Virulence factors, antibiotic resistance patterns, and molecular types of clinical isolates of Klebsiella Pneumoniae

ABSTRACT Background Klebsiella pneumoniae is armed with a wide range of antibiotic resistance mechanisms that mostly challenge effective treatment. The aims of the current study were to identify the clinical strains of K. pneumoniaealso to determine their phenotypes and molecular characterization related to antimicrobial resistance and virulence genes. Research design and methods In this investigation, clinical specimens from different hospitals located in Tehran, Iran, were collected during a nine-month period (December 2018 to August 2019). The K. pneumoniae strains were isolated and identified through standard microbial and biochemical assays. Additionally, disk diffusion, combined disk, Modified Hodge Test (MHT) and PCR were performed for antibiotic resistance and virulence gene analysis, respectively. Results Eighty-four isolates of K. pneumoniae were subjected to the study. According to the combined disk and modified Hodge test results, 27 (52%) and 15 pathotypes (62.5%) out of resistant strains of isolated K. pneumoniae were detected as ESBL and KPC producers. The virulence genes of mrkD (94%) and magA (11%) were the highest and lowest among isolates, respectively. Conclusions The high prevalence of antibiotic resistance and virulence genes in conjunction with a significant relationship between the strains revealed a high pathogenic capacity of the isolated pathotypes of K. pneumoniae.


Introduction
K. pneumoniae is a gram-negative, non-motile bacillus of the Enterobacteriaceae family with a polysaccharide capsule, which is substantial in pathogenesis and in its ability to prevent phagocytosis [1]. This bacterium is an opportunistic pathogen and one of the most common causes of nosocomial infections ranging from pneumonia, meningitis, liver abscess, urinary tract infection (UTIs) and wound infection to bacteremia and sepsis [2,3].
K. pneumoniae pathogenesis is associated with the presence of certain virulence genes that encode virulence factors and allow it to attack the mammalian immune system, which could lead to a variety of diseases [4]. Over the years, frequent use and contact with antibiotics in hospitals have created antibiotic-resistant K. pneumoniae strains, limiting available treatment options for medical intervention against Klebsiella infections, and therefore caused many problems for medical staff as well as the patients [5]. In the absence of correct and complete knowledge of antibiotic resistance mechanisms in bacteria, particularly K. pneumoniae, not only the treatment is extremely challenging, but it can inadvertently assist in the intensifying drug resistance.
Antibiotic resistance in gram-negative bacteria occurs due to enzymatic and non-enzymatic mechanisms [6,7]. Enzymatic pathways are imposed by the expression of antibiotic inactivating enzymes, while non-enzymatic pathways develop mainly because of gene mutations. This results in resistance due to changes in efflux pumps, membrane permeability or target molecules [8]. Genes encoding resistance enzymes can be derived either from the bacterium itself or from Miniature Inverted Transposable Elements (MITEs) such as the plasmid encoding beta-lactamases or aminoglycoside-modifying enzymes [9,10]. Encoding of beta-lactamases and carbapenemase enzymes by the genes is the most important mechanism of antibiotic resistance in K. pneumonia, as these enzymes lead to multidrug-resistance (MDR) and therefore pose major problems to the treatment of dangerous bacterial infections [11][12][13]. MDR has been defined as resistance to at least one antimicrobial drug in three or more antimicrobial classifications. Extended-Spectrum Beta-Lactamases (ESBLs) can inactivate broad-spectrum cephalosporins, monobactams, and penicillin such as class A beta-lactamases (TEM-1, TEM-2, and SHV-1), which form resistance to ampicillin, amoxicillin, and firstgeneration cephalosporins. Mutations in these genes trigger resistance to the third-generation cephalosporins [14,15]. In addition to penicillin and cephalosporins, CTX-M enzymes are also a group of ESBLs that impose resistance to oxyimino-beta -lactams [16]. Another group of beta-lactamases commonly found in K. pneumoniae isolates are K. pneumoniae carbapenemases (KPCs), which inflicts resistance to carbapenem antibiotics [17]. A carbapenemase identified in the accessory genome of K. pneumoniae is New Delhi Metallo-Beta-Lactamase 1 (NDM-1), a class B Metallo-beta-lactamase (MBL) encoded by the plasmid, and other examples of b-lactamases include carbapenemases (bla KPC and bla OXA48 ) [18]. It could be worth mentioning that resistance to aminoglycosides is related to overexpression of drug metabolizing enzymes such as aminoglycoside N-6ʹ-acetyltransferase-Ib AAC (6ʹ)-Ib as well as 16S rRNA methyltransferases such as ArmA and RmtB [19]. Virulence-associated genes include coding regulators of mucoid phenotype A (rmpA), type 1 and type 3 adhesines (fimH-1, mrkD), aerobactin (iron siderophore) synthase (iucC), bacteriocin biosynthesis [enterobactin (entB) and yersiniabactin (irP-1)], and serum-dependent outer membrane lipoprotein (traT), which due to the number and function of these genes, they will play a significant role in the pathogenicity of K. pneumoniae strains isolated from nosocomial infections [20,21].
The increasing prevalence of antibiotic resistance in pathogens of human and animal populations is now one of the most significant global health issues. Due to these antibiotic resistance developments in K. pneumoniae and its various mechanisms, investigation of resistance pattern may lead to appropriate prescription of antibiotics, which results in more rapid improvement of related nosocomial infections [22,23]. Additionally, this could in turn help inhibit new resistance patterns by omitting the antibiotics with borderline sensitivity from therapeutic regimens in certain regions. Thus, the current study aimed to evaluate the phenotypic and genotypic nature of antibiotic resistance and virulence genes in K. pneumoniae strains isolated from clinical specimens in Shahr-e-Qods hospital. The relationships between the strains were also investigated by PCR (ERIC-PCR).

