Identification of metallo-β-lactamases and AmpC production among Escherichia coli strains isolated from hemodialysis patients with urinary tract infection.

BACKGROUND: This study aimed to identify metallo-β-lactamases (MBLs) and AmpC β-lactamases-producing Escherichia coli isolates obtained from hemodialysis (HD) patients with urinary tract infections (UTI). METHODS AND
RESULTS: A total of 257 HD patients with UTI were included in this study, from which 47 E. coli isolates were collected. Antibiotic susceptibility was tested by disc diffusion method. MBLs and AmpC production were phenotypically detected by imipenem-ethylenediaminetetracetate and cefoxitin/boronic acid assays, respectively. The presence of MBLs and AmpC genes was examined by polymerase chain reaction (PCR). Fosfomycin and ampicillin were the most and the least effective antibiotics against E. coli isolates, respectively. Moreover, 61.7% (29/47) of E. coli isolates were multidrug-resistant with seven different antibiotypes. Antibiotype V (AMP-CIP-IMP-MEM-CPD-CRO-CTX-GEN-LEV-SXT-TOB) was the most prevalent profile. Besides, 24 (51.1%) isolates were simultaneously resistant to imipenem and meropenem. Phenotypic assay showed MBL production in 16 (66.7%) of the 24 carbapenem-resistant E. coli isolates. The distribution of MBL genes in carbapenem-resistant E. coli was as follows: blaIMP 18 (72%), blaVIM 7 (28%), and blaNDM 1 (4%). AmpC was detected in 61.7% (29/47) of the isolates using the phenotypic method. The presence of AmpC genes was confirmed by PCR in only 26 of 29 (86.7%) AmpC producers. The frequencies of blaDHA-1, blaACC, and blaCMY-2 were 6 (20.7%), 11 (37.9%), and 21 (72.4%), respectively.
CONCLUSIONS: The emergence of MBL and AmpC coproducing E. coli isolates calls for an urgent surveillance program for timely diagnosis and screening of these genes in our healthcare systems.

Introduction

Dialysis is a procedure to remove excess fluids and waste products from the human body. There are two different methods of dialysis: peritoneal dialysis and hemodialysis (HD). The second method uses a machine located outside the body to remove blood and pump it into a dialyzer [1]. HD patients are at high risk for various infections, not only due to the invasiveness of this method, but also because of the immunosuppression caused by the inflammation and uremia [2]. In HD patients with chronic renal disease, infection is the leading cause of death. These patients are at risk of urinary tract infections (UTIs) which can lead to serious complications. Because of the problems in collecting urine samples from anuric or oliguric HD patients, UTIs are initially difficult to diagnose. Previous evidence suggests that multidrug-resistant bacterial pathogens of UTIs are more common in HD patients compared with those with normal renal function [3]. Despite extensive studies on UTIs in many countries, little is known about hemodialysis patients.

Infections such as UTIs, cause high economic costs and morbidity rates, making them one of the most significant diseases in hospitals and communities. In Asia, UTIs are the main causes of mortality and morbidity. According to diagnostic and clinical reports, approximately 150 million people become infected with UTI each year worldwide [4, 5]. The prevalence of UTIs is mainly influenced by a variety of factors including gender, age, urological instruments, impaired immunity, indwelling urinary catheters, and underlying diseases such as diabetes mellitus [6]. Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus, Enterococcus faecalis, and Proteus mirabilis are the commonest bacteria causing UTIs. Community-acquired UTI is most commonly caused by E. coli bacteria, which are abundant in the gastrointestinal microflora of humans [7, 8].

Treatment of infections has been complicated by the emergence of multidrug-resistant (MDR) strains of E. coli. Beta-lactam antibiotics are prescribed drugs for UTI treatment. These agents are a class of extended-spectrum antibiotics and include all antibiotics that have a β-lactam ring in their structure [9, 10]. The worldwide distribution of E. coli harboring metallo-β-lactamases (MBLs) and AmpC β-lactamases (AmpC) is a serious threat, and due to MBL production, carbapenem resistance is progressively spreading among clinical isolates of E. coli [11]. Varied types of MBL genes have been identified around the world. The blaNDM, blaVIM, and blaIMP are among the most common MBLs [12].

