bla KPC-2 and bla OXA-48 producing Klebsiella pneumoniae found in a Turkish hospital in the Balkans

Carbapenem-resistant Klebsiella pneumoniae carrying blaKPC-2, blaOxa48, and other metallo-lactamases (MBLs) are difficult to treat. This study was conducted to study the phenotypic and genotypic features of carbapenem resistance in isolates of K. pneumoniae isolated from a tertiary care hospital in the European region of Turkey. Isolates of K. pneumonia (n=100) resistant to at least one carbapenem (imipenem, meropenem, or ertapenem) were collected from 60 patients for 14 months. Carbapenem resistance was determined via the VITEK-2 system and the E-test confirmed this. The Modified Hodge Test (MHT) and Double Disc Synergy Test (DDST) were performed. Genes were analyzed by Sybr-Green real-time, multiplex and convensional polymerase chain reaction (PCR). Phylogenetic relatedness was analyzed by ERIC-PCR. The rate of resistance by E-test against ertapenem, imipenem, and meropenem were 98, 94, and 72%, respectively; 21% of isolates were somewhat susceptible to tygecycline. The MHT positivity was 98% and DDST was negative in all cases. There were 19 and 24% of isolates positive for blaKPC-2 and blaOxa-48, respectively. ERIC-PCR showed that all blaKPC-2-positive isolates were branched into two main clusters with 80.5% similarity. The results indicate that blaKPC-2 mediated carbapenem-resistant Klebsiella pneumoniae (CRKP) infection is spreading in Turkey and blaOXA-48 endemicity continues to be a serious problem. The molecular determination of carbapenemases will be useful for patients with concurrent carbapenem-resistant Enterobacteriaceae (CRE) infections. This could prevented outbreaks and complications.

and the extended spectrum beta-lactamases (ESBLs) (Meletis, 2016).The emergence and dissemination of carbapenem-resistant Gram-negative pathogens including Pseudomonas aeruginosa, Acinetobacter baumannii as well as Enterobacteriaceae and Gram positive bacteria (e.g., Staphylococcus and Streptococcus species) is a significant contributor to patient morbidity and mortality (Patel et al., 2013).The emergence and spread of carbapenem-resistant Enterobacteriaceae is a major public health challenge.The most prevalent carbapenemases in Klebsiella pneumoniae are K. pneumoniae carbapenemases (KPC) in Amber class A which was first identified in 1996 in the United States (Yigit et al., 2001) and it is now endemic to several noncontiguous areas of the world especially China, Israel, Greece, and South America (Munoz-Price et al., 2013;Li et al., 2016).KPC-producing K. pneumoniae has only recently been reported in Turkey (Labarca, 2014).The other prevalent enzyme is Oxacillinase (OXA-like) carbapenemase in class D; specially blaOxa-48 type with strong carbapenem-hydrolyzing activity was identified first in Turkey (Poirel et al., 2004).Over the years, numerous blaOxa-48 type carbapenemase-producing Enterobacteriaceae have been reported in Belgium (Cuzon et al., 2008), Morocco (Benouda et al., 2010), Israel (Goren et al., 2011), Tunisia (Saidani et al., 2012), United Kingdom (Thomas et al., 2013), Lebanon (Hammoudi et al., 2014), India (Srinivasan et al., 2015), Argentina (Pasteran et al., 2012) and Czeck Republic (Skalova et al., 2017).The increased frequency of report shows that it has successfully spread and has even become endemic (Carrer et al., 2010;Nazik et al., 2014;Sahin et al., 2015).The main question still remains unclear; whether the carbapenemases spread within Enterobacteriaceae is chromosomal or whether horizontal transfer because the gene encoding this enzyme is located in the plasmid; for instance ST258 specific clone is associated carriage of blaKPC-2 and blaKPC-3 genes, is considered to be responsible for expansion of resistance (van Duina and Doi, 2017).
The aim of this study was to describe data collected for 14 months regarding the clinical K. pneumoniae isolates from intensive care unit (ICU) patients as well as their microbiologic and molecular features.The site was Trakya University Health Research and Application Center and State Hospital located in the Balkans.

Sample collection and bacterial diagnosis
This study was performed at a 1100-bed tertiary care center.A total of 100 K. pneumoniae isolates were isolated from 60 patients.These patients were being treated at different wards between January 2013 and February 2014.The clinical samples consisted of blood, endotracheal aspirate, urine, tissue biopsy, central venous catheter, abscess fluids, or wound.An automated VITEK-2(BioMerieux, Bruz, France) system was used to identify the isolates.Bukavaz et al. 371 The isolates were stored in the CryoBilles (AES Laboratoire, France) medium tube at the -80°C until molecular analysis.

Antimicrobial drug susceptibility testing and phenotypic screening of carbapenemase production
Antimicrobial susceptibility testing was conducted by automated VITEK-2 susceptibility card AST-GN325 (BioMerieux, Bruz, France) containing imipenem and meropenem tests according to the manufacturer's instructions.The MIC calling range for meropenem and imipenem on the AST-GN28 card was ≤0.25 to ≥16 µg/ml in doubling dilutions.The Double Disc Synergy Test (DDST) and Modified Hodge Test (MHT) were used to phenotypically screen metallo--lactamases (Ramana et al., 2013).An imipenem (10 µg) disc and indicator isolate E. coli ATCC 25922 were used for MHT.

