Molecular Analysis of Antibiotic Resistance Determinants and Plasmids in Malaysian Isolates of Multidrug Resistant Klebsiella pneumoniae

Infections caused by multidrug resistant Klebsiella pneumoniae have been increasingly reported in many parts of the world. A total of 93 Malaysian multidrug resistant K. pneumoniae isolated from patients attending to University of Malaya Medical Center, Kuala Lumpur, Malaysia from 2010-2012 were investigated for antibiotic resistance determinants including extended-spectrum beta-lactamases (ESBLs), aminoglycoside and trimethoprim/sulfamethoxazole resistance genes and plasmid replicons. CTX-M-15 (91.3%) was the predominant ESBL gene detected in this study. aacC2 gene (67.7%) was the most common gene detected in aminoglycoside-resistant isolates. Trimethoprim/sulfamethoxazole resistance (90.3%) was attributed to the presence of sul1 (53.8%) and dfrA (59.1%) genes in the isolates. Multiple plasmid replicons (1-4) were detected in 95.7% of the isolates. FIIK was the dominant replicon detected together with 13 other types of plasmid replicons. Conjugative plasmids (1-3 plasmids of ~3-100 kb) were obtained from 27 of 43 K. pneumoniae isolates. An ESBL gene (either CTX-M-15, CTX-M-3 or SHV-12) was detected from each transconjugant. Co-detection with at least one of other antibiotic resistance determinants [sul1, dfrA, aacC2, aac(6ˊ)-Ib, aac(6ˊ)-Ib-cr and qnrB] was noted in most conjugative plasmids. The transconjugants were resistant to multiple antibiotics including β-lactams, gentamicin and cotrimoxazole, but not ciprofloxacin. This is the first study describing the characterization of plasmids circulating in Malaysian multidrug resistant K. pneumoniae isolates. The results of this study suggest the diffusion of highly diverse plasmids with multiple antibiotic resistance determinants among the Malaysian isolates. Effective infection control measures and antibiotic stewardship programs should be adopted to limit the spread of the multidrug resistant bacteria in healthcare settings.


Conjugation
Conjugation was carried out by broth mating (at a ratio of 1:4) in order to test the transmissibility of ESBL plasmid from K. pneumoniae isolates as donors to a recipient Escherichia coli strain J53 AzR (resistant to azide) [18]. A total of 43 K. pneumoniae isolates were selected based on types of ESBL genes and plasmid replicons. E. coli clones carrying ESBL plasmid (transferred from the donor) were selected on Luria-Bertani agar plates containing azide (100 μg/ml) and cefotaxime (2 μg/ml).
Characterization of the transconjugants. For detection of beta-lactamase production by transconjugants, nitrocefin kit (Oxoid, UK) was used as recommended by the manufacturer. Phenotypic detection of ESBL production by the transconjugants was performed using Oxoid combination disk method (cefpodoxime and cefpodoxime with clavulanate). Antimicrobial susceptibilities of the transconjugants were determined by disk diffusion and E-test methods. Susceptibility profiles and MICs of the transconjugants were compared to those of the donors and E. coli (J53 AzR) recipient strain. All the transconjugants were screened using PCR for the plasmid replicons and antibiotic resistance genes in their donors. Representative amplicons were sequenced for the purpose of confirmation.
Plasmid DNA was extracted from the donors and transconjugants using a plasmid midi kit (Qiagen, Germany). To determine the plasmid size and number, extracted plasmids from donors and transconjugants were separated on a 0.8% agarose gel prestained with 0.5 μg/ml ethidium bromide (Thermo Scientific, US). Supercoiled DNA ladder (New England Biolabs, UK) was used as a molecular weight standard for the plasmid size estimation. Electrophoresis was performed in 0.5X TBE buffer at 80 V for 3 hr.
Transconjugants plasmids were digested with the restriction enzyme EcoRI-HF (New England Biolabs, UK) [19]. Restriction fragments were separated on a 0.8% agarose gel prestained with 0.5 μg/ml ethidium bromide (Thermo Scientific, US) at 80 V for 3 hrs and visualized following gel electrophoresis. Lambda DNA/HindIII Marker, Ready-to-Use (Thermo Fisher Scientific, Lithuania) was used as a molecular weight standard. Plasmid restriction profiles were compared using Bionumerics software, version 7.0 (Applied Maths, Kortrijk, Belgium). Cluster analysis was carried out by the unweighted pair group method with arithmetic mean (UPGMA) algorithm by defining a similarity (Dice) coefficient. Cluster designation was based on plasmid profiles with 80% relatedness.

