Antibiotic Resistant, Virulence-associated Genes, Biolm and Eux Pump Gene Expression and Molecular Typing of Klebsiella Pneumoniae Strains Recovered from Clinical Samples

Multidrug-resistant (MDR) Klebsiella pneumoniae strains are one of the most important life-threatening nosocomial pathogens. In the current study, antibiotic resistant, virulence-associated genes, gene expression of eux pumps and biolm genes as well as molecular typing of K. pneumoniae strains were investigated. A total of 505 clinical specimens were collected from hospitalized patients and K. pneumoniae strains were isolated by standard microbiological methods. Antibiotic resistant prole, prevalence of virulence associated genes, biolm and eux pump genes were investigated. The gene expression analysis of biolm and eux pump genes were analused quantitative Real Time PCR. Moreover, molecular typing of K. pneumoniae strains using Repetitive element sequence-based PCR (rep-PCR) technique was also carried out.

quinolones, tetracycline, and chloramphenicol in MDR strains of K. pneumoniae [12]. The ability to form bio lms in K. pneumoniae strains, allows it to protect against the host immune system and antibiotics in MDR isolates and there are a lot of bio lm related genes including mrk (type 3 mbriae), mH-1 (type 1 mbrial adhesion) [13,14]. Studies show that the e ux pump plays an important role in antibiotic resistance and bio lm formation [15]. There are several studies showed a correlation of K. pneumoniae antibiotic resistance, e ux pump and bio lm formation ability [16]. In the study of Subramanian et al, 2012, indicated that 80% of bio lm forming isolates from 100 clinical samples showed an MDR phenotype [17]. Various Virulence factors such as those encoding regulators of mucoid phenotype A (rmpA), bacteriocin biosynthesis [enterobactin (entB), and serum resistance-associated outer membrane lipoprotein (traT) play an important role in the pathogenicity of K. pneumoniae strains [18].
Molecular typing of K. pneumoniae strains can be used to prevent infection in the hospital and also to nd the dominant genotype among the isolates can be useful in nding the source of infection and applying prevention protocols [19,20]. Due to the high prevalence of MDR strains of K. Pneumoniae in Iran, the aim of this study was to investigate the antibiotic resistance pro le and distribution of virulence genes in K. Pneumoniae strains, e ux pump and bio lm gene expression, as well as, molecular typing of K. pneumoniae strains using Repetitive element sequence-based PCR (rep-PCR).

Bacterial isolates and identi cation
A total of 505 clinical specimens including urine, blood, sputum and cerebrospinal uid (CSF) were collected from hospitalized patients were admitted to Baghiatallah and Imam Khomeini hospitals in Iran from January 2018 to July 2019. The K. pneumoniae strains were identi ed using conventional microbiological tests including catalase, MR-VP (methyl red -Voges Proskauer), lysine iron agar, Kligler agar, phenylalanine agar, urea agar, SIM (sul de, indole, motility), blood agar, and MacConkey agar [21]. The isolated strains were stored at -20 0 C in brain heart infusion broth containing 20% glycerol for further investigation.

Phenotypic detection of e ux pump
The phenotypic and qualitative detection of the e ux pump in Klebsiella pneumoniae strains was performed by Cartwheel method. Brie y, the plates of Muller Hinton Agar culture media containing ethidium bromide were prepared and the culture medium was divided into 8 parts and the bacteria in 0.5 McFarland turbidity concentration were streaked on plates. After 24 h of incubation at 37 0 C, the plates were studied under UV transilluminator. The strains that had e ux pumps did not show emission of uorescence [23].
Phenotypic detection of bio lm formation Phenotypic detection of bio lm formation was performed using Congo red agar test. Brie y, the K. pneumoniae strains was cultured in Brain Heart Infusion agar enriched with 5% (w/v) sucrose and Congo red based on Freeman et al. 1989. The K. pneumoniae strains which formed bio lm exhibited a dry dark crystalline colonies and considered as exopolysaccharides producers [24].

Quantitative detection of bio lm
Quantitative bio lm detection test was performed by plate microtiter method in 96 house plate. In summary, rst 100 μl of 24-hour culture of strains with OD = 0.1 were added into wells. After 24 h incubation, each well was washed twice with PBS and then, stained with crystal violet for 15 min.
Finally, the stained cells were solubilized in 33% (v/v) acetic acid and their adsorption was read at 570 nm OD 570 nm. Based on the type of bio lm, the strains of bio lm former were divided into three categories: strong bio lm, moderate bio lm, weak bio lm and negative bio lm producers. In addition, the standard strain K. pneumoniae ATCC 13833 and LB broth was considered as a positive and negative control, respectively [25].

