Role of oqxA and oqxB Genes in the Development of Multidrug Resistant Phenotype among Clinical Klebsiella pneumoniae Isolates from Various Cases

This study investigated the prevalence of oqxA and oqxB genes and their effective roles in the development of multidrug resistant (MDR) phenotype among clinical isolates of Klebsiella pneumoniae. Out of 150 clinical samples, 50 (33%) isolates were recognized as K. pneumoniae according to the morphological and biochemical properties. The minimum inhibitory concentrations (MICs) assay revealed that the resistance values of the isolates were 43 (86%) against ceftriaxone (4≥64 μg/ml), 42 (84%) against ceftazidime (16≥64 μg/ml), 41 (82%) against cefepime (≥16 μg/ml), 21 (42%) against ertapenem (≥8 μg/ml), 18 (36%) against imipenem (4≥16 μg/ml), 15 (30%) against ciprofloxacin (≥4 μg/ml), 11 (22%) against levofloxacin (≥8 μg/ml), 45 (90%) against nitrofurantoin (128≥512 μg/ml), 36 (72%) against trimethoprime-sulfamethoxazole (≥320 μg/ml), and 4 (8%) against tigecycline (≥8 μg/ml). Genotype detection revealed that oqxA was found in 48 (96%) of K. pneumoniae isolates, whereas oqxB was found in 6 (12%) isolates. The MDR phenotype was observed in 40 (80%) isolates, of which 38 (95%) were harbored oqxA and/or oqxB genes. DNA sequencing of oqxA revealed the presence of three silent mutations. The phylogenetic tree of oqxA variants showed a significant deviation of these variants from K. pneumoniae species. The high prevalence of oqxA among K. pneumoniae isolates may contribute to the reduction of their susceptibility to multiple antimicrobial agents.


Introduction
Klebsiella pneumoniae is a Gram negative, non-motile, lactose fermenter, facultative anaerobic bacillus with a prominent polysaccharide capsule and belongs to the family Enterobacteriaceae [1]. It is an opportunistic pathogen that can colonize the mucosal surfaces of humans and cause serious nosocomial and community acquired infections, including urinary tract infections (UTIs), pneumonia, bacteremia, wound and burn infections, liver abscess, meningitis, cellulitis, and myositis [2].
Antimicrobial resistance becomes a worldwide problem in modern medicine that threatens public health. Since K. pneumoniae is a hospital-associated pathogen that is continuously treated with multiple antibiotics, it developed resistance abilities through multiple mechanisms against most common antibiotics in clinical usage [3]. This led to the emergence of multidrug resistant (MDR) K. pneumoniae that is responsible for high rates of morbidity and mortality due to the limited options of clinical treatment [3].
Efflux pumps are one of the resistance mechanisms employed by K. pneumoniae to be involved in both intrinsic and acquired resistance to antibiotics by decreasing intracellular concentrations of antibiotics and promoting accumulation of mutations [4]. The OqxAB is the predominant efflux pump found in K. pneumoniae that confers resistance against multiple antibiotics, including quinoxalines, quinolones, fluoroquinolones, chloramphenicol, trimethoprime, nitrofurantoin, and tigecycline [5].
The OqxAB efflux pump belongs to the resistance nodulation division (RND) family and consists of OqxA as a periplasmic part and OqxB as a transmembrane protein [4]. The oqxAB gene was first identified in 2003 on the pOLA52 plasmid in Escherichia coli from swine manure in Denmark [6]. Since then, oqxAB has been increasingly detected among K. pneumoniae as one of the plasmid mediated quinolone resistance (PMQR) mechanisms over the past decades [4,5].
The expression of the OqxAB efflux pump is regulated by RarA (regular of antibiotic resistance A) as an activator and OqxR (GntR-type transcriptional repressor) as a repressor [7]. Transposition of oqxAB gene from chromosome to plasmid has the ability to increase the expression level of OqxAB efflux pump in more than 80-fold, leading to the expansion of the MDR phenotypes [5,7]. This study aimed to investigate the prevalence of oqxA and oqxB genes and their role in the development of the MDR phenotypes in clinical isolates of K. pneumoniae.

