Background

Klebsiella pneumoniae is a Gram-negative bacterium (GNB) and a significant pathogen in nosocomial infections, particularly in intensive care units (ICUs). It is responsible for severe infections such as urinary tract infections (UTIs), pneumonia, bacteremia, neonatal meningitis, and pyogenic liver abscesses. Over recent years, the emergence of multidrug-resistant (MDR) and extensively drug-resistant (XDR) K. pneumoniae, along with the absence of new antibiotics capable of combating them, has become a serious global issue [1].

Over the past several years, carbapenem antibiotics have served as effective last-line treatments for infections caused by MDR Enterobacterales. However, the recent emergence of carbapenemase-producing Enterobacterales, particularly K. pneumoniae, has been associated with higher mortality rates (up to 40–50%), particularly in bloodstream infections (BSIs) and ICU admissions. This has led to the consideration of colistin as one of the last and most effective options for treating carbapenem-resistant K. pneumoniae (CRKP) infections. Nevertheless, the observed increase in resistance to this antibiotic indicates a worrisome trend that undermines the efficacy of this once highly efficient treatment option [2].

In 2015, the discovery of the mobilized colistin resistance (mcr) gene marked a significant development, as it was found to confer unique colistin resistance (CLR) in Enterobacterales isolated. Subsequent studies have identified Enterobacterales carrying mcr genes worldwide, spanning across livestock, food, and humans populations, suggesting the potential for horizontal transmission of colistin resistance. This has raised concerns about the emergence of pandrug resistance in Enterobacterales. Therefore, it remains crucial to continuously and precisely monitor the emergence and spread of mcr genes among bacteria. Variants ranging from mcr-1 to mcr-10 have been documented to date [10].

The objective of this study was to investigate the molecular mechanisms underlyingof colistin and carbapenem resistance in a collection of XDR CRKP isolated obtained from clinical specimens in Tehran, Iran. Additionally, the study aimed to describe the clonal relationships among these isolates.

Materials and methods

Bacterial isolates

Between August 2020 and February 2021, Milad Hospital in Tehran isolated a total of 80 non-duplicate strains of CRKP from clinical samples of both inpatients and outpatients. These strains exhibited resistance to either meropenem or imipenem during initial screening. Milad Hospital is a tertiary care facility with 1,000 beds, affiliated with the Social Assurance Organization. All isolates were obtained from clinical samples, including urine, blood, sputum, and tracheal aspirate. The bacterial isolates were reidentified as K. pneumoniae using biochemical methods including oxidase, sugar fermentation, IMViC, Kliger’s iron agar, nitrate reduction, and motility tests [11].

Assessment of antimicrobial susceptibility using the disk diffusion method

The susceptibility of CRKP isolates to 11 antibiotics specified by CLSI M100-Ed31, including ceftriaxone (30 µg), tobramycin (10 µg), piperacillin-tazobactam (10 µg), amikacin (30 µg), levofloxacin (5 µg), ceftazidime (30 µg), ciprofloxacin (5 µg), gentamicin (10 µg), meropenem (10 µg), imipenem (10 µg), and cefepime (30 µg) (MAST DISCS™ ID, UK), was determined using the standard disk diffusion method [12]. The results were interpreted according to the recommended criteria, with standard strains Escherichia coli ATCC 25,922 and Pseudomonas aeruginosa ATCC 27,853, were used as quality control strains for susceptibility testing. In accordance with the guidelines of the Centers for Disease Control and Prevention in the United States and the European Centre for Disease Prevention and Control, All isolates were identified as XDR These K. pneumoniae isolates demonstrated resistance to at least one agent in all antimicrobial categories, with the exception of two or fewer, indicating susceptibility to only one or two categories [13].

Determination of Minimum Inhibitory Concentration (MIC) against Colistin

MIC against colistin was determined using colistin (10 µg) discs (Neo-Sensitabs™, Rosco, Denmark) in the assay. The colistin broth disk elution method described in CLSI guidelines was used for the antimicrobial susceptibility test [12].

