Characterization of BKC-1 Class A Carbapenemase from Klebsiella pneumoniae Clinical Isolates in Brazil

Abstract

Three Klebsiella pneumoniae clinical isolates demonstrating carbapenem resistance were recovered from different patients hospitalized at two medical centers in São Paulo, Brazil. Resistance to all β-lactams, quinolones, and some aminoglycosides was observed for these isolates that were susceptible to polymyxin B. Carbapenem hydrolysis, which was inhibited by clavulanic acid, was observed for all K. pneumoniae isolates that belonged to the same pulsed-field gel electrophoresis (PFGE) type and a novel sequence type (ST), ST1781 (clonal complex 442 [CC442]). A 10-kb nonconjugative incompatibility group Q (IncQ) plasmid, denominated p60136, was transferred to Escherichia coli strain TOP10 cells by electroporation. The full sequencing of p60136 showed that it was composed of a mobilization system, ISKpn23, the phosphotransferase aph3A-VI, and a 941-bp open reading frame (ORF) that codified a 313-amino acid protein. This ORF was named bla_BKC-1. Brazilian Klebsiella carbapenemase-1 (BKC-1) showed a pI of 6.0 and possessed the highest identity (63%) with a β-lactamase of Sinorhizobium meliloti, an environmental bacterium. Hydrolysis studies demonstrated that purified BKC-1 not only hydrolyzed carbapenems but also penicillins, cephalosporins, and monobactams. However, the carbapenems were less efficiently hydrolyzed due to their very low k_cat values (0.0016 to 0.031 s⁻¹). In fact, oxacillin was the best substrate for BKC-1 (k_cat/K_m, 53,522.6 mM⁻¹ s⁻¹). Here, we report a new class A carbapenemase, confirming the diversity and rapid evolution of β-lactamases in K. pneumoniae clinical isolates.

INTRODUCTION

The development of antimicrobial agents represents one of the most significant achievements in modern medicine. Among the antimicrobial classes, the β-lactams have exerted a decisive impact on human health due to their extraordinary efficacy and safety (1, 2). However, the dissemination of extended-spectrum β-lactamases (ESBLs), mainly in Enterobacteriaceae, has compromised the use of penicillins and broad-spectrum cephalosporins, leading to an increasing therapeutic use of carbapenems (3, 4).

Carbapenems are stable against most β-lactamases produced by Gram-negative bacteria; however, reports of β-lactamases capable of hydrolyzing carbapenems have increased worldwide (5). These enzymes are classified according to their active site and functional structure requirements, and they are grouped into metallo-β-lactamases (MBLs) belonging to Ambler class B or into serine β-lactamases (Ambler classes A and D) (3, 5, 6). The emergence of carbapenemases is of great concern because these enzymes are generally capable of inactivating almost all β-lactams (7). In addition, carbapenemase-encoding genes are usually carried by mobile genetic elements that also harbor other resistance determinants, facilitating their spread and limiting therapeutic options (8). Moreover, phenotypic detection by clinical laboratories of carbapenemase production in Enterobacteriaceae may be compromised, since these isolates are frequently miscategorized as being carbapenem susceptible (7, 8).

The aim of this study was to investigate the mechanisms of carbapenem resistance in three Klebsiella pneumoniae clinical isolates. These investigations led to the characterization of a new plasmid-encoded class A carbapenemase named BKC, or Brazilian Klebsiella carbapenemase.

MATERIALS AND METHODS

Bacterial strains.

K. pneumoniae clinical isolates KP60134, KP60135, and KP60136 were recovered from two different hospitals located in the city of São Paulo, Brazil. These isolates were subcultured in 2008 from blood, lower respiratory tract, and perirectal swab cultures from three different patients. These isolates were referred to our facility for further investigation of carbapenem resistance mechanisms. The genetic relationships among these isolates were assessed by multilocus sequence typing (MLST) and pulsed-field gel electrophoresis (PFGE) using SpeI, as described previously (9, 10).