Materials and Methods Collection of Samples
A total of 150 clinical samples was recovered from patients who suffered from UTIs, bacteremia, pneumonia, burns, and wounds infections from different hospitals in Baghdad. The collection of samples was performed during the period from July to October in 2018, following the instructions of the ethics committee of the Ministry of Health in Baghdad according to the official approval numbered 31864.

Bacterial Isolation and Identification
The clinical samples were cultured on MacConkey agar, Blood agar, Eosin Methylene Blue agar (Oxoid, UK), and CHROMagar Orientation (Pioneer, France). All culture media were incubated at 37°C for 24hrs [8]. Then, the pure colonies were selected and identified by several biochemical assays, including oxidase, catalase, and IMVIC tests [8]. The Vitek 2 Compact System (BioMerieux, France) was used to confirm the classic identification of bacteria [9].

Molecular Characterization of oqxA and oqxB Genes
DNA was extracted from overnight K. pneumoniae colonies that grew on MacConkey agar using Wizard Genomic DNA Purification Kit (Promega, USA) according to the information from the manufacturing company. The concentrations and purity of DNA were measured by the Nanodrop (BioDrop, UK) [12]. The extracted DNA was screened for oqxA and oqxB genes using the primers (Alpha DNA, USA) that are reported in Table-1. The lyophilized product of these primers was dissolved in sterilized deionized distilled water (ddDW) (Promega, USA) to obtain 100 pmol/µl and then diluted to 10 pmol/µl according to the information from the manufacturing company. The polymerase chain reaction (PCR) was performed using Thermal Cycler (BioRad, USA). The PCR reaction mixture (20 µl) consisted of 10 µl Go Taq Green Master Mix (Promega, USA), 2 µl template DNA (84.95 ng/µl), 1 µl of both forward and reverse primers, and 6 µl ddDW [12]. The PCR thermal cycler was programmed for oqxA and oqxB genes under the optimal conditions [13] that are reported in Tables-(2 and 3), respectively. The PCR amplification was verified by electrophoresis at 100 V for 80 min on 1% agarose gel stained with 0.5 µg/ml ethidium bromide in 1X TAE buffer (Promega, USA) using a DNA ladder (100-1500 bp), supplied by Promega (USA), as a molecular weight marker. The UV-Transilluminator (Major Science, Taiwan) was used for the observation of PCR products under 320nm UV light [12].

DNA Sequencing
The sequencing of two PCR products that referred to the oqxA from both ends was done by Genetic Analyzer (Macrogen Inc., South Korea) according to the instructions of the sequencing company (Macrogen Inc. Geumchen, Seoul, South Korea). Only clear chromatographs obtained from ABI sequence files were further analyzed, ensuring that the annotation and variations are not because of PCR or sequencing artifacts. The result was compared with the reference database available at the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov). The sequencing results of the PCR products of the sequenced isolates were edited, aligned, and analyzed along with their respective sequences in the reference database using BioEdit Sequence Alignment Editor Software Version 7.1 (DNASTAR, Madison, WI, USA) (http://bioedit.software.informer.com/7.1/) [14]. Each observed variation was translated into amino acids in a reading frame corresponds to the reference amino acid sequences using the Expasy online program (http://web.expasy.org/translate/). The referring sequences were manually aligned side by side with their mutant counterparts [15].

Phylogenetic Tree Construction
The observed PCR amplicon variants of oqxA genetic loci were compared with the neighbor homologous sequences using NCBI-BLASTn suite (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE_TYPE=BlastSearch). The BLAST results of the observed variants were aligned and constructed using Clustal Omega and Simple Phylogeny Tools, respectively (https://www.ebi.ac.uk/Tools/msa/clustalo/). The full inclusive tree, including the observed variants, was visualized as a polar cladogram using Figtree Tool (http://tree.bio.ed.ac.uk/software/figtree/).