Detection of mcr-1 to mcr-10 genes by PCR

Genomic DNA was extracted using the Genomic DNA Purification Kit (QIAGEN® Kit, QIAGEN, Germantown, MD, USA) following the manufacturer’s instructions. The presence of isolates carrying mcr genes was determined through PCR amplification and subsequently confirmed by sequencing. DNA samples from E. coli SHP45 and E. coli KP37, known to carry the mcr-1 and mcr-2 genes, respectively, were utilized as positive controls in the assay. Additionally, genomic DNA from colistin-susceptible E. coli ATCC 25,922 served as the negative control. These strains were sourced from the Iranian Reference Health Laboratory. The primers for mcr genes are listed in Table 1 [14,15,16,17,18].

Table 1 The list of primers, annealing temperatures, and expected amplicon sizes for molecular detection of mcr genes and carbapenemases-producing K. pneumoniae isolates
Full size table

Detection of carbapenemase-encoding genes

Multiplex PCR was employed to detect blaNDM, blaIMP, blaVIM, blaKPC, and blaOXA-48. The positive and negative controls for PCR experiments were K. pneumoniae ATCC strain BAA-1705 and K. pneumoniae ATCC BAA-1706, respectively. PCR experiments used the specific oligonucleotide primers listed in Table 1 [19].

Genoty** with PFGE

All isolates were typed using a PFGE technique following the PulseNet Standardized Laboratory Protocol [13]. The genomic DNA from Salmonella enterica serotype Braenderup H9812 digested with XbaI (Thermo Fisher Scientific, USA) served as a molecular size marker. DNA banding patterns were analyzed using BioNumerics software, version 6.6 (Applied-Maths, Sint-Martens-Latem, Belgium). The analysis employed the Dice correlation coefficient and the UPGMA (unweighted pair group method using an arithmetic mean algorithm) method with a band tolerance and optimization set at 1.5%.By comparing the PFGE results and applying the criteria of Tenover et al. based on the number of observed band differences, a cutoff value of 80% similarity overall was set for related isolates [20].

Results

Sample Collection

Out of the 80 CRKP isolates collected from various wards of Milad Hospital in Tehran, Iran, the majority of isolates were from urine (54, 67.5%), followed by blood samples (14, 17.5%), and both sputum and tracheal aspirate each accounted for 6 isolates (7.5%). The prevalence of CRKP in different hospital wards is depicted in Fig. 1. The ICU ward had the highest rate, with 25 isolates (31.2%), while surgery and emergency wards each had three isolates (3%), showing the lowest rates, respectively. Additionally, 61.2% (n = 49) of the isolates were from female patients, while 38.8% (n = 31) were from male patients. The mean age of patients with these isolates was 49.61 ± 3 (between 1 and 93 years). Among the 80 patients, the majority, 71, were inpatients, with only 9 being outpatients.

Fig. 1
figure 1

The clustering results of the 80 carbapenem-resistant K. pneumoniae (CRKP) isolates, determined by PFGE patterns following digestion with the XbaI enzyme, were correlated with the presence of mcr, Carbapenemase genes and antibiotic resistance profiles. The information of strain is listed to the right of the patterns. The four PFGE cluster (A), (B), (C), and (D) are represented by rectangles. Full-length gels are presented in Supplementary Figs. 1, 2, 3, 4 and 5

Full size image

Antimicrobial susceptibility pattern of K. pneumonia

K. pneumoniae isolates exhibited a high resistance rate to ceftriaxone, ciprofloxacin (98.7%). The lowest resistance rates of all these isolates were observed with amikacin (47.5%). Resistance to other antibiotics was observed above 90%, as shown in Fig. 2. Among the 80 isolates, 33.7% (n = 27) were identified as CLR by colistin disk elution method.