Bacterial strains
In the present study, during a 9-month period from December 2018 to August 2019, clinical samples suspected for K. pneumoniae infections were collected from the patients admitted to several hospitals in Tehran, Iran.

Bacterial culture
Following the initial culture on Eosin-methylene blue agar (EMB), blood agar, and MacConkey agar media, the samples were incubated for 24 hours at 37°C. The resultant bacterial colonies were examined by different diagnostic and biochemical tests, including Triple Sugar Iron (TSI), Simmons citrate agar, urease and Methyl Red Voges Proskauer (MRVP), and then cultured on Sulfur-Indole-Motility (SIM) medium for further biochemical analysis. All identified K. pneumoniae strains were stored in skim milk agar stock for further analysis [24].

Antibiotics susceptibility testing
Disc fusion (Kirby-Bauer) method was used for the evaluation of antibiotic resistance by culturing the isolates on the Mueller-Hinton agar medium. Using following antibiotic discs: including tetracycline (30 µg), imipenem (10 µg), gentamicin (10 µg), meropenem (10 µg), amikacin (30 µg), ceftazidime (20 µg), cefotaxime (30 µg), ceftriaxone (30 µg), aztreonam (30 µg), and piperacillin/tazobactam (100/10 μg). The colonies from overnight cultures (16-24 hours) were dissolved in sterile saline to obtain a bacterial suspension with a turbidity corresponding to the 0.5 McFarland standard unit. The suspension was cultured on Mueller-Hinton agar whole surfaces, and the antibiotic discs were placed at regular intervals (7 discs per plate and 2 plates per microbe were used). After incubating the plates for 18-24 hours at 37°C, the diameter of the growth inhibition zone was measured. These diameters of the antibiogram growth inhibition zones were measured according to the CLSI guidelines, and the strains were classified based on how they reacted to each drug into three groups: sensitive, intermediate, and resistant [25,26].