Statistical analyses
Antibiotic resistance rates were expressed as percentages of the total number of isolates. All the statistical tests were performed using PASW Statistics version 18 (SPSS Inc., Chicago, IL, US). A p-value <0.05 was considered statistically significant. Paired-sample t-test was used to compare the MICs of transconjugants to their donors (K. pneumoniae) and recipients (E. coli strain J53 AzR), respectively.

Transmissibility of antibiotic resistance plasmids
Despite repeated attempts, conjugation experiment was successful for 27(62.8%) of 43 K. pneumoniae isolates which were selected as donors for conjugation based on their plasmid replicons (FIIK replicons alone or accompanied by different types of plasmid replicons) and ESBL genes (CTX-M-15, -3, -63 and SHV-12). Table 4 shows the characteristics of 43 donor K. pneumoniae isolates selected for conjugation experiments and the results of conjugation.
A statistically significant association (p<0.05) between the presence of CTX-M-15 and FIIK was observed in this study as FIIK replicon was the dominant replicon detected in 22 of 24 CTX-M-15 harbouring transconjugants. On the other hand, SHV-12 gene was detected from two transconjugant plasmids carrying N and R replicon types, respectively.
Analysis of plasmid extracts from 43 K. pneumoniae donor isolates revealed the detection of 1-6 plasmids with size ranging from~1.5-100 kb. A single plasmid was detected from most of the transconjugants (~100 kb in 18/27 transconjugants,~60 kb and 90 kb in one transconjugant each). In seven transconjugants, the~100 kb plasmid was accompanied by an additional

Characterization of the transconjugants
All the 27 transconjugants demonstrated β-lactamase activities using nitrocefin chromogenic detection method and were confirmed as ESBL producers by cefpodoxime combination disk method. Table 5 shows the comparison between K. pneumoniae donor isolates and their  27) transconjugants (n = 27) with respect to non-susceptibility rates (%) and MICs of various antibiotics. The transconjugants were resistant to multiple antibiotics including β-lactams, gentamicin and trimethoprim/sulfamethoxazole, but not ciprofloxacin. All the transconjugants demonstrated high resistance rates (96.3-100%) to β-lactam antibiotics except carbapenems and cefoxitin (0% for each). The transconjugants exhibited high resistance rates to β-lactam/β-lactamase inhibitor combinations (96.3% and 100% for amoxicillin-clavulanate and ampicillinsulbactam, respectively); but were all susceptible to piperacillin-tazobactam (MIC 2-4 μg/ml) in contrast to some donors (29.6%) which were non-susceptible to this drug (MIC 48-128 μg/ml). None of the transconjugants demonstrated resistance to ciprofloxacin when compared to 70.4% of the donor K. pneumoniae isolates which were non-susceptible to ciprofloxacin (MIC 2-32 μg/ml).
The full list of 27 K. pneumoniae donor isolates and their transconjugants (E. coli strain J53 AzR) with their MICs and antibiotic resistance genes are shown in S3 Table. The

Discussion
In this study, CTX-M-15 was the prevalent ESBL gene detected among 91.3% of Malaysian multidrug-resistant K. pneumoniae isolates. The finding was in agreement with most recent studies in Asia and worldwide [22][23][24]. The dramatic shift of ESBL gene types from SHV to CTX-M has been noted globally as ESBL-SHV types are currently less common compared to CTX-M types [25,26]. SHV gene was detected in 78.5% of the isolates; however, SHV-12 (6.5%) was the only ESBL-SHV gene detected in this study. SHV-5 has been previously identified as the most common ESBL gene in Malaysian K. pneumoniae isolated during a nosocomial outbreak in the pediatric oncology unit of University of Malaya Medical Center, Kuala Lumpur [27]. A subsequent study in this setting reported the emergence of SHV-12, a derivative from SHV-5 by acquisition of a single mutation (Leu35Gln) [28]. According to the latest SMART study, SHV-12 was the dominant ESBL-SHV gene in K. pneumoniae isolates from many parts of Asia-Pacific region including Malaysia, China, India, Philippines, Taiwan, Korea and Singapore [23]. This is the first report of SHV-27, -28, -75 and -83 types in Malaysian K. pneumoniae isolates. Some of these SHV types have been reported in other Asian countries, including SHV-28 which has been reported in K. pneumoniae isolates from China (GenBank accession no. AF538324), Korea [29] and India [30]. SHV-75 and SHV-27 have been documented in K. pneumoniae isolates from Thailand [24]. SHV-83 was detected for the first time in a K. pneumoniae isolate from Portugal [31], but has never been reported in South East Asia. The diversity in the SHV type of Malaysian K. pneumoniae isolates may occur through international spread by travelers or emerged via mutations in the parental SHV-1 or SHV-11 genes which are prevalent in the local isolates [32].