Results and Discussion
Fifty isolates (33%) were identified as K. pneumoniae. The bacterial isolates were found at a high prevalence in urine 15 (30%) followed by blood 12 (24%), whereas lower prevalence appeared in sputum and wounds 9 (18%), and finally burns 5 (10%). Several studies revealed that the most common site of K. pneumoniae infections is UTIs followed by bloodstream infections, pneumonia, and burns and wounds infections [2,16,17]. Beside that, other studies collected K. pneumoniae from other cases, including pus, stool, cerebrospinal fluid, catheters, and eye and ear swabs [18,19,20]. The MICs listed in Table-4 exhibited that imipenem and ertapenem were more effective against K. pneumoniae isolates than other β-lactams, in which 18 (36%) (MIC 4-≥16 µg/ml) and 21 (42%) (MIC ≥8 µg/ml) of the isolates were resistant to these drugs, respectively. A previous study [21] demonstrated that all K. pneumoniae isolates were resistant for both agents. K. pneumoniae isolates showed high resistance values for other β-lactams that included ceftriaxone, ceftazidime, and cefepime, reaching 43 (86%) (MIC 4-≥64 µg/ml), 42 (84%) (MIC 16-≥64 µg/ml), and 41 (82%) (MIC ≥16 µg/ml), respectively. The resistance rates gained by another study [19] against these three agents were relatively in line with those of this study. The production of extended spectrum beta-lactamases (ESBLs), AmpC-β-lactamases, and carbapenemases is the main reason for the high level of resistance against cephalosporins and carbapenems [22]. The increasing use of carbapenems, as one of the last resorts in the clinical treatments worldwide resulted in the dissemination of K. pneumoniae carbapenemases (KPC) in the health care settings, causing a potential threat to the public health [23]. The challenge was more aggravated when carbapenemases were found to confer resistance not only to carbapenems but also to non β-lactams, particularly aminoglycosides and fluoroquinolones [22,23]. Regarding fluoroquinolones, 15 (30%) of K. pneumoniae isolates were resistant to ciprofloxacin (MIC ≥4 µg/ml), while 11 (22%) were resistant to levofloxacin (MIC ≥8 µg/ml). Higher resistance rates for both agents were obtained by an earlier investigation [16]. The resistance for these antibiotics is mediated by the mutations in the topoisomerases encoding genes (gyrA and parC), acquisition of PMQR, and alternations of cell permeability through the loss of porins and the overexpression of the MDR efflux pumps OqxAB [4,5]. The OqxAB efflux pumps mediate low level fluoroquinolones resistance, enabling the bacteria to survive in the face of these drugs as long as possible and enhancing the chromosomal mutations in the gyrA and parC that are associated with high level fluoroquinolones resistance [5,7].
Nitrofurantoin was the least effective antibiotic against K. pneumoniae isolates, whereas tigecycline was more effective with values of 45 (90%) (MIC 128-≥512 µg/ml) and 4 (8%) (MIC ≥8 µg/ml), respectively. Moreover, 36 (72%) (MIC ≥320 µg/ml) of the isolates were resistant for Trimethoprime-Sulfamethoxazole. The resistant rates of this study were higher than those reported by a previous study [24] for both Nitrofurantoin and Trimethoprime-Sulfamethoxazole and by another study [19] for tigecycline. In addition to fluoroquinolones, the MDR efflux pump OqxAB has been linked with the cross-resistance phenotype, in which this pump contributes to the reduction of susceptibility to multiple agents, including nitrofurantoin, trimethoprime-sulfamethoxazole, chloramphenicol, and tigecycline [5,7].
The MICs indicated that 40 (80%) of K. pneumoniae isolates were MDR. The bacterial isolates were considered MDR when they exhibited resistance to at least three or more unrelated agents [25]. Currently, MDR K. pneumoniae has become an escalating global threat in preventive medicine with the limited medical treatments and the difficulty of providing alternative findings [3]. The widely misused of the antibiotic therapy and horizontal transfer of the resistance genes participate in the emergence and development of MDR K. pneumoniae [3].