ESBL screening by Combined Disk Test (CDT)
K. pneumoniae isolates' strain resistances to third-generation cephalosporin, suspected as ESBLs producers, were tested by the Combined Disc Test (CDT). Using 0.5 McFarland turbidity standards, a proper bacterial suspension was prepared from newly grown (16-24 hours) colonies of isolates and cultured on Mueller-Hinton agar using a sterile cotton swab. Afterward, discs containing the following list were used: ceftazidime (30 μg), cefotaxime (30 μg) and cefepime (30 μg), and ceftazidime + clavulanic acid (30 μg/10 μg), cefotaxime (30 μg) + clavulanic acid (30 μg/10 μg) and cefepime (30 μg) + clavulanic acid (30 μg/10 μg). The antibiotic discs were placed on the Mueller-Hinton agar, and the plates were incubated at 37°C for about 18 to 24 hours. After incubation, the diameter of the growth inhibition zone around the cephalosporin discs with and without clavulanic acid was measured and compared. An increase in the inhibition zone diameter of >5 mm was proved as ESBLs' production [27].

Modified Hodge Test (MHT) for detection of carbapenemase (KPC)-producing isolates
The modified Hodge Test (MHT) assay is a simple phenotypic test to detect the presence of the carbapenemase enzyme in bacteria [28]. Thus, based on a 0.5 McFarland turbidity standard, a suspension of E. coli ATCC 25922 was prepared in 5 ml of Mueller-Hinton broth or saline. The E. coli ATCC 25922 suspension was then diluted to 1:10 by adding 0.5 ml of it to 4.5 ml of Mueller-Hinton broth or saline. Next, a bacterial grass culture was performed from the diluted solution on Mueller-Hinton agar and dried for 3-5 minutes at room temperature. A disc of meropenem (10 μg) or ertapenem (10 μg) was then placed in the center of the plate. In a straight line, from the edge of the disc to the edge of the plate, the desired isolate was cultured. The same procedure was repeated for the other isolate in another direction. Three organisms were examined on each plate. Incubation was performed overnight at 35°C for 24 hours. Then, the presence of cloverleaf was examined at the intersection of the studied organism and E. coli 25,922 in the growth inhibition zone, indicating a positive result [29,30].

DNA extraction
The bacterial genome was extracted using the boiling method. The bacteria were first cultured in 10 ml of BHI broth medium and incubated overnight. After examining the turbidity of tubes at 600 nm, proper OD was ≥4 for genome extraction. The culture media were transferred to 2 ml microtubes and centrifuged at 9000 rpm for 10 minutes. Then, the supernatant was rinsed out, and 400 mL of distilled water was added. This was then followed by boiling for 15 minutes at 100°C in a Bain-Marie bath. The microtubes were centrifuged again for 10 minutes at 6000 rpm, and 100 μl of the supernatant, known as the extracted genome, was collected using a sampler and transferred to a new microtube. Finally, the DNA adsorption rate was measured using a spectrophotometer at 260 nm, and the extracted genomes were stored at −20°C [31].

Molecular detection of antibiotic resistance and virulence genes by Polymerase Chain Reaction (PCR)
In this study, the detection of genes involved in antibiotic resistance and virulence, including bla CTX-M , bla SHV , bla TEM , bla KPC , bla NDM , aac (6ʹ)-Ib, armA, IrP-1, rmpA, magA, and mrkD, was investigated using PCR. All the primers used in this study are listed in Table 1. The PCR protocols were performed according to the manufacturer's instructions (Amplicon 2X). The PCR products were loaded on 1% agarose gel, the green bands were observed by using UV light, and then it was analyzed using a gel documentation system [23]. Lane M contained a DNA marker (100 bp DNA ladder) for faster identification of PCR products or amplicons. The electrophoresis was performed with a potential difference of 80 V for one hour, and the stained gels were inspected using a Bio-Rad Gel Doc EZ Imager (Bio-Rad, VIC, AUS).

