Phenotypic and Enzymatic Comparative Analysis of the KPC Variants, KPC-2 and Its Recently Discovered Variant KPC-15

Sixteen different variants (KPC-2 to KPC-17) in the KPC family have been reported, and most current studies are focusing on KPC-2 and KPC-3. The KPC-15 variant, which isolated from Klebsiella pneumoniae in a Chinese hospital, was a recently discovered KPC enzyme. To compare the characteristics of KPC-15 and KPC-2, the variants were determined by susceptibility testing, PCR amplification and sequencing, and study of kinetic parameters. The strain harboring the KPC-15 showed resistance to 18 conventional antimicrobial agents, especially to cabapenem antibiotics, and the strain involving the KPC-2 also indicated resistance to cabapenem antibiotics, but both strains were susceptible to polymyxin B and colistin. The conjugation experiments showed that the changes of MIC values to the antibiotics were due to the transferred plasmids. The differences of amino acids were characterised at sites of 119 leucine and 146 lysine with KPC-15 and KPC-2. The minimum evolution tree indicated the KPC alleles evolution, and showed that the KPC-15 appeared to be homogenous with KPC-4 closely. Steady-state kinetic parameters showed the catalytic efficiency of KPC-15 was higher than that of KPC-2 for all tested antibiotics in this study. The catalytic efficiency of KPC-15 caused resistance to β-lactam antibiotics was higher than that of KPC-2. Meanwhile, an evolutionary transformation changed KPC from an efficient carbapenemase to its variants (KPC-15) with better ceftazidimase catalytic efficiency, and the old antibiotics polymyxin B and colistin might play a role in the therapy for multi-resistant strains.


Introduction
The recent emergence and spread of acquired carbapenemresistance in gram-negative bacteria, which resulting in more clinical failures in infection control, have been increasingly reported. Carbapenem-resistant Enterobacteriaceae are usually resistant not only to b-lactam antibiotics but also to most other classes of antimicrobial agents [1].
Resistance to carbapenems in gram-negative bacteria is frequently observed due to the high expression of AmpC blactamases, which are associated with the loss of outer membrane proteins [2], the increased expression of efflux pumps systems [3], and with changes in the affinity of penicillin-binding proteins (PBPs) for carbapenems [4]. Carbapenem resistance can primarily be mediated by carbapenem-hydrolysing enzymes, such as metallo-b-lactamases (MBLs), Guiana extended-spectrum (GES) and Klebsiella pneumoniae carbapenemases (KPCs) [5][6][7].
Currently, sixteen different variants (KPC-2 to  in the KPC family have been reported, and most current studies are focusing on KPC-2 and KPC-3 [21][22][23]. In the study, we characterised phenotypic and enzymatic comparative analysis for the KPC-2 enzyme and novel KPC-15 variant which was discovered recently by us [24].

Materials and Methods
The ethics committee approval This study was approved by the the Ethics Committee of Municipal Hospital of Taizhou University, and the written informed consents obtained from each of the participants. The right of research objects was protected, and we agreed this study was conducted in our hospital.