Detection of carbapenemase genes
Three different types of PCR techniques were applied via published primers (Table SD, supplementary data).SYBR-Green real-time PCR coupled with Melt-Curve analysis for blaKPC and blaNDM.
The PCR products were sequenced at Med Santek (İstanbul, Turkey) using the same primers used in PCR reactions with ABI 3100 Capillary Electrophoresis sequence detection system (Applied Biosystem, CA).

RESULTS
During the study period, 100 carbapenem-resistant K. pneumoniae (CRKP) isolates (one is susceptible to all antibiotics; clinical negative control) were isolated from different clinical materials from 60 patients.The maximum of four isolates of identical species isolated from different clinical materials or at the different dates of isolation were included.The isolates were obtained from blood (36%), urine (22%), endotracheal aspirate (21%), intravenous catheter (IVC) (15%), abscess aspiration (3%), sputum (2%), and wound tissue (1%).Most (76%) patients were in the intensive care unit and the other clinics were the internal medicine clinics (oncology, hematology, physical therapy and rehabilitation) (12%), surgical clinics (7%), and the emergency unit (5%).Antibiotic susceptibility test results indicated that all isolates were resistant to ampicillin, amoxicillin-clavulanic acid, and piperacillin-tazobactam. Between 94 and 68%   2. The MHT result was positive for 98% of the isolates, but the DDST was negative for all.Molecular analysis revealed that 19 and 24 isolates were positive for blaKPC-2 and blaOXA-48 genes, respectively.The blaKPC-2 positive isolates were isolated from 17 patients in the ICU.Four of the isolates were isolated from both the endotracheal aspirates and blood samples of two patients.The characteristics of the patients and the carbapenem MIC values are shown in Table 3. SYBR-green RT-PCR was used for only endpoint detection of blaKPC-2.The amplicon was distinguished by its specific Tm value.The melting curve analysis of the blaKPC-2 gene indicated that the products peaked at 89°C under determined conditions (Figure 1).
ERIC-PCR results showed that blaKPC-2 was positive in 17 out of 19 (two isolates were not successfully clustered).These isolates were grouped into two main clusters (A and B) that shared ~81% similarity.Cluster A was dominant with 10 isolates (~92% similarity), and cluster B had 7 isolates with ~86% similarity.The phylogenetic tree indicated a group of isolates called A 1 /A 2 that is the control isolate for K. pneumoniae ATCC ® -BAA 1705 TM .Cluster A 1 was isolated from Internal Medicine Intensive Care Units (IMICU) in the middle of 2013.İsolates B 1 /B 2 were from the Surgical Intensive Care Unit (SICU) in early 2014.Six isolates (60% of cluster A isolates) (A 1a ) displayed a single profile with 100% similarity indicative of dissemination.The isolates with five and six numbers were in A 1a , and 14 and 15 were in B 2 groups.These were in two different patients.All patient isolates shared the same patterns (Figure 2).It was found that 24 of the K. pneumoniae isolates carried blaOXA-48 that was isolated from 14 patients; six had two or more identical isolates isolated All isolates were negative for the rest of the MBLs: blaGES, blaNDM-1, blaVIM-1, blaVIM-2, blaGIM, blaSIM, and blaSPM.But the remaining 56 out of 99 isolates (57%) lost ompK 36 and 44 out of 56 (79%) isolates lost both ompK35 and 36.OmpK35 and 36 porin loss was also detected in 14 and four blaOXA-48 and blaKPC-2 positive isolates, respectively.
Before DNA sequence analyses, selected blaKPC-2 and blaOXA-48-positive samples were subjected to conventional PCR.The product sizes were confirmed on agarose gel electrophoresis (Figure 3).The DNA sequences were performed in forward directions for blaKPC-2 and both directions for blaOXA-48.The alignment of all sequenced DNA showed high identity with blaKPC-2 and blaOxa-48; sequences are available online through the GenBank databases with the accession numbers: MG171145, MG171146, MG171147, MG171148, MG171149, MG171150, MG171151, MG171152, MG171153, and MG171154.