Molecular typing
The molecular relationship and genetic diversity of the isolates were determined by ERIC-PCR using the primers ERIC-F (5ʹ-ATGTAAGCTCCTGGGGATTCAC-3ʹ) and ERIC-R (5ʹ-AAGTAAGTGACTGGGGTGAGCG-3ʹ). ERIC-PCR reactions were prepared in a volume of 25 μl, including 1 μl of each primer (final concentration 2 pmol/μl), 12.5 μl of Master Mix (Applied Biosystem), 9.5 μl of deionized water, and 4 μl of pattern DNA. At the beginning of the ERIC-PCR reaction to denature the template DNA strands, 95°C was applied for 5 min, and then, the target DNA sequences were proliferated during 30 cycles including denaturation at 92°C for 30 s; annealing at 52°C for 1 min; extension at 65°C for 8 min; Also, a final extension step at 65°C for 8 min and final storage at 4°C. The ERIC-PCR reaction products were evaluated on 1.5% agarose gel [32,33]. Comparison of ERIC-PCR patterns was performed. The presence of a band with the number 1 and the absence of a band with the number 0 were encoded in a matrix. Analysis of ERIC-PCR typing results was done according to the number and weight of bands observed in the product electrophoresis of each sample. Finally, the amplification of the products, which had been run, was done with the UPGMA algorithm on these websites: /http://insilico.ehu.es/dice_ upgma, http://genomes.urv.cat/UPGMA/index.php [34,35].

Bacterial strains and antibiotic susceptibility profile
Based on the primary culture results of the samples, 84 specimens were identified as K. pneumoniae. The results of antimicrobial susceptibility tests indicated that the highest levels of resistance were against ceftriaxone (65%), cefotaxime (64%) and ceftazidime (58%). The lowest levels of resistance were observed for meropenem (23%) and imipenem (28%). The complete results of the susceptibility test are shown in Table 2 below. This also indicates that more than 50% of the strains were multidrug resistance (MDR).

Phenotypic detection of ESBL and carbapenemase producing K. pneumoniae
The results of the combined disk test indicated that 27 out of 52 third-generation cephalosporin resistant isolates, which is about 52%, were ESBL positive. On the other hand, the results of the MHT test revealed that 15 out of 24 (which is about 62,5%) of the samples resistant to imipenem and meropenem K. pneumoniae isolates were carbapenems-positive with the ability to produce carbapenemase enzymes.

Virulence factors and antibiotic resistanceassociated genes
The results showed that from a total of 84 samples, 81 isolates (96%), 79 isolates (94%) and 77 isolates (91%) had bla CTX-M , bla SHV and bla TEM genes. Most of these genes were observed in all resistance isolates. Additionally, our results showed that 60 isolates (71%) and 50 strains (60%) had bla KPC and bla NDM genes. Among the 36 common strains, which were resistant to both third-generation cephalosporins, imipenem and meropenem, the bla KPC gene was positive in 34 cases, while the bla NDM gene was also detected in 33 of the 36 common resistant samples. Our data indicated that bla KPC and bla NDM genes had almost the same incidence among resistant K. pneumoniae, and there was no significant difference in their frequency (Figures 1, 2).
Examination of the genes involved in aminoglycoside resistance resulted in the detection of the Aac6-Ib gene in 76 isolates (90%). It was observed that 34 samples out of 36 resistant isolates were positive. Furthermore, the armA gene was identified in 60 samples (71%), which were positive among 36 common resistant strains. We found out that the Aac6-Ib and armA genes had a higher frequency among resistant specimens, while Aac6-Ib was meaningfully ample than the armA gene (Table 3, Figure 1).
Findings related to virulence genes showed that 50 isolates (60%) were positive for the Irp-1 gene, which was 34 out of 36 resistant samples. 70 samples (83%) were positive for the rmpA gene, in this way, 32 positive isolates were found out of 36 resistant samples. Moreover, in terms of the magA gene, 10 samples (11%) were positive with 6 cases related to the resistant strains. The mrkD gene was also positive in 79 samples (94%) so that out of 36 resistant isolates, 31 samples were positive. Our findings demonstrated that mrkD, rmpA, Irp-1 and magA genes had the highest frequency among resistant strains, where mrkD was significantly more abundant compared to the other tested genes (Table 3, Figures 1, 2).