Antimicrobial susceptibility testing.

The MICs of ampicillin, ticarcillin, piperacillin, cephalothin, cefoxitin, cefotaxime, ceftriaxone, cefuroxime, ceftazidime, cefepime, aztreonam, imipenem, meropenem, ertapenem, nalidixic acid, ciprofloxacin, levofloxacin, gentamicin, amikacin, tobramycin, and kanamycin were determined and interpreted by the broth microdilution method, according to the Clinical and Laboratory Standards Institute (CLSI) recommendations (11, 12), except for polymyxin B, for which European Committee for Antimicrobial Susceptibility Testing (EUCAST) breakpoints were applied (13). All β-lactams, except penicillins, were also tested in the presence of clavulanic acid at 4 μg/ml. Quality control (QC) was performed by testing Escherichia coli strain ATCC 25922 and Pseudomonas aeruginosa strain ATCC 27853, with all QC results falling within the expected ranges (12).

Carbapenemase activity.

β-Lactamase activity was determined spectrophotometrically using crude extracts of sonicated cells and imipenem as a substrate. The assay was carried out at 299 nm for up to 8 min. Inhibition assays were performed by incubating the cell extract with either 25 mM EDTA, 7.5 mM NaCl, or 10 μg/ml clavulanic acid for 15 min at room temperature. The results were compared with those obtained in the absence of inhibitor.

Isoelectric focusing.

Crude extracts of β-lactamases and the purified BKC protein were tested for pI determination using polyacrylamide gel containing Ampholine at pH 3.5 to 9.5 (Amersham Pharmacia Biotech, Piscataway, NJ, USA), as previously described (14).

Detection of β-lactamase-encoding genes and DNA sequencing.

Total DNA of K. pneumoniae isolates was extracted using the QIAamp DNA minikit (Qiagen, Courtaboeuf, France) and used as the template in PCR experiments. Amplification of bla_TEM, bla_SHV, bla_CTX, bla_GES, bla_KPC, bla_OXA-48, metallo-β-lactamases (bla_IMP, bla_VIM, bla_SPM, bla_GIM, bla_SIM, and bla_NDM), plasmid AmpC (bla_CMY, bla_DHA, bla_ACC, bla_ACT, bla_FOX, and bla_MIR), OXA-ESBL (bla_OXA-1, bla_OXA-2, bla_OXA-5, bla_OXA-7, bla_OXA-18, bla_OXA-45, and bla_OXA-46), and OXA carbapenemase-encoding genes (bla_OXA-23, bla_OXA-24, bla_OXA-51, and bla_OXA-58) was performed using specific primers and cycling parameters, according to the target sequence (15–21). Amplicons were purified with the QIAquick PCR purification kit (Qiagen, Hilden, Germany) and sequenced in both strands using the Applied Biosystems 3500 genetic analyzer equipment (Applied Biosystems, PerkinElmer, USA). The nucleotide sequence and the respective deduced amino acid sequences were analyzed using the Lasergene software package (DNAStar, Madison, WI, USA) and compared with the sequences available on the Internet using the BLAST tool (http://www.ncbi.nlm.nih.gov/blast/).

Plasmid profile and transference of carbapenem resistance.

Plasmid extraction of the K. pneumoniae isolates was performed using the methodology of Kieser (22) and the QIAprep spin miniprep kit (Qiagen), according to the manufacturer's instructions. Conjugation experiments were performed using the clinical isolate KP60136 as a donor and E. coli J53 as the receptor strain on plates supplemented with azide (150 μg/ml) and ticarcillin (200 μg/ml). The plasmids obtained from K. pneumoniae KP60136 (p60136) were transferred by electroporation into E. coli strain TOP10. The transformants were selected on Luria-Bertani (LB) agar plates supplemented with 0.5 μg/ml imipenem. The MICs were determined using the CLSI broth microdilution method (11, 12).

Plasmid pyrosequencing.

