RAPD PCR Profile, Antibiotic Resistance, Prevalence of armA Gene, and Detection of KPC Enzyme in Klebsiella pneumoniae Isolates

The increasing prevalence of multidrug-resistant Klebsiella pneumoniae strains isolated from hospitals shows the limitation of recent antibiotics used for bacterial eradication. In this study, 81 K. pneumoniae isolates were collected from three hospitals in Tehran. Antibiotic susceptibility test showed the highest rates of resistance to cefotaxim (85.5%) and ceftazidime (78.3%), and the lowest rates of resistance were detected for colistin (16.9%), streptomycin (16.8%), and chloroamphenicol (21.7%). Eleven different resistance patterns were observed. Sixty-six out of 81 isolates (81.5%) were found to be multidrug resistant (MDR), and 35.8% of them belonged to A3 resistance pattern. 7.4% and 66.7% were KPC enzyme and armA gene positive, respectively. RAPD PCR assay of these bacteria showed 5 clusters, 16 single types, and 14 common types, and there was not any correlation between genetic patterns of the isolates and presence of resistance agents. Simultaneous detection of resistance-creating agents could be an important challenge for combination therapy of MDR K. pneumoniae-caused infections.


Introduction
In the 1980s, gram-negative pathogens were successfully defeated using cephalosporins, carbapenems, and uoroquinolones [1]. K. pneumoniae is one of the most widespread nosocomial pathogens causing urinary tract infections, bacteremia, and pneumonia in di erent parts of the world [2,3]. Extensive use of antibiotics and thereby the development of various resistance mechanisms have led to the emergence of MDR (multidrug resistance) K. pneumoniae [4]. KPC-(Klebsiella pneumoniae carbapenemase-) producing K. pneumoniae strains are the most common carbapenamaseproducing pathogens worldwide [5] that have also been reported in Iran in many studies [6][7][8]. Seventeen di erent types in the KPC family have been reported to date [9], among which one family is located on a conjugative plasmid that encodes resistance to gentamicin and tobramycin [10]. e armA gene, which is responsible for resistance to the majority of the aminoglycosides and is the most prevalent in Asia [11], was found to be located on the same plasmid of KPCproducing strains reported formerly from Italy [12], China [13], and Poland [14]. Coexistence of the resistanceinducing agents can result in the emerging of MDR K. pneumoniae strains. Epidemiological characterization of MDR K. pneumoniae can help to prevent the spread of resistant strains [15]. RAPD PCR (random ampli cation of polymorphic DNA PCR) is a simple, rapid, inexpensive, and widely used typing method which does not require advanced knowledge of DNA sequences of the target organism. RAPD typing has been successfully employed for epidemiological analysis of many bacteria [16] as well as clinical isolates of K. pneumonia [17].
is study aimed to detect the RAPD PCR ngerprint, antibiotic resistance patterns, and the presence of armA gene in clinical KPC-producing isolates and nonproducing isolates of K. pneumoniae. September 2015 to March  2016, clinical samples were collected from both inpatients  and outpatients attending three hospitals in Tehran, Iran. e samples were transferred to the microbiology laboratory, and biochemical tests were performed to identify K. pneumoniae. e pure isolates were stored at −20°C in Trypticase soy broth containing 20% glycerol.

Detection of armA Gene.
Bacteria were cultured in Luria-Bertani (LB) broth medium and were incubated overnight at 37°C. DNA was extracted from K. pneumoniae isolates by using Genomic DNA Isolation Kit (Global Gene Network Bio, cat. no. k-3000, South Korea) according to the manufacturer's instruction. e presence of armA gene was evaluated using PCR technique with the armA-F: TCGGAACTTAAAGACGACGA and armA-R: CCATTCCC-TTCTCCTTTCCA (designed in this study) sequences. e PCR product was analyzed by electrophoresis in a 1% (W/V) agarose gel in TBE bu er at 95 V for 45 min, stained with ethidium bromide, and observed under UV lighting using Gel Doc. en, DNA sequencing was performed by Bioneer Company (Korea), blasted in NCBI, and analyzed by Finch TV software.

