mutant and its parental strains harboring each of the three genes. As shown in Fig. 3, carbapenemase activities produced by blaKPC-2a, blaKPC-2b, and blaKPC-2c were c. 5, 13, and 2 μmol min−1 mg−1 protein in each of the mutant and parental strains, respectively. These results suggest that carbapenemase activity produced by the blaKPC-2 genes is unlikely to be responsible for the lower levels of β-lactam MICs in the mutant strains.
Transcriptional levels and carbapenemase activities from the mutant and its parental strains harboring the blaKPC-2 genes were similar, suggesting that mutation on acrAB is more likely to be associated with the lower MICs to β-lactams than with the carbapenemase activity. However, the high-level β-lactam resistance in Enterobacteriaceae may require both AcrAB and blaKPC-2 genes. Interplay between multidrug efflux systems and other resistant factors such as β-lactamase, porin, target mutations, and lipopolysaccharide is commonly associated with high levels of resistance to antibiotics in Gram-negative clinical isolates (Quale et al., 2006; Kallman et al., 2009; Lee & Ko, 2012; Singh et al., 2012; Yamasaki et al., 2013). Overall results suggest that the variability of β-lactam MICs, ranging from susceptible to high-level resistant, in KPC-producing Enterobacteriaceae may be associated with the promoter deletions of blaKPC-2 genes as well as with alterations of AcrAB.
This work was supported by a grant from the National Institutes of Health (5SC3 GM094053). The authors would like to thank Drs. J. Quale and D. Landman (SUNY Downstate Medical Center, Brooklyn, NY) providing clinical isolates of K. pneumoniae.