The emergence of plasmid mediated quinolone resistance qnrA2 in extended spectrum β-lactamase producing Klebsiella pneumoniae in the Middle East

Background Klebsiella pneumoniae is one of the most important opportunistic pathogens causing serious complications in patients in hospitals and community. The clinical significance of K. pneumoniae is mainly due to its ability to acquire multiple antibiotic resistance genes. In this study we report the findings of a survey of plasmid mediated quinolone resistance in Extended-Spectrum β-lactamase (ESBL)-producing K. pneumoniae in Kuwait. Methods Clinical samples were collected from the microbiology laboratories of three major hospitals. Isolates were confirmed as ESBL-producers by disc diffusion method and PCR for the presence of bla genes. Antimicrobial susceptibility testing and genetic analysis were performed to detect the presence of a number of genes conferring resistance to β-lactam and fluoroquinolone antimicrobial agents including blaSHV, blaTEM, aac (6')-Ib-cr, qnrA, qnrB and qnrS. Pulsed-field gel electrophoresis (PFGE) was used for typing the isolates. Results In total 173 ESBL-producing K. pneumoniae were detected. qnr genes were identified in 27 (15.6 %) isolates and aac(6′)-Ib Ib-cr gene in 26 (96 %). One (3.7 %) contained qnrA2, 21 harbored qnrB1 (78 %) and 5 (18.5 %) contained qnrS. Twenty one (78 %) isolates contained all three bla genes. PFGE showed diverse profiles. Conclusion We identified for the first time the emergence of the mobile fluoroquinolone resistance qnrA2 in a clinical isolate in the middle east and also showed the dissemination of aac (6')-Ib-cr, qnrB, and qnrS genes among ESBL-producing K. pneumoniae in Kuwait. The abundance of plasmid mediated resistance to fluoroquinolones among ESBL-producing K. pneumoniae is alarming as it facilitates therapy failure. Preventing the spread of these isolates is crucial if we are to sustain the effectiveness of the limited choices we have left in antimicrobial therapy.


Background
Members of the Enterobacteriaceae family in particular the multi-drug resistant Klebsiella pneumoniae are amongst the opportunistic pathogens. K. pneumoniae causes a wide range of infections from urinary and respiratory tract infections to bacteraemia, particularly in elderly or debilitated patients. Antibiotic resistance is an important factor that hinders therapy and delays improvement in patients' health. In K. pneumoniae resistance to antimicrobial agents is caused by different mechanisms however the production of extended spectrum β-lactamases (ESBLs) such as CTX-M and other enzymes including TEM-1, TEM-24, SHV-12 and the plasmid-mediated AmpC (CMY-2) are considered highly important [1][2][3]. CTX-M-15 producing strains can often contain variants of an aminoglycoside-modifying enzyme expressed by aac(6′)-Ib-cr genes. Aminoglycosidemodifying enzymes cause reduced susceptibility to aminoglycosides and to some fluoroquinolones [4]. Fluoroquinolone resistance in K. pneumoniae can arise by mutations in the chromosomal type II topoisomerases and changes in the expression of efflux pumps and porins. Furthermore fluoroquinolone resistance can also be facilitated by plasmid-mediated resistance determinants qnr that express the pentapeptide repeat proteins. These proteins protect the quinolone targets (DNA gyrase or topoisomaerase IV enzymes) from the inhibitory activity of fluoroquinolone antibiotics [5,6]. In view of the genetic heterogeneity variants of qnrA, B and S-like genes share 94 % to 99 %, 85 % to 99 % and 90 % nucleotide identity respectively [7].
Following the rise in the use of cephalosporins in hospitals; the incidence of infections caused by ESBL-producing K. pneumoniae has soared [2,10]. Resistance to 3 rd generation cephalosporins is predominantly associated with the presence of bla CTX-M genes located on plasmids [11]. Therefore plasmid mediated resistance to fluoroquinolones among ESBL producing K. pneumoniae is alarming as it facilitates therapy failure. Moreover; the exchange of the genetic information among bacterial species threatens the efficacy of fluoroquinolones and increases the dependency on carbapenem antibiotics for treatment. The purpose of the present study was to determine the prevalence of plasmid-mediated fluoroquinolone resistance in ESBL-producing K. pneumoniae in hospitalized patients in Kuwait.