The increasing emergence of AmpC (pAmpC) β-lactamases mediated by plasmids is also of growing concern. Six families of pAmpC β-lactamases have been defined by Pérez-Pérez and Hanson as EBC, MOX, FOX, CIT, DHA, and ACC. Among E. coli strains, the CMY-2 type is the most commonly known pAmpC. In recent decades, the frequency of AmpC-producing E. coli strains, which are mostly MDR, has increased significantly. Therefore, the detection of AmpC-positive strains is of utmost importance for appropriate treatment [13].

Considering the importance of these issues and knowing that studies on the prevalence of β-lactamase genes in E. coli strains isolated from HD patients with UTIs from all over the world, and especially from Iran are very rare, this study aimed to identify the MBLs and AmpC β-lactamases-producing E. coli isolates in HD patients with UTIs.

Methods

Study design and specimen collection

This descriptive cross-sectional study was conducted from October 2019 to July 2020. Written informed consent was obtained from all patients who participated in this study before the start of the work. A total of 257 HD patients with clinical suspicion of UTI (fever, chills, burning sensation during urination, cloudy urine, frequent and low urine output, unpleasant odor of urine) admitted to Army hospitals in Tehran city, were included in this study. The Army hospitals are located in Tehran, the capital of the country of Iran. These hospitals are for general referral and have all specialized and advanced medical departments. All patients with suspected UTI symptoms had their urine samples collected by clean-catch midstream protocol. A urine specimen was deemed positive for UTI if it contains a single microorganism with a count of ≥ 105 CFU/ml. E. coli isolates were identified by phenotypic and biochemical tests including Gram staining, lactose fermentation on MacConkey agar, triple sugar iron agar reaction, indole production, motility test, and Simmons' citrate test [14]. All culture media used in this study, were prepared and purchased from Merck Co, Germany.

Antibiotic susceptibility testing (AST)

AST was conducted using disc diffusion method, according to the recommendations of Clinical and Laboratory Standards Institute (CLSI) [15]. The AST was performed with the following antibiotic discs: imipenem (IMP, 10 μg), meropenem (MEM, 10 μg), amikacin (AMK, 30 μg), gentamicin (GEN, 10 μg), tobramycin (TOB, 10 μg), ampicillin (AMP, 10 μg), cotrimoxazole (SXT, 25 μg), fosfomycin (FOS, 10 μg), piperacillin–tazobactam (PTZ, 100/10 μg), ceftriaxone (CRO, 10 μg), ceftazidime (CAZ, 30 μg), cefpodoxime (CPD, 10 μg), cefotaxime (CXT, 30 μg), nitrofurantoin (NF, 300 μg), ciprofloxacin (CIP, 10 μg), and levofloxacin (LEV, 10 μg) (Mast, UK). E. coli ATCC 25922 was used as a control strain for the susceptibility tests. Isolates resistant to at least three classes of antibiotics were considered as MDR isolates.

Phenotypic detection of MBLs

The presence of MBLs was screened by the combined imipenem-ethylenediaminetetracetate (IMP-EDTA) disc test. Briefly, isolates to be tested were inoculated on a Mueller Hinton Agar (Merck Co, Germany) plate using lawn culture method. One disc of IMP (10 μg) alone and another disc in combination with EDTA (750 μg/ml) were applied 20 mm apart and incubated for 18–24 h at 37ºC. Strains were confirmed as MBL producers once an increase of ≥7 mm in the inhibition zone of IMP-EDTA combination disc was observed compared to the IMP disc alone [16].

Phenotypic detection of AmpC

The cefoxitin (FOX) disc (30 μg)/boronic acid was used to screen AmpC-producing isolates as previously described [17]. Isolates with the inhibition zones of <18 mm in diameter against FOX discs were considered potentially positive for AmpC screening and were subjected to further testing, using FOX disc alone and in combination with boronic acid (Sigma, USA). The presence of a 5-mm or larger increase in the inhibition zone diameter of the FOX disc in combination with boronic acid compared to the FOX alone, was considered as positive AmpC production.