CTX-M-3 has been described as a common ESBL gene in some Asian countries including Korea, Taiwan and China [33,34]. To the best of our knowledge, this is the first report of CTX-M-3 and CTX-M-63 (1.1% each) in Malaysian K. pneumoniae isolates. CTX-M-63 is an enzyme belongs to CTX-M group-8 which was first identified in a Japanese K. pneumoniae isolate in 2005 (GenBank accession no. AB205197), and later reported among isolates from other bacterial species including Morganella morganii (GenBank accession no. EU660216) and Salmonella enterica in Thailand [35].
OKP β-lactamases are chromosomally-encoded enzymes specific for K. pneumoniae, sharing the same ancestor origin with both SHV and LEN genes [36]. In this study, the detection of OKP gene in two isolates was by accident as the gene was amplified using SHV sequencing primers [21]. OKP β-lactamases are phylogenetically related to SHV; thus, the detection of OKP gene using SHV primers is possible as reported in a previous study [37]. In general, OKP gene is rarely reported in ESBL-producing K. pneumoniae isolates [38] and has not been reported before in South East Asia.
Except for a report of CMY detection in Malaysian K. pneumoniae isolates [39], limited information is available on AmpC genes in Enterobacteriaceae isolates in Malaysia. DHA-1 was identified recently in a Malaysian K. pneumoniae isolate in the SMART Asia-Pacific study [23]. The detection of DHA-1 gene in two isolates investigated in this study suggests the emergence of the AmpC gene among Malaysian K. pneumoniae isolates. Additionally, regional variations in the distribution of AmpC genes has been reported in Enterobacteriaceae isolates from the Asia-Pacific region according to the latest SMART study [23]. DHA-1 gene was more common in Enterobacteriaceae isolates from Philippines and Singapore compared to CMY gene which was dominant in Enterobacteriaceae isolates from Taiwan, India, South Korea and Vietnam [23].
A big proportion (74%) of the multidrug resistant K. pneumoniae isolates investigated in this study were non-susceptible to gentamicin. Resistance to gentamicin is common in ESBLproducing K. pneumoniae isolates, as reported in several parts of the world including China [3] and Taiwan [40]. In this study, only a few isolates (5.4%) were non-susceptible to amikacin. This finding concurs with recent surveillance studies of K. pneumoniae isolates from the Asia-Pacific region which indicted that less than 10% of the isolates were non-susceptible to amikacin [1,41].
The association between non-susceptibility to gentamicin and the presence of aacC2 gene encoding the gentamicin modifying enzyme AAC(3)-II have been documented [42,43]. In agreement with these reports, aacC2 was detected in most gentamicin non-susceptible isolates in this study. Additionally, one of the gentamicin-resistant isolates harboured aadB gene which is known for conferring resistance to gentamicin [44]. aadB gene has been reported recently in K. pneumoniae isolates from Malaysia [22,45], in contrast to aacC2 gene which is reported for the first time in Malaysia. Both aacC2 and aadB genes have been reported in gentamicin-resistant K. pneumoniae isolates from China, with higher prevalence of the former (60%) than the latter (3.6%) [3]. Similar observation was noted in this study as aacC2 gene (67.7%) was more common than aadB gene (1.1%). aacC1, which confers resistance to gentamicin, has been detected at a very low rate (2.7%) among aminoglycoside-resistant K. pneumoniae isolates in a Chinese study [3]; however, it was not detected in this study.
High-level resistance to multiple aminoglycosides is common in K. pneumoniae due to the spread of 16S rRNA methylases in multidrug resistant bacteria worldwide [3]. Except for strain NDM-2012 carrying armA gene [14], none of the isolates understudied were positive for 16S rRNA methylases (armA and rmtB) or phosphotransferase gene (aphA6), two genes associated with amikacin resistance [43].