The sequencing of the plasmid obtained from the T3 transformant strain was carried out using the GS FLX 454 Roche platform (Roche Diagnostics GmbH, Penzberg, Germany). The Newbler software version 2.8 (Roche Diagnostics GmbH) was employed for plasmid sequence assembly.

Analysis of the outer membrane proteins.

Alterations in the ompK35 and ompK36 genes were investigated by PCR and sequencing, using the primers listed in Table 1. The expression of OmpK35 and OmpK36 was also analyzed by SDS-PAGE, as previously described (23).

TABLE 1.

Sequences of primers tested for outer membrane protein analysis and BKC-1 cloning and detection

Primer	Sequence (5′–3′)
Pre-OmpK35 F1	GGATGGAAAGATGCCTTCAG
Pre-OmpK35 F2	AATGAGGGTAATAAATAATGATGAAGC
Pre-OmpK35 R	CGAGGTTCCATTGTGATTACTG
OmpK35 F	TGATGAAGCGCAATATTCTGG
OmpK35 R	CCAGCCGCTTTGGTGTAAT
Pre-OmpK36 F	TTGTTGGATTATTCTGCATTTTG
Pre-OmpK36 R	TCTTACCAGGGCGACAAGAG
OmpK36-int F	CAGCACTGATGCCATCATAG
OmpK36-int R	ACGACGTAACGTCCTGGACC
BKC-1 F	ACATAATCTCGCAACGGGCG
BKC-1 R	TCGCCGGTCTTGTTCATCAC
BKC-F_Nde	TAATGCCATATGACGATCACATTTTCGCGCCGGCAG
BKC-R_Not	ATTATCGCGGCCGCTCAGGCCTCGGCGGCAATGCGACCA

Enzyme purification.

The bla_BKC-1 gene was amplified by PCR using the primers BKC-F_Nde and BKC-R_Not (Table 1) containing the NdeI and NotI restriction sites, respectively. The amplicon fragment was digested with both enzymes and cloned into pET-26b+ (Novagen, Merck KGaA, Darmstadt, Germany) within the corresponding insertion sites. The plasmid obtained was transformed into E. coli strain DH5α and then subsequently into E. coli strain BL21(DE3). Clone selection was performed on LB agar plates supplemented with 50 μg/ml kanamycin.

Production of the BKC-1 enzyme was induced in 1 liter of LB broth by adding 0.2 mM isopropyl-β-d-thiogalactopyranoside (IPTG), followed by overnight incubation at 20°C. The β-lactamases were released from the periplasm by osmotic shock, as follows: the cells were harvested by centrifugation at 4°C and washed with 10 mM CaCl₂. After a new centrifugation step, the pellet was resuspended in 20% sucrose, 30 mM Tris-HCl (pH 8.0), and 1 mM EDTA. This solution was harvested, and the pellet was carefully resuspended in cold deionized water. After complete resuspension, the cellular debris was removed by centrifugation, and the supernatant, containing the enzymes, was purified. The purification was performed by ion-exchange chromatography in an Äkta Purifier (GE Healthcare, Orsay, France). The solution containing the enzymes was loaded into a DEAE Sepharose column (GE Healthcare) previously equilibrated with binding buffer (Tris 50 mM [pH 8.0]), and eluted by a linear gradient of NaCl. The fractions obtained were tested against nitrocefin, and those with β-lactamase activity were pooled and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The fractions with >90% purity were concentrated using an Amicon Ultra-15 centrifugal filter device (10,000 nominal molecular weight limit [NMWL]; Millipore, Billerica, MA, USA). The concentration of BKC-1 was determined using the Bradford assay.

Kinetics.