Molecular detection of MBLs and AmpC genes

The presence of AmpC genes (blaCMY-2, blaDHA-1, and blaACC) [18] and MBLs genes (blaIMP, blaVIM, and blaNDM) [19] was detected by polymerase chain reaction (PCR). Genomic DNA extraction was performed using the boiling method [20]. The PCR was performed in a final volume of 25 μl consisting of DNA template (50 ng), dNTPs (100 μM), Taq buffer (5×), Taq DNA polymerase (1 U; Cinnagen, Iran), and forward and reverse primers (25 pM each). The PCR mixtures were subjected to thermal cycling. PCR reactions included 30 amplification cycles in a Mastercycler (Eppendorf, Germany) under the following conditions: denaturation at 95 °C/5 min, annealing at 55 °C/30 s, and extension at 72 °C/45 s, with a final extension at 72 °C/6 min. Amplified products were visualized using electrophoresis on a 1% agarose gel stained with safe stain (Sinaclon, Iran), in a Tris-Borate-EDTA buffer (Promega, USA). Water was used as a negative control in the study, and the positive controls were K.pneumoniae ATCC 700603, P.aeruginosa ATCC 27853, P.aeruginosa ST 147, K.pneumoniae KP696465, E.coli KX 342010 and E.coli KX342011.

Statistical analysis

Data were entered and statistically analyzed using Microsoft Excel 2019 (Microsoft Corporation, USA) and the Statistical Package for the Social Sciences (SPSS) software version 22 (IBM SPSS Statistics, USA) [21]. The results were presented as descriptive statistics in the form of relative frequency.

Results

Bacterial isolates and antibiotic resistance

In this study, a total of 47 E. coli isolates were obtained from urine samples of HD patients, with an overall prevalence of 18.3% (47/257). The isolates were obtained from 21 (44.7%) males and 26 (55.3%) females (female/male ratio =1.23), respectively. The mean age of the patients was (31 ± 1) years, and ranged from 2 to 65 years. The E. coli isolates showed the highest susceptibility (87.2%) and esistance (100%) to FOS and AMP antibiotics, respectively. PTZ and NF were the second and third most effective antibiotics, respectively. More than 70% of the isolates were resistant against third-generation cephalosporin (TGC) antibiotics. As well, 24 (51.1%) isolates were simultaneously resistant to IMP and MEM. The susceptibility rate results for all antibiotics tested are shown in Table 1. The results of AST showed that 61.7% (29/47) of E. coli isolates were MDR with seven different antibiotypes (Table 2). Antibiotype V (AMP-CIP-IMP-MEM-CPD-CRO-CTX-GEN-LEV-SXT-TOB) was the most prevalent profile (31.0%).

Table 1

The results of antibiotic susceptibility testing

Antibiotic	Total Escherichia coli, N (%)
	Sensitive	Intermediate	Resistant
Gentamicin	12 (25.5)	8 (17.0)	27 (57.4)
Amikacin	28 (59.6)	10 (21.3)	9 (19.1)
Levofloxacin	27 (57.4)	2 (4.3)	18 (38.3)
Ciprofloxacin	18 (38.3)	–	29 (61.7)
Tobramycin	14 (29.8)	5 (10.6)	28 (59.6)
Fosfomycin	41 (87.2)	–	6 (12.8)
Cefpodoxime	7 (14.9)	4 (8.5)	36 (76.6)
Cefotaxime	9 (19.1)	4 (8.5)	34 (72.3)
Ceftriaxone	7(14.9)	–	40 (85.1)
Ceftazidime	4 (8.5)	7 (14.9)	36 (76.6)
Ampicillin	–	–	47 (100)
Cotrimoxazole	9 (19.1)	–	38 (80.9)
Meropenem	23 (48.9)	–	24 (51.1)
Imipenem	19 (40.4)	4 (8.5)	24 (51.1)
Piperacillin–tazobactam	38 (80.9)	–	9 (19.1)
Nitrofurantoin	36 (76.6)	–	11 (23.4)