Bacterial isolates
In-patient clinical samples were collected from the microbiology laboratories of three major governmental hospitals that serve the six governorates of Kuwait, namely Al-Ahmadi hospital, Al-Amiri hospital and Adan hospital. All three hospitals are tertiary health care providers with bed capacities of 300, 500 and 600, respectively. The average number of specimens processed every day in each hospital varies from 500 to 700 which includes samples from out-patient and in-patient specialists units. A table was created based on the patients' records containing information such as age, gender, hospital, and type of specimen. Specimens were processed by clinical staff members of the diagnostic microbiology laboratories using standard protocols. The species identification was carried out by the API (bioMérieux, Marcy l'Etoile, France) and the VITEK 2 systems (Vitek AMS; bioMérieux Vitek Systems Inc., Hazelwood, MO, USA). The results were analyzed according to the Clinical and Laboratory Standards Institute (CLSI) (2012) guidelines [12]. The isolates were stored in 10 % skim milk at −70°C.

Susceptibility testing
Antibiotic susceptibility tests were performed by the Kirby-Bauer method disc diffusion test following the CLSI (2012) criteria and recommendations [12]. The antibiotics tested were ampicillin, ampicillin/sulbactam, amoxicillin, amoxicillin/clavulanic acid, piperacillin/tazobactam, ceftazidime, cefepime, ceftriaxone, cefazolin and cefuroxime. The Minimum Inhibitory Concentration (MIC) was determined by agar dilution method and E-test when available for the following antibiotics: trimethoprim gentamicin, cefotaxime, imipenem, meropenem, ciprofloxacin, levofloxacin and tetracycline. Isolates that showed resistance to at least three classes of antibiotics were considered as MDR. Isolates that were detected as resistant to cefoxitin were further investigated for the presence of an ampC β-lactamase by using multiplex PCR [13,14].

Double-disc synergy method
ESBLs were detected as previously described using the disc approximation and double-disc synergy methods and confirmed with cefotaxime and ceftazidime E-test ESBL strips (AB Biodisk, Biomerieux-diagnostics, Durham, NC, USA) [2]. For the disc approximation test, clavulanate diffusion from an amoxicillin-clavulanate (AMC30) disc was used to test for synergy with cefotaxime, ceftazidime, cefuroxime, cefepime and cefixime (Oxoid) as described previously [15]. For the double-disc synergy test, a ceftazidime disc (30 μg) was placed 30 mm away from a disc containing amoxicillin-clavulanate (60/10 μg). ESBL production was considered positive when an enhanced zone of inhibition was visible between the β-lactam and β-lactamase inhibitor-containing discs. For the E-test, ESBL strips containing ceftazidime and ceftazidime-clavulanate and strips containing cefotaxime and cefotaxime-clavulanate were used to determine the MIC ratio according to the manufacturer's instructions (AB Biodisk, Biomerieux-diagnostics, Durham, NC, USA). Cultures were incubated aerobically at 37°C for 18-24 h. CTX-M-15 β-lactamase enzyme displays a catalytic activity toward ceftazidime.
Isolates resistant to ciprofloxacin and for which the ceftaxidime or cefotaxime MICs were >8 mg/L were screened for ESBLs and qnr genes. Double disc and combination disc tests were used for ESBLs confirmation.
Amplified PCR products were purified with Qiagen purification kit (Qiagen, Limburg Netherlands) according to the manufacturer's instructions and both strands were sequenced by automated ABI3100 DNA sequencer (Applied Biosystems, Foster City, CA, USA). The BLAST program of the National Centre for Biotechnology Information (http://www.ncbi.nlm.nih.gov) was used to search and compare databases for similar nucleotide sequences.

Pulsed-Field Gel Electrophoresis (PFGE)
PFGE was performed by using the restriction enzyme XbaI

Statistical analysis
Out of the total number of isolates included hereby, the percentage of the different ESBL-producing bacteria were calculated in each of the three hospitals participating in the study.