Bacterial strains
The two strains of K. pneumoniae, which were numbered Kp1769 and Kp1241, were isolated from samples of sputum and blood, respectively. The strain Kp1241 was isolated from the hepatobiliary unit in June 2012, and the strain Kp1769 was isolated from the ICU in September 2011 in respiration unit in the Taizhou Municipal Hospital of Taizhou University. The strain Kp1769, which was confirmed involving the bla KPC-2 gene, was used to compare the characteristics of the kinetic parameters with the recently discovered bla KPC-15 variant in this study. These two strains were assigned to K. pneumoniae using a Vitek GNI + card (bioMérieux, France), and species identification was confirmed by a sequence analysis of the 16 S-23 S rRNA (Midi Labs, USA).
According to a previously reported protocol [26], conjugation experiments were performed in lysogeny broth (LB) with the strain E. coli J53Az R (resistance to sodium azide) as the recipient and with the strains Kp1241 and Kp1769 as the donor strains. Donor and recipient cells in logarithmic phase (0.5 mL of each) were added to 4 mL of fresh LB, which was followed by incubation at 35uC for 18-24 h without shaking. The transconjugants were selected on trypticase soy agar (TSA) plates containing sodium azide (0.3 g/L) and imipenem (0.02 g/L) following incubation for 18-24 h at 35uC. PCR amplification and sequencing PCR was performed for the amplification and sequencing of pre-bla KPC-15 , the bla KPC-15 variant, and its surrounding genes (bla TEM , tnpA and tnpR) in K. pneumonae using the primers shown in Table 1. PCRs were also performed for the amplification of the transformant conjugated with E. coli J53Az R . The primers were synthesised by Sangon Biological Technology (Shanghai, China). The PCR conditions for pre-bla KPC-15 and the bla KPC-15 variant were as follows: 3 min at 94uC and 30 cycles of 1 min at 94uC, 1 min at 52uC, and 1 min at 72uC, followed by an elongation step for 10 min at 72uC, which produced a band of ca. 1,431 bp for bla KPC-15 , encompassing the entire KPC coding region. The same PCR conditions were used to amplify bla TEM , tnpA (ISKpn8 and ISKpn6-like transposases) and tnpR genes. The primers of tnpA 1 , tnpA 2 and tnpA 3 were used to amplify the ISKpn8 transposon, and the tnpA primer was used to amplify the ISKpn6-like transposon. The positive PCR products were also sequenced by Sangon Biological Technology (Shanghai, China). Then, the sequencing results were assembled using the contigExpress Program software and the sequences were obtained for the genetic environment of the bla KPC-15 variant from the strain Kp1241.
Plasmids were isolated from the strain Kp1241 using an Axygen kit (USA) in accordance with the manufacturer's instructions. Plasmids were then separated by electrophoresis through a 0.6% agarose gel in the presence of 16TBE buffer at a constant voltage of 90 V for 90 min. These plasmid DNA fragments served as samples for the amplification of bla KPC-15 gene by PCR, where the initial position of the bla KPC-15 gene plasmid was determined. The estimated sizes of plasmid DNAs were determined according to reference [26]. Based on these data, we mapped the genetic sequence of the bla KPC-15 upstream and downstream regions and determinded the start sites of transcription and the promoter regions of relevant genes.
b-lactamase preparation and the kinetic parameters for a comparison of KPC-15 and KPC-2 Cultures of K. pneumoniae harboring bla KPC-15 and bla KPC-2 were grown overnight at 37uC in 100 ml of trypticase soy broth with amoxicillin (100 mg/mL). Bacterial suspensions were disrupted by sonication (20 s at 20 KHz64 times) and centrifuged (30 min, 20,000 g, 4uC). The supernatants, which contained the crude enzyme extracts, were collected. Then, the crude enzyme extracts were mixed with 50 mM Tris-Cl, pH 7.4, with 1 mM magnesium sulphate, and lysed with 40 mg/L lysozyme. Next, 1.0 U/ml benzonase nuclease was added to digest nucleic acids, and a 2.0 mM concentration of EDTA was added to complete the periplasmic fractionation. The lysed cells were centrifuged at 12,000 rpm for 10 min to remove the cellular debris. The supernatant was further enriched and purified for b-lactamase by using preparative isoelectric focusing as described previously [27,28]. The b-lactamase purification step was performed in accordance with reference [29].
Steady-state kinetic parameters were determined using a diode array spectrophotometer (Agilent, USA). Each assay was performed with 10 mM PBS, pH 7.4, at 25uC. In all assays, the enzyme was maintained at 10 nM, whereas substrate concentrations were varied from 50 to 200 mM. Kinetic assays were performed under steady-state conditions to determine the kms and kcats of the antimicrobial agents: imipenem, meropenem, ceftaz- idime, cefotaxime, cefepime, aztreonam and nitrocefin for the enzyme according to a previously established model represented in the kinetic parameter equation [30,31] and in reference [29]. Concentrations varied from 25 to 250 mM for imipenem; 50 to 250 mM for meropenem; 150 to 500 mM for ceftazidime; 100 to 500 mM for cefotaxime; 250 mM to 1.5 mM for cefepime; and 50 mM to 500 mM for aztreonam. A final concentration of 100 mM nitrocefin was used as the reporter substrate.