Trimethoprim/sulfamethoxazole resistance is commonly observed in ESBL-producing K. pneumoniae isolates [3]. High resistance rate against trimethoprim/sulfamethoxazole (90.3%) as noted in this study, was probably attributed to the presence of sul1 and/or dfrA genes. In fact, both sul1 (encoding dihydropteroate synthases) and dfr (encoding dihydrofolate reductase) have been reported as causes of trimethoprim/sulfamethoxazole resistance in Gram-negative bacteria [46].
There is limited data on the plasmids of multidrug resistant K. pneumoniae isolates in Malaysia. The detection of 1-6 plasmids (~1.5-100 kb) in our isolates reflects the complexity of multidrug resistant isolates, in agreement with a previous study which described significantly higher number of plasmids in multidrug resistant K. pneumoniae isolates [47]. The existence of several plasmids within the same isolate may increase the possibility of genetic reassortment and recombination events and contribute to the plasmid diversity by recruiting new resistance genes into the plasmid scaffold [48].
In this study, 1-4 of 14 plasmid replicons types were detected from majority (95.7%) of the isolates. The finding suggests the existence of highly diverse plasmids carried by the Malaysian multidrug resistant K. pneumoniae isolates. The FIIK replicon was the dominant replicon type (90.3%) which was identified as a single replicon in more than half (53.8%) of the 93 isolates investigated in this study. K. pneumoniae isolates carrying FIIK replicon has been reported to have great ability to diffuse and persist in time [49,50], mainly because the bacteria are equipped with both virulence and antibiotic resistance determinants on the FIIK plasmids [10].
The detection of a few isolates (4.3%) carrying IncF replicons (FIIK, FIA, FIB and/or FII) is in line with previous studies which reported the detection of multi-IncF-replicon plasmids in K. pneumoniae and other species from Enterobacteriaceae [8,47]. IncF plasmids are notorious for their ability to evolve rapidly in order to adapt to the host environment; thus, mutations and recombination are common in these plasmids [10,51]. Sequence analysis revealed diversity in the IncF replicons identified in this study. FIIK replicons were differentiated into seven alleles based on sequence variation, of which, two were considered novel (FIIK-9 and -10). In addition, various alleles from FIA, FIB and FII replicon types were identified in this study.
There is a paucity of data on FIIK alleles around the world due to the limited publications available in this field [10,11]. Most studies investigating plasmid replicon types in K. pneumoniae used the old version of PCR-based plasmid replicon typing (PBRT) method developed in 2005 [17], which utilized primers specific for the detection of 18 replicons including IncF replicons (FIA, FIB, FIC, FII and F), but not for FIIK replicon [17]. The new PBRT scheme updated in 2010 has included new primers for the detection of IncF plasmids specific for Salmonella, Klebsiella and Yersinia spp. (FIIS, FIIK and FIIY, respectively) [10]. Hence, the real prevalence of FIIK might not been updated in many parts of the world [47,52]. The only database available for comparative analysis of FIIK alleles is provided by the Plasmid MLST (pMLST) website (http://pubmlst.org/plasmid/).
The second most common replicon identified in this study was the R replicon which was detected in 20.4% of the isolates harbouring FIIK replicon. Both FIIK and R replicons have been described previously to have an association with CTX-M-15 in Spanish K. pneumoniae isolates [49]. The R replicon has also been identified on a resistance plasmid associated with the spread of KPC gene amongst Canadian K. pneumoniae isolates [53] and on resistance plasmids bearing CTX-M-15, qnr and/or aac(6 0 )-Ib-cr in Spanish K. pneumoniae isolates [54]. These reports suggest the contribution of R replicon in the dissemination of resistance genes.
The remaining replicons detected in this study (FIB-M, HIB-M, N, HI2, Y, I1, A/C, X1 and K) were less common (1.1-7.6%) among the isolates. This was expected as some plasmid replicon families have a narrower distribution in certain geographical regions [7,9]. Strain NDM-2012 (carrying both NDM-1 and OXA-232 genes), the only carbapenem-resistant isolate investigated in this study, harboured four plasmid replicons i.e. FIIK, R, FIB-M and HIB-M, of which, the latter two are novel replicons reported from the plasmid of a Moroccan K. pneumoniae isolate carrying NDM-1, CTX-M-15 and qnrB1 genes [55]. FIB-M and HIB-M were not limited to strain NDM-2012 as they were also detected in several isolates investigated in this study. Plasmids can be exchanged among different isolates. Additionally, the resistance genes can be acquired or lost from a plasmid scaffold; thus, a particular resistance gene may not be linked exclusively to plasmids from a specific replicon family [8]. This may explain our observation of isolates carrying the same replicon(s) but exhibiting variability with respect to the antibiotic resistance genes.