Bacterial isolates
All three governmental hospitals participated during our study period from 2010 to 2012; however there were inconsistencies in the level of strain contribution from each hospital. Therefore under-representation of ESBL-mediated K. pneumoniae might exist. A total of 832 ESBL-producing Enterobacteriaceae isolates were obtained Table 1 The antibiotic sensitivity testing for K. pneumoniae isolates was performed followed by double-disc synergy tests. The results were confirmed with PCR and sequencing the product.

PFGE
PFGE showed diverse profiles among the isolates tested (Fig. 1). Only two isolates (KOC 7 and KOC 64) from urinary tract infection had identical PFGE patterns. The majority of isolates with similar qnr genes; displayed similar antimicrobial resistance phenotypes but their PFGE profiles were not identical. In total three isolates (ADA-45, ADA-49 and KOC-37) were not typable by PFGE. Table 3 The profiles of qnr positive K. pneumoniae isolates and other resistance genes studied

Discussion
In the present study we detected a similar rate of ESBLproducing K. pneumoniae (21 %) as previously reported (28 %) [19]. However, we found 15.6 % (27/173) of ESBL-producing K. pneumoniae contained a plasmid mediated qnr gene. Previous research undertaken in Kuwait in 2007 did not identify either qnrA or qnrS in ESBLproducing Enterobacteriaceae and reported a low prevalence of 4.7 % (3 out of 64 isolates) of qnrB in E. coli [7].
The key findings in our study are the emergence of qnrA2 in the Middle East for the first time and also the spread of qnrS among the clinical ESBL-producing K. pneumonia isolates in Kuwait. qnrA is the first plasmidmediated quinolone resistance identified however, in K. pneumoniae it is not as widespread as qnrB [25]. qnrA confers resistance to quinolones such as nalidixic acid and increases minimum inhibitory concentration (MIC) values of fluoroquinolone up to 20-folds [26]. It has also been shown that when both qnrA and aac-(6`)-Ib-cr are present in the same cell, the level of ciprofloxacin resistance is increased fourfold more than that conferred by qnrA alone [5]. Therefore the detection of qnrA2 gene in an ESBL producing K. pneumoniae from the catheter tip of a 14 year old female patient was alarming and a challenge for empirical antibiotic therapy such as fluoroquinolones in clinical settings. The rise of plasmidmediated fluoroquinolone resistance is concerning for antimicrobial treatment of K. pneumoniae whereby the only option left will be carbapenems.
In this study bla CTX-M-15 was the predominant βlactamase and in 93 % of the cases (25 out of 27 isolates) it was associated with the presence of aac(6)-1b-cr. We also identified qnrB1 and qnrS in isolates harbouring aac(6′)-Ib Ib-cr and bla CTX-M . These isolates were resistant to penicillins, most cephalosproins, β-lactamase inhibitors as well as fluoroquinolones posing a risk to combination βlactam/β-lactamase inhibitor therapy. The correlation between ESBL and ciprofloxacin resistance is an indication of the important contribution of the plasmid-mediated resistance to fluoroquinolone resistance in this region.
Urine samples comprised 67 % (18/27) of the specimens followed by wound swabs 15 % (4/27). We did not find any correlation between antimicrobial resistance characteristics or specimen type and PFGE profiles. In fact overall the PFGE profiles of the tested isolates were diverse implying that the spread of qnr genes is not concentrated to only a specific clonal lineage. A detailed longitudinal study is required to identify the clonal identity of ESBL-producing qnr-positive K. pneumonia isolates.

Conclusion
This manuscript reports the genotypic variation among strains of qnr positive K. pneumonieae in hospitalized patients in Kuwait. The presence of qnr and aac-(6`)-Ib-cr genes especially when they co-exist in ESBL-producing isolates contribute to the increase in fluoroquinolone resistance.
In conclusion the clinical importance of our findings is that the circulating MDR plasmids that contain qnr, aac-(6`)-Ib-cr and bla genes are concerning in hospitals as they increase the chances of K. pneumoniae adaptability to environmental pressures such as antibiotics. If guidelines are not in place regarding the strict use of antibiotics we will face antimicrobial therapy failure in hospitalized patients. Continuous monitoring of plasmid mediated fluoroquinolone ESBL-producing Enterobactericeae in clinical settings is crucial.