Genotyping of K. pneumoniae isolates by ERIC-PCR analyses
ERIC-PCR differentiated the isolates into four clusters (G1-G4) with 70% similarity (Figure 3). In the strains that were studied, the maximum number (40) belonged to the G4 cluster, and the minimum number (8) belonged to the G3 cluster. This also showed that ten of the studied strains were in G2 and 26 strains were in the G1 cluster (Figures 3, 4).

Discussion
The present study revealed the picture of antibiotic resistance, virulence and genetic relationship among the clinical strains of K. pneumonia, which were recovered from the patients admitted to the Bahman Hospital and other laboratories in Shahr-e-Qods.
K. pneumoniae causes several types of infections in humans, including respiratory, bloodstream and urinary tract infections (UTIs), which are commonly seen in hospitalized or immunocompromised patients [36,37]. These infections are often treated with beta-lactams and other effective antibiotics against Enterobacteriaceae. Nevertheless, the antibiotic-resistant and highly pathogenic species of K. pneumoniae are rapidly spreading around the world [2]. Bacterial resistance depends vastly on populational and geographical factors. Thus, the body of different people can provide different environments for the bacteria to grow, multiply and be affected by the drugs . Therefore, the study of bacterial resistance in a specific population can give an appropriate overview of effective drugs for the healthcare staff to provide an effective antibiotic regimen to ensure improved recovery of patients.
Our findings demonstrated that out of 84 isolates identified as K. pneumoniae, more than 50% of the samples had multidrug resistance, with the highest resistance, respectively, against ceftriaxone, cefotaxime and ceftazidime. Notably, in the tested isolates, a high susceptibility to meropenem, imipenem and tetracycline was observed. In line with the present study, Rocha et al., reported that K. pneumoniae ESBL-positive isolates sampled from ICU patients showed the highest resistance to ceftriaxone [38]. Kim D et al. found that during the period from 2013 to 2015 the resistance rates of Klebsiella pneumoniae to cefotaxime, cefepime, and carbapenem were 38-41%, 33-41%, and <0.1-2%, respectively [39]. According to them, all isolates are sensitive to imipenem. In a different report compared to our results, the highest sensitivity was observed against ceftriaxone, ciprofloxacin and gentamicin, respectively. However, in the present study, the highest resistance was to ceftriaxone [40].
Moreover, our results of the combined disk test indicated that 27 out of 52 resistant samples (52%) cases were positive for ESBL, almost corroborating evidence with the findings of a study by Rupinder et al., who stated that ESBL production was observed in 48% of E. coli, 44% of K. pneumoniae and 50% of P. aeruginosa isolates in a tertiary hospital in Patiala, Punjab [41].
The results of the genes involved in ESBL resistance showed that the prevalence of bla CTX-M , bla SHV and bla TEM genes from a total of 84 samples were 81(96%), 79(94%) and 77(91%), respectively. Therefore, the highest frequency was related to the bla CTX-M gene. In the study of Pishtiwan et al., the frequencies of bla TEM , bla SHV and bla CTX-M genes were 64.7%, 35.2% and 41.1%, respectively [42]. Also, according to Ugbo et al., the prevalence of bla SHV , bla TEM and bla CTX-M was identified to be 55%, 35% and 45% [43]. It is worth mentioning that the genotypic approach can be the method of choice for distinguishing ESBL-producing strains from Enterobacteriaceae because phenotypic tests for ESBL detection only confirm ESBL production.
Analysis of the detection of carbapenemase-producing resistance genes showed that bla KPC (71%) was ampler than bla NDM (60%) without any significant differences in resistant species. Liu et al. stated that among the tested isolates of K. pneumoniae, bla NDM and bla KPC genes had the highest frequency [44]. This discrepancy may be attributed to genetic variations among strains associated with human populations and antibiotic regimens. In accordance with our data, Xiufeng et al. reported that bla KPC and bla NDM were highly detected in carbapenems-resistant K. pneumoniae samples isolated from a Chinese hospital [45].