Fig. 2
figure 2

Antimicrobial resistance pattern of 80 K. pneumoniae. TM: Tobramycin, CRO: Ceftriaxone, TZP: Piperacillin/tazobactam, AK: Amikacin, LVX: Levofloxacin, CAZ: Ceftazidime, CIP: Ciprofloxacin, GM: Gentamicin, MEM: Meropenem, IPM: Imipenem, FEP: Cefepime, CS: Colistin

Full size image

Detection of mcr genes

In this experiment, we used multiplex PCR screening to determine the prevalence of the mcr-1 to mcr-10 genes among the clinical K. pneumoniae isolates (Fig. 3). The mcr-1 gene was detected in 3.7% (3 out of 80) of the obtained isolates. None of the studied isolates were found to carry the mcr-2 to mcr-10 genes. Table 2 indicates that there were no significant correlations between the CLR isolates and the presence mcr-1 genes.

Fig. 3
figure 3

The electrophoresis analysis using (2%) agarose gel for the detection of mcr − 1 genes, and the M: DNA ladder 100 bp; lanes (1) is a negative control, lanes (2) is a positive control, lanes (3): a K. pneumoniae clinical isolate

Full size image
Table 2 The frequency of the mcr-1 gene among phenotypically colistin-resistant isolates (N = 27)
Full size table

Molecular analysis of carbapenemase genes

The isolates were examined by multiplex PCR for blaOXA-48, blaVIM, blaKPC, blaIMP, and blaNDM, and confirmed by sequencing (Fig. 4). The frequency of carbapenemase genes is displayed in Table 3. The gene encoding the OXA-48 enzyme was the most prevalent among the studied isolates and was identified in 36 strains (45%). It was followed by blaVIM, blaIMP, blaKPC, and blaNDM, in 14 (17.5%), 12 (15%), 10 (12.5%), and 4 (5%) strains, respectively. Additionally, co-existence of blaOXA-48 and blaKPC, and blaOXA-48, blaKPC, and blaVIM genes were observed in 3 strains (3%) and 1 strain (1.2%), respectively. Furthermore, the results indicated no significant correlations between the CLR isolates and the detected carbapenemase genes.

Fig. 4
figure 4

PCR results for carbapenemase-encoding genes; M: DNA ladder 100 bp, lanes (1) is a negative control, lanes (2) is a positive for blaKPC (798 bp), lanes (3): positive for blaVIM (390 bp), lanes (4): positive for blaOXA−48 (438 bp), lane (5): positive for blaIMP, lane (6): positive for blaNDM, lane (7): positive for blaIMP and blaNDM

Full size image
Table 3 Prevalence of carbapenemase genes among isolates (N = 80)
Full size table

Population of K. pneumoniae strains

A clonal analysis was performed on the 80 CRKP strains isolated from Milad hospitals. The PFGE dendrogram revealed four clusters based on an 80% similarity level, designated Clusters A to D, with the highest number of isolates belonging to Cluster A and the lowest number to Cluster C. (Fig. 1). A total of 40, 26, 4, and 10 CRKP isolates were identified in Clusters A, B, C, and D, respectively. Isolates from Clusters B, and D were obtained from different wards, while Cluster A isolates were primarily from the ICU (isolates no. 20, 32, 44, 5, 55, 67, 8, 61, 56, 68 ,57, 59, 17, 29, 41, 79 ,64, 63, 14, 76, 13, 16, 54, 66, and 52). Additionally, Cluster C consisted of clinical isolates from the internal ward under study. Furthermore, looking at pulsotypes, it was evident that each pulsotype had a similar antibiotic sensitivity pattern and carbapenemase genes. The most frequently detected carbapenemase gene was blaOXA-48. PFGE profiles demonstrated that the mcr-1-harboring K. pneumoniae was found in pulsotype P10 of Cluster A in the ICU and CCU wards. All of these isolates exhibited identical PFGE patterns and a 100% resistance profile to all antibiotics in our study except for tobramycin and gentamicin, and harbored the blaOXA-48 and blaKPC genes.