Molecular detection of virulence associated genes
The genomic DNA of the strains was extracted by the DNA extraction kit (Bioneer, Korea) according to the manufacturer's protocol. The frequency of entB, Trat and rmpA virulence gene, mdtk, tolC and AcrAB e ux pump, mH-1, mrkA and mrkB bio lm associated genes were detected by PCR. The PCR conditions were as follow: Initial denaturing at 94 0 C for 5 min followed by 30 cycles, each cycle contained 1 min at 94 0 C for denaturation, 30 second for annealing (Table 1) and 60 S for extension steps and nally one cycle for nal extension at 72 0 C for 10 min. The primer of target genes are given in Table 1. Gene expression analysis of AcrAB and mrkA gene AcrAB e ux pump and mrkA bio lm gene expression analysis was done using quantitative Real Time PCR method in K. pneumoniae strains. Brie y, 100 µl of each bacterium was added into 96 well plate and incubated for 18 h at 37 0 C aerobically. Then, each well was washed using PBS and adherent cells were scraped off using LB broth. Subsequently, total RNA of collected strains were extracted using an RNA extraction kit (Qiagen, USA) according to instruction protocol.
The extracted RNA was converted to complementary DNA (cDNA) by a cDNA synthesis kit (Fermentase, Lithuania) based on manufacturer's instructions. In order to perform the Real Time PCR, each cDNA was used as a template in 20 µl nal volume containing 2 μl cDNA, 10 pmol of each primer (Table 1), and 10 µl Power SYBR Green PCR Master Mix (Applied Biosystems) using Bioneer Real-Time PCR equipment (Korea). The 16S rRNA was used as a housekeeping gene to normalize the levels of mRNA expression and the relative expression of AcrAB e ux pump gene was calculated using ΔΔCт method.
Molecular typing of MDR K. pneumoniae strains Molecular typing of K. pneumoniae strains was done using the Repetitive element sequence-based PCR (rep-PCR) method. The rep-PCR was performed using two following primers: Forward: REP1 5'-III ICG ICG ICA TCI GGC-3' Reverse: REP2 5'-ICG ICT TATCIG GCC TAC-3' as described previously. The PCR ampli cation products were visualized using electrophoresis on 1.5% agarose gel and stained with safe red. Finally, the banding patterns and size were determined via Image Lab 4.0. The obtaining results were analyzed by gel compare II software using Dice correlation coe cient and the UPGMA method [31].

Statistical Analysis
All tests of this study are repeated three times and one way ANNOVA test was used for statistical analysis. In addition, P <0.05 was also considered signi cant.

Phenotypic detection of bio lm formation and e ux pump
Phenotypic detection of bio lm formation was done using Congo red agar test and our results showed that 77% (77 strains) of isolates exhibited black colonies, which presumably indicate bio lm formation (Fig 1A). The other strains formed white colonies which re ecting no bio lm formation. Moreover, the Cartwheel results showed that 90% (90 isolates) of strains had e ux pumps (Fig 1B).

Quantitative bio lm production
The results of quantitative bio lm production test is shown in Table 2

Frequency of bio lm, e ux pump and virulence associated genes
The prevalence of bio lm, e ux pump and virulence associated genes are given in Table 3. The mrkA, mrkD and mH genes encoding type 1 and type 3 mbrial adhesion engage in bio lm formation were present in all bio lm former strains. Moreover, the e ux pump genes including AcrAB, TolC and mdtK were observed 41 (41%), 33 (33%) and 26 (26%) strains, respectively. The AcrAB was more prevalent in K. pneumoniae strains comparing to other e ux pump genes. In addition, the AcrAB e ux pump gene was more prevalent in urine samples comparing to other clinical specimens.
In addition, the virulence related genes including enterobactin biosynthesis gene (entB), outer membrane protein coding gene (traT) and mucoid phenotype A (rmpA) was seen in 80 (80%), 62 (62%) and 48 (48%) strains, respectively. As reported in Table 3, analysis of selected genes showed that bio lm was more pronounced among virulence associated gene positive than among negative strains. The entB virulence gene was detected in all blood, CSF and sputum isolates. There were 8 virulence pro les (V1-V8) Based on virulence detected gene and V1 was the most prevalent virulence type. isolates NO.) selected K. pneumoniae strains were used for AcrA and mrkA gene expression analysis. The results of Real Time PCR showed that AcrA and mrkA gene were up-regulated signi cantly in MDR isolates comparing to non-MDR isolates. There was a signi cant relationship between MDR isolates, AcrA and mrkA gene expression (P<0.05) (Figure 2).