Table 2

Different antibiotypes of 29 multidrug-resistant (MDR) E. coli isolates

MDR profile	Antibiotypes	Number of isolates (%)
I	CIP–CPD–CRO–CTX–SXT–TOB–NF	1 (3.4)
II	AMP–CAZ–CIP–PTZ– IMI–MEM–CPD–CRO–CTX–FOS–GEN–SXT–TOB–LEV	4 (13.8)
III	AMP–CAZ–CIP–PTZ–IMI–MEM–CPD–CRO–CTX–GEN–LEV–SXT–TOB	5 (17.2)
IV	AMP–AN–CIP–CPD–CRO–CTX–GEN–SXT–TOB–NF	3 (10.3)
V	AMP–CIP–IMI–MEM–CPD–CRO–CTX–GEN–LEV–SXT–TOB	9 (31.0)
VI	AMP–AN–CIP–IMI–MEM–CPD–CRO–CTX–GEN–SXT–TOB–NF	6 (20.7)
VII	CIP–CPD–CRO–CTX–SXT–NF	1 (3.4)

AMP Ampicillin, AN Amikacin, CAZ Ceftazidime, CIP Ciprofloxacin, CPD Cefpodoxime, CRO Ceftriaxone, CTX Cefotaxime, FOS Fosfomycin, GEN Gentamicin, IMI Imipenem, LEV Levofloxacin, MEM Meropenem, NF Nitrofurantoin, PTZ Piperacillin–tazobactam, SXT Cotrimoxazole, TOB Tobramycin

Detection of MBLs and AmpC genes

The results of the IMP-EDTA combined disc test confirmed the MBL production in 16 (66.7%) out of 24 carbapenem-resistant E. coli isolates. PCR results showed that all 24 carbapenem-resistant E. coli isolates had at least one MBL gene. The distribution of MBL genes among carbapenem-resistant E. coli isolates was as follows: blaIMP 18 (72%), blaVIM 7 (28%), and blaNDM 1 (4%). Two isolates showed coexistence of MBL genes. Results from FOX disc/boronic acid detected 61.7% (29/47) potential AmpC producers. However, the presence of AmpC genes was confirmed by PCR in only 26 of 29 (86.7%) AmpC producers. The frequency rates of the resistance genes blaDHA-1, blaACC, and blaCMY-2 among AmpC β-lactamase-producing strains were 6 (20.7%), 11 (37.9%), and 21 (72.4%), respectively. Ten isolates showed coexistence of AmpC genes. Overall, the frequencies of MBL and AmpC genes were 51.1% (24/47) and 55.3% (26/47), respectively. Fifteen (31.91%) E. coli isolates had no MBL and AmpC genes, and 18 (38.3%) isolates had both of these genes, simultaneously (Table 3). The electrophoresis image of some MBL and AmpC positive isolates is shown in Fig. 1.