Conjugation is the most common mechanism of horizontal dissemination of resistance plasmids with much higher success rates in nature than under laboratory conditions [6,56]. In this study, conjugative transfer of ESBL plasmids to E. coli strain J53 AzR recipient strain was only successful in 27 (62.8%) of 43 K. pneumoniae donor isolates. Approximately same rate (61%) of conjugation has been documented for multidrug resistant K. pneumoniae isolates investigated in a Czech study [11]. The success rate of conjugation could be affected by the selection of the recipient strain. Higher conjugative transfer rates of ESBL plasmids have been observed from K. pneumoniae to E. coli (47%) than to Salmonella (20%) recipient strains [11]. In this study, conjugation was only attempted using E. coli strain J53 AzR (kindly provided by Dr. Jacoby G.A.), a recipient strain widely used for conjugation experiments [3].
The replicative and transferability properties of plasmids are related to their incompatibility groups and a relationship may exist between FIIK allele and conjugative efficiency of the plasmids [6]. IncF plasmids are characterized by extensive mutations, insertions, deletions and recombination events which may affect the tra genes encoding transferase proteins for mating aggregation and DNA movement into the recipient cell and, consequently, plasmids conjugative efficiency [57,58]. All plasmids carrying FIIK-4 (1.1%), FIIK-5 (5.4%) and FIIK-8 (5.4%) were not successfully transferred to the E. coli recipient strain. The loss or mutations in tra genes in the plasmids carrying certain FIIK alleles such as FIIK-4, -5 and-8 in this study might have resulted in less efficient donors for conjugation [58]. Hence, FIIK-4, FIIK-5 and FIIK-8 replicons were found less common (1.1%, 5.4% and 5.4%, respectively) as compared to FIIK-2 and FIIK-7 replicons (48.4% and 24.7%, respectively) in the isolates understudied.
It has been reported that conjugative plasmids are bigger in size compared to the non-conjugative counterparts [59]. Several authors have reported the recovery of conjugative resistance plasmids of 30 kb from multidrug resistant K. pneumoniae isolates [11,48,54]. However, resistance genes can be carried by small plasmids (<10 kb), also known as mobilizable resistance plasmids, which can be disseminated to a new host with the help of the conjugative plasmids [59,60]. In agreement with these reports, this study showed the transfer of a big plasmid (~60-100 kb) by conjugation from 27 donor K. pneumoniae isolates to the E. coli recipient strain. The big plasmid was accompanied by 1-2 small (~3-5 kb) plasmids in seven transconjugants. The co-transfer of multiple plasmids from donors K. pneumoniae isolates to E. coli recipient strain has been reported previously [19]. The small plasmids may be mobilizable resistance plasmids or helper plasmids which provide mobilization proteins essential for conjugation [7].
In this study, SHV-12 ESBL gene was detected from two transconjugant plasmids carrying N and R replicon types, respectively. Both replicons have been previously documented in plasmids carrying SHV-12 in K. pneumoniae isolates [61,62]; however, SHV-12 was not confined to plasmids with a particular replicon type due to the ability of this gene to move among different plasmid scaffolds [8,63]. The association between the presence of CTX-M-15 and FIIK plasmid replicon on transconjugant plasmids was noted in this study, in agreement with previous reports [10,11,48]. Three transconjugant plasmids carrying CTX-M-15 (n = 2) and CTX-M-3 (n = 1) were devoid from any replicons. Non-typable plasmids (lacking replicons) have been previously reported in antibiotic resistant K. pneumoniae [47,64] and other members of Enterobacteriaceae [8].
Full sequencing of resistance plasmids in K. pneumoniae [65,66] and other bacterial species including E. coli [58,67] showed the structural linkage of various antibiotic resistance genes which were clustered in a specific region within the plasmid known as multi-resistance region.
In this study, ESBL genes (CTX-M-15 or SHV-12) amplified from the transconjugant plasmids were accompanied by various collections of other resistance genes including those encoding broad-spectrum β-lactamases (SHV-11, OXA-1 and TEM), PMQR genes (aac(6 0 )-Ib-cr and qnrB), aminoglycoside resistance gene (aacC2 and aac(6 0 )-Ib) and trimethoprim/sulfamethoxazole resistance genes (sul1 and dfrA). The detection of multiple antibiotic resistance genes from the same plasmid suggests the possibility of having multi-resistance regions carried in the plasmids of the Malaysian K. pneumoniae isolates. Full plasmid sequencing approach is required to prove this hypothesis.