Antimicrobial susceptibility testing and conjugation experiment
The MICs of 20 antimicrobial agents for the strains Kp1241 and Kp1769 were depicted in Table 2. The results suggested that the strain Kp1241 was resistant to 18 antimicrobial agents, and particularly exhibited high MIC values for imipenem, meropenem, and ertapenem at 16, 16, and 32 mg/mL, respectively; however, the strains Kp1769 showed MIC values for imipenem, meropenem, and ertapenem at 2, 2, and 4 mg/mL, respectively. And the strain Kp1241 remained susceptible to polymyxin B and colistin, so it was as the strain Kp1769. By the conjugation experiments, the transconjugant of strain Kp1241 exhibited a phenotype of resistance to imipenem, meropenem, ertapenem, cephalosporin, aminoglycosides, ampicillin/sulbactam, piperacillin/tazobactam and cefoperazone/sulbactam, but, the transconjugant of strain Kp1769 indicated intermedium values to imipenem, meropenem, and ertapenem. Hence, the results of conjugation experiments showed that the changes of MIC values of the antibiotics were due to the transferred plasmids.
Amino acid variations, the minimum evolution tree of amino acid sequences and the kinetic parameters for KPC-15 and other KPC enzymes Amino acid variations between novel bla KPC-15 and other bla KPC genes were listed in Figure 1, and basing on the data of Figure 1, we drew the minimum evolution tree of amino acid sequences for KPC-2 to KPC-17 ( Figure 2) using MEGA 5.05 software. The tree obviously indicated the KPC alleles evolution, and showed that the KPC-15 appeared to be homogenous with KPC-4 closely.
Steady-state kinetic parameters were measured five times and averaged as shown in Table 3. Consistent with the MIC data, the catalytic efficiency (kcat/km ratio) of the KPC-15 enzyme was lowest for ceftazidime, at 0.0142 mM 21 s 21 , and was higher at 5.26-fold compared with that of the KPC-2 enzyme, at 0.0027 mM 21 s 21 . However, the catalytic efficiency of the KPC-15 enzyme was highest for nitrocefin, at 9.2 mM 21 s 21 , and was a little higher at 1.  Table S1. In general, the catalytic efficiency of KPC-15 was higher than that of KPC-2 for all tested antibiotics in this study (Table 3).

PCR sequencing analysis and genetic organisation
The analysis of the 8,997 bp nucleotide sequence for Kp1241 was shown in Figure 3. There were 5 different genes in the sequence, including the genes tnpR, bla TEM-12 and bla KPC-15 , and the transposons ISKpn8 and ISKpn6-like [24]. The sites of the transcriptional promoters for the bla TEM-12 and bla KPC-15 genes were indicated as +1 and the 210 and 235 regions were also shown; however, there were no obvious transcriptional promoters in tnpR, ISKpn8 and ISKpn6-like transposons. Between the tnpR gene and the ISKpn8 transposon, there was a 109 bp nucleotide interval; the ISKpn8 transposon nucleotide sequence was 3,020 bp. Between the ISKpn8 transposon and the bla TEM-12 gene, there was a 206 bp nucleotide interval, which included the site of the transcriptional promoter for the bla TEM-12 gene, +1 (g, start codon), 210 region (tataac) and 235 region (ttattg). Between the bla TEM-12 and bla KPC-15 genes, there was a 223 bp nucleotide interval, which included the site of the transcriptional promoter for the bla KPC-15 gene, +1 (g, start codon), 210 region (gattac) and RBS (aaggaa, the ribosome binding site), but no obvious 235 region was shown in the bla KPC-15 gene. Between the bla KPC-15 gene and the ISKpn6-like transposon, there was a 249 bp nucleotide interval, and the ISKpn6-like transposon had a 1,320 bp nucleotide sequence.