Table 3

MBLs and AmpC profiles of all 47 Escherichia coli isolates

Strain ID	Gender	AmpC producer	MBL producer	AmpC genes	MBL genes
Ec 1	M	+	+	blaDHA-1, blaACC	blaIMP
Ec 2	F	+	−	blaCMY-2	blaIMP, blaVIM
Ec 3	M	−	+	−	blaIMP
Ec 4	M	+	+	blaCMY-2	blaVIM
Ec 5	F	−	−	−	−
Ec 6	F	+	−	blaDHA-1, blaACC, blaCMY-2	blaVIM
Ec 7	M	+	+	blaCMY-2	blaIMP
Ec 8	F	+	−	blaCMY-2	blaVIM
Ec 9	F	−	−	−	−
Ec 10	M	−	+	−	blaIMP
Ec 11	F	+	+	−	blaVIM
Ec 12	F	+	−	blaACC, blaCMY-2	−
Ec 13	F	+	−	blaCMY-2	blaIMP
Ec 14	M	−	+	−	blaIMP
Ec 15	M	−	−	−	−
Ec 16	F	+	+	blaACC, blaCMY-2	blaVIM
Ec 17	F	−	−	−	−
Ec18	M	+	+	blaCMY-2	blaIMP
Ec 19	F	−	−	−	−
Ec 20	M	+	+	blaDHA-1, blaACC	blaIMP
Ec 21	F	−	−	−	−
Ec 22	M	+	+	blaACC	blaVIM, blaNDM
Ec 23	M	+	−	blaCMY-2	−
Ec 24	F	+	−	blaCMY-2	blaIMP
Ec 25	M	−	−	−	−
Ec 26	M	+	−	blaACC	−
Ec 27	F	+	−	blaCMY-2	blaIMP
Ec 28	M	−	−	−	−
Ec 29	F	+	−	blaCMY-2	−
Ec 30	M	−	−	−	−
Ec 31	F	+	−	blaDHA-1, blaCMY-2	blaIMP
Ec 32	F	+	−	−	−
Ec 33	M	−	−	−	−
Ec 34	M	+	−	blaCMY-2	blaIMP
Ec 35	F	+	−	blaCMY-2	−
Ec 36	F	+	+	blaDHA-1, blaCMY-2	blaIMP
Ec 37	M	−	+	−	blaIMP
Ec 38	F	+	−	blaACC	−
Ec 39	F	+	−	blaACC, blaCMY-2	−
Ec 40	M	−	−	−	−
Ec 41	F	+	−	−	−
Ec 42	M	−	+	−	blaIMP
Ec 43	F	+	+	blaACC, blaCMY-2	blaIMP
Ec 44	F	−	−	−	−
Ec 45	F	+	−	blaCMY-2	−
Ec 46	M	−	−	−	−
Ec 47	F	+	+	blaDHA-1, blaACC, blaCMY-2	blaIMP

Ec Escherichia coli, F female, M male, MBL Metallo-β-lactamase, AmpC AmpC beta-lactamase

Fig. 1

Agarose gel electrophoresis (1.0%) of blaDHA PCR products. From left to right: L indicates DNA Ladder (100 bp), lanes 1 and 6 demonstrate blaDHA-positive (405 bp) Escherichia coli isolates, lanes 2,3,4,5,7,8,9,10, and 11 indicate blaDHA-negative E. coli isolates, NC indicates negative control, PC indicates positive control of blaDHA gene (405 bp) (a). Agarose gel electrophoresis (1.0%) of blaIMP PCR products. From left to right: L indicates DNA Ladder (100 bp), lanes 1, 2, 3, 7, and 10 demonstrate blaIMP-positive (568 bp) E. coli isolates, lane 4, 5, 6, 8, and 9 show blaIMP-negative E. coli isolates, NC indicates negative control, PC indicates positive control of blaIMP gene (568 bp) (b)

Discussion

At present, a paucity of epidemiological information is available on E. coli strains causing UTIs, as well as on their antibiotic resistance profiles and resistance mechanisms in HD patients from many countries including Iran. The novelty of this study was to address these issues in a region of Iran. The present study investigated the prevalence and antibiotype patterns of UTI-causing E. coli isolates in HD patients against 16 antibiotics.

The results showed an overall prevalence of 18.3% for UTI-causing E. coli isolates in HD patients. This rate was much lower than the rate (43%) reported by Sadeghi et al. [22] from Isfahan, Iran, but in kidney transplant patients. In another research in USA, a frequency rate of 29.3% was reported for E. coli isolates in HD patients [23]. The rationale behind these discrepancies is likely the differences in studied population, patient screening programs, and predominant pattern of UTI-inducing pathogens in each region.