In the case of genes involved in aminoglycoside resistance, the Aac6-Ib and armA genes had the frequency of 90% and 71%, while among the resistant samples, Aac6-Ib was the most frequently detected gene. Aligned with our results, Cirit et al. also reported Aac6-Ib to have the highest frequency among the genes involved in aminoglycoside resistance in nosocomial K. pneumoniae isolates [46]. In another study conducted by Li et al. in China, from a total of 223 isolates of K. pneumoniae, 13 isolates (5.8%) contained armA and 8 isolates (3.6%) contained rmtB. One hundred and ten isolates of K. pneumoniae were phenotypically resistant and after PCR, 11.8% of the isolates contained armA gene and 7.3% of the strains contained rmtB gene [47]. These values are lower than our findings about the armA gene.
According to the results of the modified Hodge testing of 24 samples resistant to imipenem and meropenem, 15 (62.5%) samples were positive for carbapenemases. Examination of virulence genes showed that among 84 samples, mrkD, rmpA, Irp-1 and magA genes showed the frequencies of 94%, 83%, 60% and 11%, respectively. This also showed that among 36 resistant samples, mrkD, rmpA, Irp-1 and magA genes had the highest frequency, while mrkD was significantly more abundant compared to the other genes. Based on these data, mrkD is evidently highly associated with multidrug resistance. In line with our findings, Liu et al. identified mrkD as the most common virulence gene with a prevalence of 100% [48]. In 2004, an initial study of the magA gene was performed by Fang et al., who identified the gene as a virulence factor in the pathogenesis of K. pneumoniae. In this study, the magA gene was observed in 52 invasive strains (liver abscesses) and 15 noninvasive strains [49]. El Fertas-Aissani et al. [50] obtained the opposite result from the present report. In this regard, none of the studied strains carried the magA gene. They also examined the rmpA gene, which contained 3.7% of the 54 strains of K. pneumoniae isolated from different clinical specimens, which is a lower percentage than the present study. In another study, Liu et al. reported that rmpA and magA were the most abundant genes among the 117 isolates of K. pneumoniae, respectively [44].
Since the presence of some virulence factors can be involved in the pathogenicity of bacteria, knowledge of the existence of these factors and their prevalence can be  a good way to identify and treat the studied strain. The phylogenetic tree was drawn based on the UPGMA algorithm, and the genetic relationship between the isolates was identified. The isolates were differentiated into four clusters, G1-G4, with 70% similarity. Moreover, in the studied strains, the maximum number (40) belonged to the G4 cluster, and the minimum (8) belonged to the G3 cluster.
According to the research of Ferreira et al. [51], the dendrogram obtained from ERIC-PCR results showed a genetic relationship between 25 studied K. pneumoniae. In their study, the clusters were determined using the method (UPGMA) and dice's similarity coefficient. Based on this, it was found that although the bacteria were isolated from different patients, K. pneumoniae in the bloodstream had a high genetic relationship with each other. El-Badawy et al. [52] showed that most isolates have different origins by genotyping K. pneumoniae isolates using ERIC-PCR method. It was shown that 32 isolates belonged to 18 different single roots, indicating that the prevalence of K. pneumoniae in different parts of the hospital was due to poor infection control.

Conclusion
The present study revealed a high prevalence of resistance strains of K. pneumoniae. The highest resistance was observed for ceftriaxone and cefotaxime, and the lowest resistance was reported against meropenem and imipenem. Furthermore, the bla CTX, bla KPC , Aac6-Ib and mrkD seemed to be the highest associated genes with multidrug resistance. The high prevalence of antibiotic resistance and virulence genes in conjunction with the significant relationship between the strains reveals a high pathogenic capacity of the isolated pathotypes of K. pneumoniae.
Our findings demonstrated a direct relationship between the frequency of the genes involved in the development of virulence and resistance. It can also provide a highly effective model for physicians of relevant medical centers to prescribe more suitable antibiotic regimens aimed at improved clinical efficiency and faster recovery of patients. These findings emphasize the choice of more effective    antibiotic regimens for the treatment of infections caused by K. pneumonia.