Genotypic screening revealed that oqxA was detected in 48 (96%) of K. pneumoniae isolates with a PCR product of 392 bp, whereas oqxB was found only in 6 (12%) of isolates with a PCR product of 512 bp, as shown in Figures-(1 and 2), respectively. The studies conducted by several researchers [19,26,27] showed similar high incidences of oqxA among the used K. pneumoniae isolates. In addition, a number of studies [19,28,29] exhibited that oqxB was more prevalent among the isolates than that in this study. The high prevalence of oqxA in K. pneumoniae isolates represents a potential reservoir for the spread of this gene [5]. The horizontal transfer and transposition of these genes from chromosome to plasmid enhance the ability to increase the level of MDR phenotypes [3].
As previously mentioned, MDR phenotype was observed among 40 (80%) of K. pneumoniae isolates, of these 38 (95%) harbored oqxA and/or oqxB (Table-5) that play an influential role in broaden bacterial resistance for various classes of antimicrobial drugs [4]. The correlation between phenotype (MICs) and genotype (PCR) of fluoroquinolones resistance revealed that 17 (34%) of K. pneumoniae isolates showed resistance against ciprofloxacin and/or levofloxacin, among which 16 (94%) possessed oqxA and/or oqxB, as illustrated in Table 5. Earlier studies [26,27,28,29] explained that the reduced susceptibility of the used K. pneumoniae isolates against multiple agents, particularly fluoroquinolones, may be due to the high prevalence of oqxA and/or oqxB among their isolates.   DNA sequencing was applied on the oqxA-A (KP23) and oqxA-B (KP13) isolates in order to exhibit the MDR phenotype. The results presented in Figure-3 revealed that no mutation was observed in the DNA sequence of the isolate oqxA-A (KP23), although it showed resistance for more antibiotics that used in this study. This can be explained that isolate may be harbored other MDR encoded genes that did not detect by this study. Otherwise, three point mutations were seen in the DNA sequence of the isolate oqxA-B (KP13). Two of these mutations were transition, in which led to the replacement of guanine with adenine (G˃A) at the position 54 and cytosine with thymine (C˃T) at the position 210. The third mutation was transversion that led to the replacement of guanine with cytosine (G˃C) at the position 326. The translation of protein ( Figure-4) revealed that these point mutations didn't cause any substitution in the amino acid sequences (silent mutations). The mutant DNA sequence of the isolate KP13 was registered in NCBI at the accession number LC381732. The phylogenetic tree (Figure-5) indicated the presence of seven species all over scanned of oqxA variants sequence-related species. The total number of the aligned DNA sequences in this tree was 100. In relation to both oqxA variants, the comprehensive involved organisms were included K. pneumoniae, Enterobacter hormaechei, Escherichia coli, Escherichia albertii, Salmonella enterica, and Salmonella typhimurium. This tree did not provide enough information about the real identity of the local isolates of K. pneumoniae. This fact is reflected from the position occupied by the nonmutant oqxA-A (KP23) variant, in which this variant was positioned very close to three species, namely Salmonella enterica (MF418180.1), Escherichia coli (KY075659.1), and K. pneumoniae (CP014647.1), respectively. Furthermore, the mutations observed in oqxA-B (KP13) variants (g.54G>A, g.210C>T, and g.326G>C) did not go further than these three species. This sort of positioning adds another layer of confirmation about the limited power of oqxA fragment based tree in providing concrete information about the actual identity of such microbiological identified K. pneumoniae local isolates [5,6]. Nevertheless, oqxA gene is found to be commonly located on the chromosome of K. pneumoniae isolates [4,5]. So, K. pneumoniae may constitute a potential reservoir for the spread of oqxA gene, posing a threat to public health [5]. Moreover, the phylogenetic tree constructed by [7] which included various types of Enterobacteriaceae, suggested that oqxAB was originated from the chromosome of K. pneumoniae.

Conclusion
The high prevalence of oqxA among K. pneumoniae isolates may contribute to reduce the bacterial susceptibility to multiple antimicrobial agents, leading to increase the selection of MDR phenotype among the studied isolates. K. pneumoniae may represent a potential reservoir for the spread of this gene.

Acknowledgment
We thankfully the administration of Medical City Hospitals in Baghdad to help us during the collection of samples.