RAPD PCR.
Molecular typing of K. pneumoniae isolates was performed using RAPD PCR analysis with the primer 640 (CGTGGGGCCT) (Faza Biotech, Tehran, Iran). Reaction mixtures (25 µl) contained: 12.5 µl (Exprime Taq Premix (2X) Mastermix (Global Gene Network Bio, cat. no. G-5000)), 7.5 µl Distilled water, 200 pmol of primer, and 3 µl DNA template. e program used was as follows: 5 min at 94°C followed by 36 cycles of 1 min at 94°C, 1 min at 36°C, and 2 min at 72°C followed by a nal extension at 72°C for 9 min. PCR products were then electrophoresed on 1% agarose gels and visualized by ethidium bromide staining. To determine the similarity rate among the isolates, they were analyzed by the unweighted pair-group method with arithmetic mean (UPGMA) using GelClust software.

Determination of Antibiotic Resistance and KPC Enzyme
Detection. Antibiogram showed that the resistance rates of the isolates were as follows: o oxacin 65%, cipro oxacin 68.7%, nor oxacin 66.3%, gentamicin 66.3%, amikacin 51.8%, tobramycin 56.7%, kanamycin 79.5%, ticarcillin 82%, streptomycin 16.9%, cefotaxim 85.5%, ceftazidime 78.3%, azetreonam 79.5%, imipenem 45.8%, trimethoprim 74.7%, chloramphenicol 21.7%, and colistin 16.9%. One colistinresistant isolate had a MIC of 8 µg/ml, and the other all had MIC values of 4 µg/ml. All the colistin-resistant isolates except one of them belonged to one hospital. Five colistinresistant K. pneumoniae exhibited resistance to imipenem of which one was positive for KPC production. e antibiotic resistance pattern of the isolates is shown in Table 1. e highest rate of the resistance belonged to the A9 pattern (64.2%) which found to be resistant to three ouroquinolons. Also, the lowest rate of resistance belonged to the A4 pattern (4.9%) which was resistant to cipro oxacin, o oxacin, noroxacin, imipenem, amikacin, and colistin. MDR is generally determined as resistance to three or more classes of antibiotics [19]. In the present study, 66 out of 81 isolates were resistant to 3 or more classes of antibiotics; thereby, 81.5% of the isolates were found to be MDR. Also, the modi ed Hodge test showed that 6 isolates (7.4%) were positive for KPC production with 3 resistance patterns (Table 2).

Molecular Detection.
e PCR showed that 54 isolates (66.7%) of the isolates had armA gene. Aminoglycoside resistance patterns in armA positive isolates were presented in Table 3.

RAPD PCR.
Genetic analysis of RAPD showed 30 distinct patterns from D1 to D30, as 5 distinct clusters ( Figure 1). e information of isolates is shown in Table 4. e isolates were considered as the same pattern if the level of similarity was ≥85%. e numbers of the isolates in each cluster (from cluster 1 to 5) were 18,19,14,26, and 4, respectively. e highest redundancy belonged to pattern D1, D8, and D20. ere was not any correlation between genetic patterns of the isolates and presence of armA gene or KPC production.
e isolates which showed 100% similarity, belonging to di erent clusters, all found to have armA gene.