DISCUSSION
Carbapenem-resistant K. pneumoniae is a major problem in nosocomial infections with high mortality rates especially in immunocompromised patients in the intensive care unit (Patel et al., 2008;Ulu et al., 2015).CRE were listed as one of the most urgent antibiotic resistance threats by the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) (Band et al., 2018).According to the CDC tracking program; KPC and OXA-48 type CRE have  spread rapidly, with cases reported in all 50 and 26 states in USA, respectively (CDC, 2017).The European Survey on Carbapenemase-Producing Enterobacteriaceae (EUSCAPE) reported that Greece, Italy, Malta, and Turkey have endemic carbapenemase-producing Enterobacteriaceae (CPE).Greece has VIM and KPC, Italy has KPC, and Malta and Turkey have OXA-48 (Dortet et al., 2017).The first blaKPC-2-producing K. pneumoniae in Turkey was reported in 2014 in Istanbul.It was subsequently found in 4.052 Escherichia coli isolates including two blaKPC-2 isolates and five blaOXA-48 positive isolates (Kuskucu et al., 2016).
To the best of our knowledge, this is the third and one of the largest study to identify Turkish blaKPC-2 and the first one to use real-time PCR for screening of blaKPC in Turkey.The blaOXA-48 was first identified in Turkey and then subsequently emerged in the Middle East, India, Europe, North Africa, and Saudi Arabia (Aktas et al., 2008;Azap et al., 2013;Carrer et al., 2008Carrer et al., , 2010;;Daikos and Markogiannakis, 2011;Gulmez et al., 2008;Karabay et al., 2016;Al-Zahrani et al., 2018).The blaOXA-48 is frequently carried in K. pneumoniae but it was also identified in other Enterobacteriaceae members like E. coli, Proteus mirabilis, Enterobacter species (Fursova et al., 2015;Yu et al., 2017).Here, a nearly similar positive number of blaKPC-2-and blaOXA-48 was detected (2017) reported that the tigecycline resistance rate was 18% in the USA and 100% in Taiwan and Greece (Bathoorn et al., 2016).The previous reports indicate that meropenem can be used for screening of CPE (Fattouh et al., 2016).However, in this study, the meropenem resistance rate was not significantly increased in blaKPC-2-positive isolates (90% vs. 70%; intermediate resistance was not included).The mortality rates for blaKPC-2 and blaOXA-48 positive isolates were 59% and 50, respectively.All patients had serious multiple underlying medical comorbidities.The clonal relationship of the blaKPC-2-positive isolates suggests that the epidemic has expanded beyond the hospital and resistance is spreading to other cities as well as to Greece and Bulgaria where medical tourism is popular.In addition, analyses of multi-locus sequence typing (MLST) of blaKPC-2 is worth further study because it can identify resistant and virulent clones like ST258 (in America and Europe) and ST11 (in Europe and Asia) (Andrade et al., 2014;Tzouvelekis et al.,  In contrast, other Turkish studies have found coexpression of MBLs (VIM, IMP and NDM) or AmpC type enzymes (CMY-1,2, CTX-M and SHV) with blaOXA-48 and blaKPC2 (Baran and Aksu, 2016;Karabay et al., 2016;Cizmeci et al., 2017).However, OmpK35 and 36 porin loss was found in fourteen blaOXA-48 and four blaKPC-2 positive isolates.Furthermore, 56 of 99 isolates (57%) showed moderate to high intensity carbapenem resistance, but no production of carbapenemases.These lost ompK 36; 44 out of 56 isolates (79%) lost both ompK35 and 36.This study shows that colistin (97% susceptibility) and tigecycline (79% susceptibility) might still be treatment options for CRKP infections in our center.Nevertheless, all isolates had aminoglycosides resistance; 32 to 65% of the isolates were blaOXA-48 and blaKPC-2-positive.The co-existence of plasmidmediated blaKPC-2 carbapenemase in K. pneumoniae with the 16S rRNA-methylase rmtB-encoding gene offers aminoglycoside resistance.It allows methylate binding site of drugs used clinically (Sheng et al., 2012).16S rRNA-methylase rmtB genes were screened by single PCR in carbapenemase-positive isolates, but none of the isolates were positive.

Conclusion
SYBR-Green RT-PCR is economical and useful tool for rapid screening of beta-lactamases rather than conventional PCR when coupled with post amplification melt-curve analysis.This offers reliable and discriminative results for carbapenemases.Furthermore, this study indicates that blaKPC-2-mediated CRKP infection is spreading in Turkey.Furthermore, blaOXA-48 is endemic and continues to be a serious problem.This situation is alarming and should be investigated via molecular characterization to find efficient antibiotics that decrease CPKP mortality.Molecular determination of carbapenemases was proposed to be useful for patients with concurrent CRE infections.This could prevent outbreaks and complications.

Figure 1 .
Figure 1.Amplification plot vs. cycle and Melt Curve plot; negative derivative of the normalized fluorescence generated by reporter during PCR amplification vs. temperature of K. pneumoniae blaKPC-2 positive control isolate BAA 1705.

Figure 3 .
Figure 3. Detection of blaKPC-2 and blaOXA-48 by conventional PCR.The amplicons were separated by agarose gel electrophoresis and visualized with etidium bromide staining.First lane on both gels shows the molecular size marker (100 to 1000 bp) and lanes 1-9 (on first gel), 1-6 (on second gel) shows the presense of carbapenemases.NC: Negative control, PC: positive control.

Table 1 .
MIC values of the strains according to the VITEK-2 system.

Table 2 .
Susceptibility of the strains for carbapenems and tigecycline according to E-test.

Table 3 .
Demographic, clinical and laboratory characteristics of the patients infected with CRKP positive for blaKPC-2.

Table 4 .
Demographic, clinical and laboratory characteristics of the patients infected with CRKP positive for blaOXA-48.