Hydrolysis of the antimicrobial agents was monitored by following the change of absorbance resulting from the opening of the β-lactam ring using the purified BKC-1 in a Synergy H1 Hybrid multimode microplate reader (Biotek, Winooski, VT, USA) and measuring at 30°C in 100 mM sodium phosphate (pH 7.0). The Lineweaver-Burk's linearization of the Henri-Michaelis-Menten equation and a direct nonlinear regression with the hyperbolic equation from GraFit 5 (GraFit data analysis software; Erithacus Software) were used to fit the data and determine the V_max and K_m values. Low and very high K_m values were determined as K_i values, using nitrocefin as the reporter substrate. The extinction coefficients (Δε) of the antimicrobials were determined through the slope of the difference in absorbance of intact and degraded molecules of each antimicrobial tested. All reported values are the means of three independent measurements.

The K_i of the inhibitors (clavulanic acid and tazobactam) was determined in a competition assay measuring the initial steady-state velocity of a mixture of inhibitor and the reporter substrate (10 μM nitrocefin). The value of K_{i app} was corrected to determine the K_i according to the equation K_i = K_{i app}/(1 + [S]/K_m), where [S] is the substrate concentration and K_m is the affinity constant of the substrate. K_{i app} was calculated using the equation v_o/v_i = 1 + [I]/K_{i app}, where v_o is the velocity of hydrolysis without the inhibitor, v_i is the velocity of hydrolysis in the presence of the inhibitor, and [I] is the molar concentration of the inhibitor.

Nucleotide sequence accession number.

The nucleotide sequence of the p60136 plasmid was deposited in GenBank database with the accession number KP689347.

RESULTS

The three K. pneumoniae clinical isolates showed an identical PFGE pattern and also belonged to a novel sequence type (ST) (named ST1781), which belongs to clonal complex 442. The antimicrobial susceptibility profiles of the K. pneumoniae clinical isolates are shown in Table 2. Resistance to all β-lactams, including carbapenems (ertapenem MICs, 64 to >256 μg/ml; imipenem MICs, 4 to 128 μg/ml; and meropenem MICs, 8 to 32 μg/ml), quinolones (nalidixic acid MICs, >256 μg/ml; ciprofloxacin MICs, 32 to >32 μg/ml; and levofloxacin MICs, 8 to 32 μg/ml), amikacin (MICs, 64 to 128 μg/ml), and kanamycin (MICs, 256 to >256 μg/ml) were observed. A reduction of ≥2 log₂ dilutions was noted for ceftazidime (8-fold), cefepime (4-fold), imipenem (2-fold), and meropenem (2-fold) MICs in the presence of clavulanic acid.

TABLE 2.

Antimicrobial susceptibility profile of K. pneumoniae clinical isolates and E. coli T3 transformant and TOP10 receptor strains

Antimicrobial	MICs (μg/ml) for:
	KP60134	KP60135	KP60136	TOP10	T3	BL21(DE3)	BL21 + pET-BKC-1
Ampicillin	>256	>256	>256	4	>256	2	>256
Piperacillin	>256	>256	>256	2	>256	1	>256
Ticarcillin	>256	>256	>256	2	>256	2	>256
Cephalothin	>256	>256	>256	1	>256	2	>256
Cephalothin + clavulanic acid	>256	>256	>256	1	>256	2	4
Cefoxitin	128	128	128	2	4	1	1
Cefotaxime	>256	>256	>256	0.06	>256	≤0.015	16
Cefotaxime + clavulanic acid	>256	>256	>256	0.06	128	≤0.015	≤0.015
Ceftriaxone	>256	>256	>256	0.06	>256	≤0.015	8
Ceftriaxone + clavulanic acid	>256	>256	>256	0.03	256	≤0.015	≤0.015
Cefuroxime	>256	>256	>256	0.25	>256	1	256
Cefuroxime + clavulanic acid	>256	>256	>256	0.25	>256	1	1
Ceftazidime	>256	>256	>256	0.25	>256	0.06	8
Ceftazidime + clavulanic acid	64	32	64	0.25	256	0.06	0.06
Cefepime	>256	>256	>256	0.06	>256	0.06	4
Cefepime + clavulanic acid	128	128	128	0.03	128	≤0.015	0.12
Aztreonam	>32	>32	>32	0.25	>32	0.06	8
Aztreonam + clavulanic acid	>32	>32	>32	0.25	>32	0.06	0.06
Imipenem	8	4	128	0.125	8	0.06	0.5
Imipenem + clavulanic acid	4	2	32	0.125	2	0.06	0.06
Meropenem	16	8	32	≤0.015	8	≤0.015	0.12
Meropenem + clavulanic acid	4	4	32	≤0.015	4	≤0.015	0.03
Ertapenem	64	64	>256	≤0.015	64	0.03	0.12
Ertapenem + clavulanic acid	64	32	>256	≤0.015	64	0.03	0.06
Nalidixic acid	>256	>256	>256	2	2	NTa	NT
Ciprofloxacin	32	>32	32	≤0.03	≤0.03	NT	NT
Levofloxacin	8	32	16	≤0.03	≤0.03	NT	NT
Gentamicin	0.25	0.25	0.5	0.5	0.5	NT	NT
Amikacin	128	64	128	2	128	NT	NT
Tobramycin	4	2	4	1	1	NT	NT
Kanamycin	>256	256	>256	2	>256	NT	NT
Polymyxin B	0.25	0.25	0.25	≤0.12	≤0.12	NT	NT