Discussion
In recent years, many reports have described the presence of KPC b-lactamases (particularly KPC-2), and discovered they can cause resistance to carbapenem or reduce susceptibility to carbapenem in strains of the family of Enterobacteriaceae from different areas in China [5,16,32]. It seems that KPC enzymes are becoming widely disseminated in China, and plasmids of different sizes may harbor the bla KPC-2 gene [33]. The genetic environments of the bla KPC gene have been characterised in some studies, and various transposon elements seems to be responsible for the rapid spread of bla KPC [33][34][35].
In this study, the susceptibility tests indicated that the strain Kp1241 showed resistance to quinolones, aminoglycosides, and the beta-lactam antibiotics, but susceptibility to polymyxin B and colistin with MICs of 0.5 and 1 mg/mL (Table 2), respectively. This result suggested that the old antibiotics, polymyxin B and colistin, might play an important role in the treatment of the strain Kp1241 [36].
Comparing the amino acids with other KPCs, KPC-15 showed more than two sites of amino acid changes (A119L and G146 K), and emerged to be homogenous with KPC-4 closly (Figures 1 and  2). The changes of the amino acids would influence the kinetic properties of the KPC enzymes. In spite of different values for the kms and kcats, the catalytic efficiency (kcat/km) of KPC-15 was higher than that of KPC-2 for all tested antibiotics (Table 3). These data implied that the effect of KPC-15 caused resistance to b-lactam antibiotics (including imipenem, meropenem, ceftazidime, cefotaxime, aztreonam and cefazolin) was higher than that of KPC-2. Moreover, an evolutionary transformation changed KPC from an efficient carbapenemase to its variants (KPC-15) with better ceftazidimase catalytic efficiency. Because of the multiresistance to quinolones or to aminoglycosides antibiotics, in addition to carbapenemase, the strain Kp1241 might involve other genes, such as quinolone or aminoglycoside genes.
According to the literature, there is a distinct genetic environment with the chimera of several transposon-associated elements in China, such as Tn3, ISKpn8 and an ISKpn6-like element on plasmid pKP048 [33]. With conjugation and gene mapping studies, we acquired three plasmids of different lengths. The bla KPC-15 variant is on a ca. 73-kb plasmid and carries ISKpn8 and an ISKpn6-like element determinant [24]. The transformant was also much less resistant than the original isolate, it was most likely due to the lack of any permeability lesion. The genetic environment of the bla KPC-15 gene in the plasmid showed a structural and context similarity to those characteristics found in plasmid pKP048 [33], except that a truncated bla TEM-12 gene and the bla KPC-15 variant were inserted and located between ISKpn8 and the novel bla KPC-15 variant [24]. The nucleotides upstream of the bla KPC-15 translational start codon were also shown in Figure 3. The start codon was indicated as G by the +1 residue, and the putative 210 promoter elements were shown as gattaa and labeled as 210 below. These data were consistent with a study by Yigit et al. [8]; however, no obvious 235 region was discovered in the promoter region, which was different from previous report [8]. Only a 233 bp sequence was discovered between the bla TEM-12 gene and the bla KPC-15 variant in this study (Figure 3).

Conclusions
The study suggested that old antibiotics (polymyxin B and colistin) might play a role in therapy. The novel KPC-15 had more than two sites of amino acid changes (A119L and G146 K), and was homogenous with KPC-4 closly. The catalytic efficiency of KPC-15 causing resistance to b-lactam antibiotics was higher than that of KPC-2. Moreover, an evolutionary transformation changed KPC from an efficient carbapenemase to its variants (KPC-15) with better ceftazidimase catalytic efficiency.