Rep-PCR typing
According to the dendrogram, Repetitive element sequence-based PCR (rep-PCR) revealed 11distinct patterns of K. pneumoniae isolates (Figure 3). The 11 rep genotypes were designed rep1 to rep11. The rep type 4-7 were the most common and, followed by type rep 1, 9 and 3, which consisted non-MDR isolates (Group I:

Discussion
Multidrug-resistant (MDR) K. pneumoniae strains are an important cause of several life-treating infections, worldwide [32]. The extensive use of antimicrobial agents led to a high prevalence of MDR K. pneumoniae strains [33]. The increasing rate of K. pneumoniae strains resistant to multiple antimicrobials is a global public health problem [34]. In this study, the prevalence of MDR K. pneumoniae isolates was 92%. The high rate of MDR strains was also shown in other studies. Moreover, 48% and 47% of K. pneumoniae strains were resistant to imipenem and meropenem, respectively. From the results, it can be concluded that there has been a signi cant increase in carbapenem resistant K. pneumoniae isolates in Iran. It seems that the production of carbapenemase and metallobetalactamase have an important role in carbepenems resistance [35]. Our results is supported by some previous reports. Manjula et al, indicted 90.2% of isolates were MDR and the majority of MDR strains were resistant to a high range of antibiotics including penicillin, cephalosporin, uoroquinolone, aminoglycoside, and sulfonamide [36].
One of the possible reason for high rate of antimicrobial resistance is lack of strict policies for use of antibiotics in Iran. Another mechanism of multidrug resistant is e ux pumps which used by K. pneumoniae strains [37]. The e ux pumps could reduce the intracellular concentration of antibiotics which is as an important cause of bacterial survival [38]. In our study, the AcrAB e ux pump is most common e ux pump in K. pneumoniae strains comparing to mdtk and it was signi cantly correlated with MDR phenotype. Our results are consistent with other reports which indicated that the multidrug e ux pump system (AcrAB-TolC) in K. pneumoniae strains is responsible for antibiotics especially uoroquinolones such as cipro oxacin, tetracycline and beta-lactam antibiotics in MDR isolates.
In addition, 77% of K. pneumoniae strains were bio lm former and 89% of bio lm former were MDR. Until now, it has been shown a signi cant correlation between MDR phenotype and the bio lm forming ability of K. pneumoniae strains [39]. In addition, the relationship between antibiotic resistance and bio lm formation has also been studied in strains grown under the antibiotic dose especially in sub-minimum inhibitory concentration [40]. There was a signi cant relationship between strains of K. pneumoniae isolated from urine and bio lm formation. Most of urine originate strains exhibited strong bio lm capacity. In the current study, K. pneumoniae strains isolated from clinical specimens harbored high prevalence of e ux pump, bio lm and virulence associated genes. The virulence associated genes were also dominant in MDR strains. Type 1 mbriae ( mH-1) and Type 3 mbrial adhesion (mrkA and mrkD) are the most common bacteria cells adhesive agent which can mediate the binding of K. pneumoniae strains to endothelial and epithelial cells of urinary tracts cause to urinary tract infections [41,42]. Type 3 mbrial adhesion plays an important role in the bio lm formation of K. pneumoniae strains, however, their exact role in the bio lm formation remains unclear [43]. The results of Nirwati et al, 2019 study on bio lm formation in K. pneumoniae strains isolated from clinical samples showed that 85.63% of the strains were bio lm and that is more than the results of our study [44]. In our study, the mH, mrkA and mrkD were detected in all types of urine, blood and CSF isolates especially in urine isolates. The enterobactin biosynthesis gene (entB), serum resistance-associated outer membrane lipoprotein (traT) and regulators of mucoid phenotype A (rmpA) were detected in 80%, 62% and 48% of K. pneumonia strains. The traT gene encodes an outer membrane protein which plays an important role in conjugation and inhibition of complement cascade and act as invasin [45]. Molecular typing of bacteria is a very suitable method for differentiation of microbial pathogens and among the typing methods, rep-PCR technique is widely used for genotyping of bacterial strains. [46]. In our study, out of 100 K. pneumoniae strains, rep-PCR could differentiated into 11 distinct patterns and most MDR strains were put in the same patterns. Our data con rmed the Lai et al results which showed pathogenic K. pneumoniae strains are heterogeneous, because of variation in genome sequences [47]. Our results showed the correlation of MDR strains with rep-PCR patterns. However, the rep-PCR revealed no statistically signi cant correlation with virulence type. In addition, the rep-PCR results showed that the same rep-type in two studied hospitals indicated that the same clonal distribution of K. pneumoniae in two hospitals. Our ndings can help to interpretation of MDR K. pneumoniae outbreaks associated with speci c patterns in the future.

Conclusion
In our study, we reported the high prevalence of MDR K. pneumoniae strains with resistance to multiple antimicrobial agents, the ability to formed bio lm and presence of e ux pump, bio lm and virulence associated genes which can be a great barrier to treatment of K. pneumoniae related infections. Moreover, high genetic similarity of MDR strains in hospitals showed clonal dissemination of K. pneumoniae strains that requiring control tools. However, further studies are needed to investigate other epidemiological aspects of the K. pneumoniae strains.