^aNT, not tested.

PCR analysis of the porins showed that the three isolates had an intact ompK35 gene. However, on the SDS-PAGE gel, the band corresponding to the OmpK35 protein of KP60136 was absent, suggesting that some posttranslational event might have occurred, leading to the absence of the OmpK35 band on the SDS-PAGE gel. The sequencing results showed that the three isolates had a premature stop codon in the ompK36 gene, corroborating the SDS-PAGE results.

The β-lactamase-encoding genes bla_CTX-M-2 and bla_SHV-110-like were detected in all isolates; however, PCR experiments failed to identify any known carbapenemase-encoding genes. The hydrolysis of imipenem was observed for all isolates, and it was inhibited by the addition of clavulanic acid but not by EDTA or NaCl, suggesting the production of a class A carbapenemase.

The plasmid profile showed the presence of four plasmids (2.7, 8, 10, and 60 kb) in all clinical isolates. Conjugation experiments failed to obtain any transconjugants. The KP60136 plasmid DNA was transferred to E. coli TOP10 by electroporation, and a transformant, named T3, which carried a single 10-kb plasmid (named p60136), was obtained. Susceptibility testing showed that T3 exhibited high-level β-lactam resistance, including resistance to the three tested carbapenems, with MICs similar to those of the clinical strain KP60136. Strain T3 was also resistant to amikacin and kanamycin (Table 2) and did not carry bla_CTX-M-2 or bla_SHV-110-like β-lactamase-encoding genes.

To determine whether a carbapenemase-encoding gene was responsible for the carbapenem resistance detected in strain T3, p60136 was fully sequenced. It contains 9,786 bp with 10 open reading frames (ORFs), as shown in Fig. 1. Seven ORFs were responsible for encoding proteins involved in plasmid replication and mobilization (repA, repB, repC, repression protein F, mobA, mobB, and mobC genes) and showed that p60136 belonged to the incompatibility group Q (IncQ). Among the remaining three ORFs identified, one corresponded to an insertion sequence that belongs to the IS1380 family and possessed only 91% identity with ISApr9, characterizing a new IS, ISKpn23. Downstream of the ISKpn23 gene, there were two other ORFs, one containing a new β-lactamase-encoding gene, bla_BKC-1, and a new variant of an aminoglycoside phosphotransferase (3′)-VI gene, respectively. The aph3A-VI gene encodes a protein with four amino acid substitutions (Glu49Ala, Ala112Val, Thr156Ala, and Thr173Asn) compared to that encoded by aph3A-VIa.

FIG 1.