According to the results of AST, MEM and IMP were among the relatively least effective antibiotics, with resistance rates of about 50%. These results were in contradiction with previous reports by Critchley et al. [24] from USA, Raeispour et al. [25] from Iran, and Duicu et al. [26] from Romania who showed the efficacy of more than 90% for these two antibiotics. The emergence of carbapenem-resistant Gram-negative bacteria in recent years has posed a new challenge to medical centers, both in Iran and other parts of the world [27, 28]. This phenomenon may arise from the high prevalence of different carbapenemase genes that can develop resistance to carbapenem antibiotics [28]. Although FOS was the most effective antibiotic, its resistance rate (12.8%) was higher than the rate reported in South Korea (6.7%) [29].

Of the three aminoglycosides studied here, AMK showed the highest efficacy compared to GEN and TOB, supporting previous studies from USA [24] and Gabon [30]. However, a survey by Kushwaha et al. [31] from Nepal showed an opposite result: GEN had higher efficacy against UTI-inducing E. coli isolates than AMK and TOB. In this study, more than half of the E. coli isolates were resistant to GEN and TOB, which was higher than the report of Naziri et al. [32] from Iran.

This higher resistance rate could be explained by the excessive parenteral administration of these antibiotics in our region. Our study explored a high resistance rate, ranging from 72.3 to 85.1%, against TGC antibiotics, which was in agreement with previous researches conducted by Raeispour et al. [25] and Lee et al. [33], but in contrast to findings achieved from USA [24] and Gabon [30]. The high resistance rates against TGCs is possibly rooted in increasing the spread of extended-spectrum β-lactamase genes in UTI pathogens in Iran and other countries during recent decades, which were not investigated in the current study [34, 35].

The resistance rates against CIP and LEV were 61.7% and 38.3%, respectively. Reports from Iran by Halaji et al. [34] and Raeispour et al. [25] stated the resistance rates of 54.3% and 34% for CIP, respectively. These rates were lower than those of the current study. The administration of CIP, as the first-line therapy, against UTI-inducing bacteria might be the reason for this higher resistance rate. Another finding obtained in the present work was the exceptional efficacy of NF with a susceptibility rate of 76.6%.

In line with our result, Lee et al. [33] reflected the high susceptibility rates (74.3%-100%) of NF against different UTI-inducing bacteria and stated that 99.2% of E. coli isolates were NF-susceptible. Similarly, Raeispour et al. [23] implied that 90% of UTI-inducing E. coli isolates were susceptible to this antibiotic, as well. In another study in Gabon, which was comparable with ours, Mouanga Ndzime et al. [30] identified NF as the most effective drug against UTI-inducing E. coli isolates.

Likewise, a high proportion (80.9%) of E. coli isolates were susceptible to PTZ, confirming the results observed by Critchley et al. [24] in the USA and Halaji et al. [34] in Iran. In this study, AMP and SXT (with resistance rates of 100% and 80.9%, respectively) were among the less effective antibiotics, a result consistent with those observed in the Gabon [30], USA [24], and Iran [34].

This result for AMP was not far-fetched because in recent years, many Gram-negative bacilli carrying β-lactamase enzymes have readily become resistant to a wide range of β-lactam antibiotics. As SXT has always been one of the first-line drugs for the treatment of UTIs, according to the observed evidence, it is recommended to prescribe this drug with regard to the results of antibiotic susceptibility tests.