Discussion

In recent decades, the escalating prevalence of antibiotic-resistant GNB, particularly Klebsiella spp., has emerged as a significant global health threat, particularly within ICUs. CRKP stands out as the most frequently implicated microorganism causing nosocomial infections. The global increase in multidrug-resistant K. pneumoniae strains has led to increase the use of colistin to treat these infections, resulting in the emergence of colistin resistance worldwide [21, 22]. An important consideration in the management of nosocomial infections caused by K. pneumoniae are periodic surveillance to identify the resistant strains, optimizing available infection control policies, and treatment options in different areas of hospitals [23].

The objective of this study was to investigate the molecular mechanisms underlying colistin and carbapenem resistance among a collection ofXDR CRKP isolated from clinical specimens in Tehran, Iran. Additionally, the study aimed to describe the clonal relationships among these isolates. The utilization of molecular methods, particularly PFGE, has proven invaluable in comprehending the epidemiological aspects of such infections and identifying their sources. In our study, PFGE served as the molecular ty** method, revealing a high genomic relatedness among CRKP isolates. Epidemiological investigations such as PFGE are essential for identifying bacterial isolate outbreaks and transmission among patients, as well as within hospital wards. Additionally, PFGE plays a crucial role in obtaining important information on resistance transmission through the dissemination of clonal complexes worldwide [24].

The reported colistin-resistant rate in Iran is approximately 11.6% [25]. However, data from neighboring countries indicate that colistin resistance ranges from 0 to 31.7 [26]. These discrepancies between reports could stem from variations in the methods used to study resistance, the availability of colistin in healthcare settings, inadequate infection control programs, and increased utilization of colistin in clinical settings. Consistent with expectations for colistin-resistant isolates in our study, the majority also exhibited resistance to other clinically relevant antimicrobial agents. [27, 28]( Fig. 1).

Isolates carrying mcr-1 genes exhibited resistance to colistin by the colistin broth disk elution method (MIC ≤ 4 mg/L), and remarkably, all of these isolates displayed identical PFGE patterns, indicating their origin from a single clone. Remarkably, findings from the current study indicate that mcr-1- negative K. pneumoniae isolates displayed substantial colistin resistance. This observation aligns with previous studies which have shown that K. pneumoniae strains with chromosomal mutations in the mgrB gene also exhibit elevated levels of colistin resistance [29, 30]. Critical alterations in mgrB, such as disruptions in the promoter or coding sequence, are believed to result in the silencing of the gene or the generation of truncated forms of mgrB. Consequently, the inactivation of mgrB by any of these occurrences leads to the activation of the PhoP/PhoQ system, which subsequently activates the PmrA response regulator. This activation of PmrA is responsible for modifying the lipopolysaccharide, which is the target of polymyxins [31]. Previous investigations have indicated that the prevalence of mcr-1 in Enterobacterales ranges between 0.1 and 1% [32, 33]. In Iran, the widespread utilization of colistin in clinical practice, primarily due to the dissemination of carbapenemase-producing Enterobacterales, has led to the selection of multidrug-resistant bacteria in hospital settings [34]. Although these findings suggest a low prevalence of the mcr-1 gene among CRKP isolates, regular surveillance efforts are crucial to continually assess the epidemiology of mcr-1 among CRKP strains.

In this study, out of a total of 54 urine samples collected from hospitalized patients, 38 (70.4%) were from female patients and 16 (29.6%) were from male patients. Therefore, the results of the present research, like many previous studies, show that women are more susceptible to urinary tract infections with K. pneumoniae than men [35].

In our study, PFGE analysis revealed the presence of four clusters of related strains and 29 pulsotype strains. These findings indicate low diversity, suggesting a clonal population structure characterized by continuous exchange of K. pneumoniae strains among patients within the same and different hospital wards. This pattern aligns with previous epidemiological studies conducted in Iran, which revealed frequent transmission of K. pneumoniae strains among patients within medical centers. Additionally, our results are consistent with other epidemiological studies demonstrating a polyclonal population structure of K. pneumonia [36]. Within our study, we observed multiple clones simultaneously circulating and persisting, contributing to the endemic presence of K. pneumoniae within our hospital, despite the implementation of infection control measures such as hand hygiene, colonization surveillance among high-risk patients, and contact precautions.