Map of the 10-kb plasmid obtained from the T3 transformant strain. The arrows represent the genes and their transcription direction: second darkest gray, mobilization module; darkest gray, replication module; lightest gray, antimicrobial resistance-encoding genes; and second lightest gray, insertion sequence.

bla_BKC-1 was found to encode an enzyme with 313 amino acids that contains the conserved motifs S₇₀TFK, S₁₃₀DN, E₁₆₆, and K₂₃₄TG, which are characteristic of class A β-lactamases. The phylogenetic tree from the amino acid sequence alignment of BKC-1 with other class A β-lactamases is represented in Fig. 2. The alignment of BKC-1 with β-lactamases commonly reported in clinical isolates showed very low identity to KPC-2 (39.5%), SHV-18 (38.7%), and TEM-3 (38.3%). In contrast, the highest identity (63.0%) was observed with an uncharacterized β-lactamase encoded by a gene carried on the megaplasmid pSymA of Sinorhizobium meliloti strain AK83 (GenBank accession no. WP_010968024). Isoelectric focusing (IEF) analysis of the clinical isolates revealed the presence of three bands with estimated pI values of 6.0, 7.2, and 8.3, corresponding to BKC-1, SHV-110-like, and CTX-M-2, respectively. The IEF profile of the T3 strain showed the presence of a single band with pI 6.0.

FIG 2.

Phylogenetic tree of BKC-1 containing the S. meliloti β-lactamase, the class A carbapenemase, and the main ESBL enzymes found in clinical isolates. Alignment was performed using the MegAlign program (Lasergene software package; DNAStar, Madison, WI, USA).

The genetic environment of the bla_BKC gene was investigated by EMBOSS (http://emboss.bioinformatics.nl). We searched for direct repeats (DRs) and inverted repeats (IRs) in all sequence regions that flanked the aph3A-VI, bla_BKC, and ISKpn23 genes. Since this tool performs searches using only one nucleotide sequence, the intergenic regions were merged for this search. We identified the inverted repeats of ISKpn23, corresponding to the four-nucleotide region 5′-AGCC-3′. We believe that this structure is a noncomposed transposon comprising the aphA3-VI, bla_BKC, and ISKpn23 genes. We detected no DRs in the analyzed sequence. Future studies may allow us to characterize a potentially new transposon.

The kinetic parameters of BKC-1 are shown in Table 3. The BKC-1 enzyme was able to hydrolyze penicillins, cephalosporins, monobactams, and carbapenems but not cephamycins. BKC-1 exhibited highest affinity for carbapenems, followed by penicillins (low K_m values). However, the carbapenems were not efficiently hydrolyzed due to their very low k_cat values. Among the cephalosporins, cephalothin, cefuroxime, and cefotaxime were more efficiently hydrolyzed, whereas ceftazidime had k_cat/K_m values 28-fold smaller than those of ceftriaxone. Oxacillin was the best substrate for BKC-1 (according to k_cat/K_m values) and also had the highest k_cat values among all the β-lactams tested. Imipenem was the more efficiently hydrolyzed than meropenem and ertapenem. Although meropenem and ertapenem have similar K_m values, the catalytic efficiency (according to k_cat/K_m values) of meropenem was higher due to its k_cat values, which were 2-fold higher than those of ertapenem. The inhibition measurements showed a K_i of 0.08 μM and 1.86 μM for clavulanic acid and tazobactam, respectively.

TABLE 3.

Steady-state kinetic parameters of substrate hydrolysis by purified BKC-1 β-lactamase

Substrate	kcat (s−1)	Km (μM)	kcat/Km (mM−1 s−1)
Benzylpenicillin	34.206	78.7	434.6
Ticarcillin	1.632	32.7	49.9
Nitrocefin	22.4	20.9	1,071.8
Oxacillin	14,306.6	267.3	53,522.6
Cephalothin	118.4	170.4	694.9
Cefoxitin	—a	—	—
Cefotaxime	0.4648	223.9	2.1
Ceftriaxone	4.325	127.8	33.8
Cefuroxime	55.315	366.1	151.1
Ceftazidime	0.1126	92.9	1.2
Cefepime	1.69	174.3	9.7
Aztreonam	2.224	1,200.7	1.9
Ertapenem	0.0016	1.7	0.9
Imipenem	0.031	4.4	7.0
Meropenem	0.0034	1.51	2.3

^a—, no measurable hydrolysis was detected.