One notable finding of this study was the relatively high prevalence of MDR (61.7%) E. coli isolates. Previous studies have reported various MDR frequencies, ranging from 55.1 to 100% for UTI-causing E. coli isolates [25, 30, 34]. In this study, 7 different antibiotypes were seen in MDR isolates. Previous study from Saudi Arabia showed 70 different antibiotypes in E. coli isolates collected from clinical samples [36]. In another study by Alanazi et al. [37], 7 different antibiotypes were seen in UTI-causing E. coli isolates against ciprofloxacin, ampicillin and co-trimoxazole antibiotics. In this study, according to the results of the antibiogram, the most effective antibiotic against these MDR isolates was the FOS that can be considered for treatment of the MDR UTI-causing E. coli. These results necessitate the development of a surveillance program for antibiotic consumption to control the spread of MDR E. coli isolates in healthcare systems of our country. In recent years, the outbreak of E. coli sequence type 131 (ST131) isolated from urine culture has become one of the global health problems due to the high prevalence of multidrug resistance. This clone has a wide range of virulence factors including siderophores, adhesins, and toxins which disrupt host defense mechanisms. The majority of these factors are found on mobile genetic elements (MGEs) or pathogenicity islands (PAI), which are capable of being horizontally transmitted among different species [38, 39]. MGEs have a huge impact on bacterial genomes, including causing marked differences in genome size and pathogenicity. Although bacteria have several mechanisms for resisting lateral gene transfer, MGEs play a major role in bacterial evolution and contribute greatly to adaptation to new and changing ecological niches. Most of the resistance genes are acquired by horizontal transfer of plasmids and other MGEs, and this process has been associated with the successful dissemination of particular lineages. Also, MGEs have an important role in virulence gene acquisition and forming new subpopulations among pandemic clones such as E. coli ST131 [38, 39].

This study investigated the production of MBL and AmpC β-lactamases in UTI-inducing E. coli isolates in HD patients. To the best of our knowledge, this was the first study in this regard in Iran. Today, there are multiple phenotypic methods for detecting β-lactamases, but owing to the multiplicity of different types of these enzymes, none of the phenotypic methods can identify all types. Therefore, their detection is mostly based on molecular methods such as PCR or real-time PCR [40, 41].

In our survey, AmpC and MBL phenotypes were detected in 29 (61.7%) and 16 (34%) UTI-causing E. coli isolates, respectively. Similar to these results, Helmy et al. [42] reported a prevalence of 66.6% for AmpC-producing E. coli isolates in Egyptian UTI patients using phenylboronic acid method. A higher prevalence rate (87.5%, n = 175/200) was also reported for the detection of phenotypic AmpC in UTI-inducing E. coli isolates collected from New Zealand by Drinkovic et al. [43]. In opposition to our findings, Jamil et al. [44] stated a lower frequency (16%, n = 12/75) of UTI-causing E. coli isolates harboring MBLs in Pakistan using phenotypic method. Variations in the results seem to arise from differences in the phenotypic test methods, quality of materials used, and dissimilarity in geographical area.

The results of PCR assay revealed a frequency of 20.7%, 37.9%, and 72.4% for blaDHA-1, blaACC, and blaCMY-2 AmpC β-lactamases-producing isolates, respectively. In conformity with our results, findings by Helmy et al. [42] from Egypt and Drinkovic et al. [43] from New Zealand represented the blaCMY-2 (89% [n = 90/101]) as the predominant type of AmpC in UTI-inducing E. coli isolates. The extensive dissemination of blaCMY-2 among E. coli isolates could be linked to unique transposon-like element ISEcp1, which is thought to play a role in the transmission of blaCMY-2 from the Citrobacter freundii chromosome to other Enterobacteriaceae [42]. In another study conducted in Egypt, the prevalence rate of 50% was reported for blaDHA AmpC in UTI-causing E. coli, while other AmpC genes were not detected [45]. In a study from Nepal, two AmpC genes blaCIT and blaDHA were detected in 30.6% and 31.3% of clinical E. coli isolates, respectively [46]. Another experiment by Shayan and Bokaeian [47], showed a prevalence of 5.0% and 0.0% for blaCMY-2 and blaFOX genes, respectively. They did not investigate the remaining AmpC genes. In a recent study from Iran, no blaMOX and blaECC genes were found in uropathogenic E. coli isolates, while, blaCIT, blaFOX, blaDHA, and blaEBC were detected in 73.6%, 10.5%, 10.5%, and 15.8% of E. coli isolates, respectively [48]. Currently, the clinical microbiology laboratories do not have a reliable method of detecting and confirming AmpC β-lactamases. For this reason, the molecular characterization of β-lactamases, in particular the AmpC family, is of crucial importance in in terms of molecular epidemiology and genetic contributing factors [48].