In this study, based on the results of PFGE analysis, all CRKP isolates within the largest cluster (Cluster A) carring the blaOXA−48 gene. The OXA-48 gene, a class D carbapenemase, is situated within a composite transposon known as Tn1999. This gene is bordered by the carbapenemase gene and facilitates the mobilization of an intervening DNA segment. Studies have shown that blaOXA−48-carrying plasmids enable both clonal and horizontal transfer, thereby facilitating transmission between patients and healthcar workers. The presence of Cluster A suggests continuous exchange of K. pneumoniae strains not only within single hospital wards but also between different hospital wards. This emphasizes the role of widespread dissemination within a hospital setting [37].

In Cluster A, one isolate (K58) from the internal ward exhibited a band pattern resembling those from the ICU, suggesting a potential transfer of agents between the ICU and internal ward. Remarkably, our study is reported the Co-existence of mcr-1, blaOXA-48 and blaKPC genes in Cluster A. A matter of concern as such plasmids possess a significant risk of inter- and intra- wards dissemination in the hospital. Therefore, strict epidemiological surveillance, infection control measures, and antibiotic stewardship are required to curb this menace of colistin resistance from dissemination.

Cluster B, comprising three urine isolates (K25, 27and 42) from internal ward, along with one urine isolate (K39), one sputum isolate (K37), and one trachea aspirit isolates (K43) from emergency ward displayed a similar resistance and carbapenemase gene pattern (blaIMP). These findings strongly indicate the likelihood of interhospital transfer among patients within the internal and emergency wards.

Notably, isolates within Cluster C demonstrated an identical antimicrobial susceptibility profile and harbored the carbapenemase gene. Cluster C isolates shared an identical antimicrobial susceptibility profile and carried the blaNDM gene. This implies that these isolates were probably introduced to the ward through patients, clients, or medical staff. The genetic persistence within this cluster likely facilitated bacterial survival, colonization, and spread.

Three isolates in Cluster D were from Pediatric wards. Two isolates were associated with outpatient cases exhibited similar resistance patterns, carried carbapenemase genes (blaKPC), and shared identical genetic patterns. An important observation within this cluster was that isolates from both outpatient and inpatient wards showed comparable band patterns.This suggests the widespread transmission of strains across various hospital wards, potentially facilitated by outpatients and employees working outside the hospital. Therefore, if insufficient attention is given to controlling these strains, there is a risk of encountering a high rate of potential epidemics in the future. This serves as a serious warning for physicians and the infection control team.

Given the importance of investigating the molecular epidemiology of K. pneumoniae, numerous studies have been conducted worldwide. For instance, studies conducted in India on the carbapenemase-positive K. pneumoniae isolates [38] and in Iran on ESBL K. pneumoniae isolates revealed five and four clusters, respectively [39].

In contrast, an Iranian study on carbapenemase-positive K. pneumoniae isolates collected from various wards of a reference hospital. Their PFGE analysis revealed 11 clusters [20]. However, compared to our recent study, a significant disparity in genomic patterns observed may be attributed to the wide distribution of samples and the diverse origins of the strains.

In another study conducted in Iran, an analysis of 165 K. pneumoniae strains isolated from diverse samples revealed 17 clusters through PFGE analysis, with an 80% similarity rate [36]. In this study, the genetic diversity among isolates was high; one reason for this could be the diversity of sample sources and because our samples were from diverse sources.

Conclusion

Our findings indicate that molecular methods, such as PCR, offer a rapid and sensitive approach for detecting genes associated with antibiotic resistance, including mcr and carbapenemase genes. Additionally, using methods such as PFGE to analyze the clonality of resistant pathogens and investigate outbreaks of healthcare-associated infections can aid in identifying possible routes of dissemination and persistence of resistance among hospitalized patients.

Surveillance of carbapenem and colistin resistance prevalence in Iran is imperative. Furthermore, new therapeutic strategies, including the re-evaluation and utilization of older drugs, should be assessed and implemented in the country.