DISCUSSION

Despite a diverse arsenal of antimicrobials, bacterial infections still represent a major cause of morbidity and mortality worldwide, especially due to the emergence and spread of multidrug-resistant isolates. Infections caused by carbapenem-resistant Enterobacteriaceae are very worrisome, because they have been associated with high mortality rates (7, 24, 25). Here, we report the identification of a novel class A carbapenemase, BKC-1, in K. pneumoniae clinical isolates belonging to a single clone that spread between two hospitals of the city of São Paulo. Patient-to-patient transmission might be one of the reasons for justifying the dissemination of BKC-producing K. pneumoniae, since interhospital spread of resistant Gram-negative clones has frequently been reported in Brazil (26–28).

Initially, BKC-1-producing K. pneumoniae isolates were identified as carbapenemase producers by a routine laboratory employing the modified Hodge test (MHT), and these isolates were referred to a research laboratory for further investigation. Since the search for carbapenemase-encoding genes initially had negative results, the previously obtained MHT results were misinterpreted as false positives. The production of BKC was confirmed only by the referral laboratory after the measurement of imipenem hydrolysis rates by testing crude extract cells of the isolates grown in agar plates supplemented with 0.5 μg/ml imipenem and using an extended time of reading (up to 8 min). Despite the weak imipenemase activity of BKC, this enzyme was responsible for conferring resistance to all tested β-lactams (penicillins, cephalosporins, monobactams, and carbapenems).

bla_BKC-1 was carried by p60136, a 10-kb IncQ plasmid. Plasmids belonging to the IncQ group are characterized by their small size and their ability to replicate and mobilize in a very broad range of hosts, being considered promiscuous plasmids (29). IncQ plasmids are not self-conjugative but can be mobilized at a high frequency by auxiliary conjugative plasmids, which might justify why p60136 was not transferred by conjugation (29). The association of the bla_BKC-1 with ISKpn23 is of great concern, because this insertion sequence may have facilitated the acquisition of bla_BKC-1 and further acted in the dissemination of bla_BKC-1 to other bacterial species. The fact that the G+C content of BKC-1 is high (68.79%) also corroborates this hypothesis that BKC-1 was acquired from a distant bacterial species. Besides, the presence of inverted repeats upstream the ISKpn23 gene and downstream the aph(3′)-VI gene suggests that a genetic element carrying the ISKpn23, bla_BKC-1, and aph3A-VIa genes was probably transposed.

The new variant of an aph(3′)-VI gene carried by p60136 encodes a new aminoglycoside-modifying enzyme with 98.5% identity with the APH(3′)-VIa enzyme. APH(3′) is the most widespread phosphotransferase group among clinical pathogens, but the variant APH(3′)-VI was described in only a few Enterobacteriaceae and Acinetobacter species isolates (30–32). The APH(3′)-VI variants usually confer resistance to most aminoglycosides, including kanamycin, amikacin, neomycin, paromomycin, ribostamycin, butirosin, and isepamicin (33).

Among all β-lactamases, BKC-1 possesses the highest identity with a β-lactamase found in the plasmid of S. meliloti, an environmental bacterium. S. meliloti is a Gram-negative alphaproteobacterium, belonging to the Rhizobiales order. This environmental species has been found in many types of soil worldwide, either in its free form or in symbiosis with certain leguminous plants, especially Medicago sativa (alfalfa), where it is involved in the nitrogen fixation process (34, 35). The soil is rich in microorganisms that produce β-lactams and, consequently, genes conferring resistance to these compounds, such as β-lactamases, are also abundant in this ecological niche (36–39). bla_BKC-1 might have originated from S. meliloti or another environmental species, and after several evolutionary processes, it is now detected in K. pneumoniae human clinical isolates.