This study revealed the high prevalence rate of MBL genes in 24 carbapenem-resistant isolates, as all of them harbored at least one MBL gene. In the current study, the blaIMP (72%) was the most prevalent MBL gene followed by blaVIM (28%) and blaNDM (4%). The emergence of MBL-producing UTI-inducing E. coli isolates could be considered a serious threat to health communities because this pathogen is resistant to numerous antibiotics. In a study by Naeem et al. [49], the NDM-positive rate was 38.5% (n = 10/26) among E. coli strains isolated from urine samples that was higher than the current result. In another study from Sudan [50], the blaVIM (16.7%) was the most prevalent MBL gene in clinical E. coli isolates followed by blaIMP (8.3%) and blaNDM (2.8%).There is a paucity of data regarding the blaVIM and blaIMP harboring UTI-causing E. coli in Iran. In a recent study from southwest Iran, each of blaIMP-1 and blaIMP-2 genes were detected in 8.3% of uropathogeneic E. coli isolates, while the blaNDM was detected in 75.0% of isolates. Also, the blaVIM-1 and blaVIM-2 were not detected [51]. In another study by Deldar Abad Paskeh et al. [52] from Iran, a lower occurrence rates of blaIMP (8.7%) and blaVIM (9.8%) genes were reported in UTI-causing E. coli compared to the current study. In a recent study from Taiwan, an increasing rate of NDM-positive E. coli isolates has been reported in clinical samples. They reported a prevalence rates of 39.1%, 30.4%, 21.7%, and 8.7% for NDM, IMP-8, KPC-2, and VIM-1, respectively [53]. NDM is a member of the amber class B β-lactamases, which can hydrolyze virtually all β-lactams except monobactam [53]. In China and its neighboring countries, IncX3 plasmids are the most common types of plasmids carrying blaNDM in Enterobacteriaceae [54], while in a recent study from Iran, blaNDM–1 gene was located on both conjugative plasmids: IncFII ∼86-kb to ∼140-kb and IncA/C [55]. There are various classes of chromosomal and plasmidic MBLs including Sao Paulo metallo-β-lactamase (SPM), German imipenemase (GIM), and Adelaide imipenemase (AIM) that were not screened in this study and could pave the way for further epidemiological research in the future [56].

The present research showed that 38.3% of the isolates harbored both MBL and AmpC genes simultaneously. In this regard, co-occurrence of β-lactamase genes has been reported in several investigations [45, 57], which was in harmony with the results of this study. In recent years, the occurrence of different β-lactamases in various Gram-negative bacteria has been reported in various studies [58-61]. Due to the financial and traffic constraints caused by the COVID-19 virus pandemic, the current study has the following limitations: the lack of MBLs and AmpC genes sequencing, the lack of antibiogram results for some newer and reserved antibiotics including ceftazidime/avibactam, imipenem/relebactam, ceftolozane/tazobactam, colistin, and tigecycline, and lack of evaluation of other bacteria contributed to UTI in HD patients. Also, due to the non-shareable privacy of those patients referred to Army hospitals, we did not have any clinical data of individuals to evaluate the correlation of infections by MDR E. coli in HD patients with other important predisposing conditions to detect if any significant trends exist in the prevalence of such MDR strains among different susceptible groups.

Conclusion

This survey portrayed the high resistance rates of UTI-causing E. coli isolates harboring AmpC and MBLs in HD patients against AMP, SXT, and TGCs. In light of this information, it is recommended to choose treatments based on the antibiogram results, in order to hinder the further spread of MDR isolates. Moreover, the higher efficacy of FOS, PTZ and NF against UTI-inducing E. coli isolates, compared to other antibiotics, made them suitable options for empirical therapy. The emergence of MBL and AmpC co-producing E. coli isolates calls for an urgent surveillance program for timely diagnosis and screening of these genes in our healthcare systems. In future, such program can be of great help in preventing the spread of antibiotic resistance in our country, Iran. Another suggestion is to evaluate the other infections and their etiological pathogens such as septicemia in HD patients to prepare a suitable epidemiological data bank in this regard in Iran.