Discussion
In the present study, 81 K. pneumoniae were isolated from clinical samples. Antibiotic resistance patterns and their relationships among di erent clonal isolates were investigated. Eleven antibiotic patterns were found (A1-A11), showing much diversity in the resistance patterns. In the study by Hassan et al. [16] and Ben-Hamouda et al. [17], di erent antibiotic resistance patterns were reported, too [20]. In the present study, 81.5% were MDR. e presence of multidrug resistance community acquired K. pneumoniae highlights the need for accurate planning to control and prevent of the dissemination of MDR strains. Most of the KPC-producing isolates harbored armA gene and were resistant to carbapenems, aminoglycosides, and uoroquinolones. According to our previous study, Real Time PCR showed an increased expression level of OqxAB and AcrAB e ux pumps in uoroquinolone-resistant isolates in comparison with the sensitive ones (data not shown). So, the role of e ux pump in creating uoroquinolone-resistant strains could be identi ed in these isolates [21]. Also, the aminoglycoside resistance rates suggested 16S rRNA methylase activity. In the present study, approximately 70% of the aminoglycoside-resistant strains carried the armA gene as Zhou et al. reported [22]. erefore, some of these K. pneumoniae isolates have three features of resistance: KPC, e ux pumps, and armA gene. So they can be resistant to uoroquinolones, cephalosporins, carbapenems and a spectrum of aminoglycosides as well. ese strains can turn out to an important challenge for community and hospital o cials by disseminating among the patients in hospitals and making the treatment process more di cult. Coexistence of the active e ux pump, armA gene, and KPC enzymes in K. pneumoniae can help to resist against the combination therapy. is hypothesis is also recommended by Zacharczuk et al. [14] and Jiang et al. [13] who observed KPC production and armA gene in clinical isolates of K. pneumoniae.
Furthermore, about 17% of the isolates were resistant to colistin, among which all but one isolate had MIC 4 µg/ml. Although colistin-resistant isolates were related to ve di erent clusters, 42.9% of them belonged only to the fourth cluster from one hospital, indicating a genetically speci c circulating cluster. Colistin is considered as e ective treatment against MDR and carbapenem-resistant bacteria, such as K. pneumoniae, but resistance to this agent has begun to emerge. So, more studies to determine the best treatment for infections caused by resistant K. pneumoniae are needed. In the present study, genotyping analysis showed di erent   genetic patterns among pathogenic K. pneumoniae isolates. Lim et al. [23], Ben-Hamouda et al. [17], Pai et al. [24], and Eftekhar and Nouri [25] performed separately the genotyping of clinical K. pneumoniae strains and showed that they were genetically diverse and heterogeneous. erefore, this tool has got the ability to identify related and unrelated isolates. e correlation between the antibiotic resistance patterns and RAPD analysis demonstrated that di erent genetic patterns had di erent antibiotyping pro les. Also, KPC-producing K. pneumoniae were found to belong to di erent clusters, and the results of RAPD PCR implicated that there is no correlation between the genetic patterns and presence of armA gene or KPC-producing K. pneumoniae, as Ma et al. [26] reported that nonclonally spread of ESBLproducing K. pneumoniae strains with armA or rmtB mediates aminoglycoside resistance. It seems that because most resistance genes are carried by the mobile genetic elements, they can easily transmit among the bacteria.
Of importance, the low number of the isolates was one of our limitations. Collecting more clinical isolates, using more   powerful discriminating typing methods such as PFGE and analysis of armA and bla kpc genes' expression level may improve the quality of our results in the following studies.

Conclusion
Although uoroquinolones seem to be a good choice for treating Klebsiella infections, several studies have shown increasing resistance of K. pneumoniae isolates to these agents. Aminoglycosides have been considered as an adequate therapeutic against both gram-negative and grampositive pathogens and also combination therapy with β-lactams and aminoglycosides is well accepted for the treatment of the systemic infections caused by K. pneumoniae, so simultaneous detection of resistance creating agents to uoroquinolones, β-lactams, and aminoglycosides in these strains is of clinical importance. Emphasis on the suitable use of antibiotics, e ective infection control measures, and identi cation of antibiotic resistance mechanisms by molecular procedures are necessary to reduce the incidence of infections caused by antibiotic-resistant organisms.

Conflicts of Interest
e authors declare that they have no con icts of interest.