Some of the genes detected in the donors (aacC2, aac(6 0 )-Ib-cr, DHA-1, sul1 and dfrA) were not detected from the transconjugants. This may be attributed to their locations on a different plasmid other than the ESBL plasmid transferred during the conjugation experiment [19]. Furthermore, chromosomal location of these genes may be suspected as recent reports have confirmed the location of some plasmid-mediated antibiotic resistance genes such as aac(6 0 )-Ib-cr and armA on K. pneumoniae chromosome [54,68].
In general, MICs of β-lactams, β-lactam/β-lactamase inhibitors, aminoglycosides, and fluoroquinolones of the transconjugants were significantly lower than their donor K. pneumoniae isolates. This finding suggests the presence of other resistance mechanisms in the donors, for instance, chromosomal-mediated resistance genes, reduced intracellular drug accumulation due to active efflux pump and/or porin loss [69,70].
All the transconjugants carrying ESBL genes [SHV-12 (n = 2), CTX-M-3 (n = 1) and CTX-M-15 (n = 24)] demonstrated high resistance rates (96.3-100%) to β-lactam antibiotics except carbapenems and cefoxitin (0% for each). The transconjugants exhibited high resistance rates to β-lactam/β-lactamase inhibitor combinations (96.3% and 100% for amoxicillin-clavulanate and ampicillin-sulbactam, respectively); but were all susceptible to piperacillin-tazobactam. In fact, the ability of β-lactam/β-lactamase inhibitor combination to inactivate βlactamases is dependent on the total quantity of the enzyme that needs to be inhibited; thus, βlactamases hyperproduction or the concomitant presence of multiple β-lactamases may reduce the bacterial susceptibility to these combinations [71][72][73]. It is possible that lower β-lactamases net amount was produced by the transconjugants due to the absence of the chromosomal non-ESBL SHV enzymes (SHV-1, -28, -27, -75, -83, -144, -11) from these transconjugants; thus affecting their susceptibility levels to piperacillin-tazobactam [73,74]. None of the transconjugants demonstrated resistance to ciprofloxacin when compared to some of the donor K. pneumoniae isolates which were resistant to ciprofloxacin due to the presence of chromosomal gyrA and/or parC mutations [12]. All gentamicin-resistant transconjugants were positive for aacC2 gene, confirming the role of this gene in conferring resistance to gentamicin [42]. Trimethoprim/sulfamethoxazole resistance was detected in transconjugants carrying sul1 and/or dfrA genes, confirming the role of these genes in conferring resistance to trimethoprim/sulfamethoxazole [46].
Restriction analysis of plasmids extracted from 27 transconjugants in this study revealed high diversity of these plasmids (15 profiles) even amongst plasmids of the same replicon type. Genetic events such as insertions, deletions, reassortment and recombination may have happened during plasmid evolution contributing to the observed plasmid diversity [58,64]. The finding of diverse plasmids carrying multiple resistance genes in the Malaysian K. pneumoniae isolates was on the contrary to the findings of a recent Chinese study whereby, a single epidemic plasmid was implicated in the spread of CTX-M-15 gene in the K. pneumoniae isolates [19]. In that study, conjugative transfer of a 90 kb IncFII plasmid was documented from K. pneumoniae isolates collected from different hospitals in southern China. These plasmids were highly related as indicated by restriction analysis of transconjugant plasmids carrying CTX-M-15 gene alone in contrast to the parental K. pneumoniae isolates which harboured multiple resistance genes such as DHA-1, qnrB, qnrS, aacC2, and aac(6 0 )-Ib [19]. These findings indicate that the plasmids circulating among K. pneumoniae isolates from different geographical locations may vary in the type and distribution of resistance genes [6].
In conclusion, this is the first study describing the characterization of plasmids in Malaysian multidrug resistant K. pneumoniae. The results of this study suggest that highly diverse plasmids with multiple antibiotic resistance determinants are spread among the Malaysian isolates. The location of resistance genes on conjugative plasmids and the ability for co-transference en bloc is an alarming finding as their dissemination may increase multidrug resistance rates among the Malaysian K. pneumoniae isolates unless more strict infection control measures and antibiotic stewardship programs are adopted to limit the spread of the multidrug resistant bacteria.
Supporting Information S1