Genetic relatedness in extended-spectrum beta-lactamase-producing Escherichia coli from clinical isolates using enterobacterial repetitive intergenic consensus polymerase chain reaction

Introduction

One of the main approaches to modern medicine used to combat infections is the use of antibiotics. A defining moment in humanity’s history that revolutionized medicine and saved uncountable lives was the discovery of antibiotics. The “golden era” of antibiotics from the 1930s to 1960s gave rise to many new antibiotics.[1] Unfortunately, these “magic bullets” have been accompanied by the emergence of antibiotic-resistant pathogens.[2]

The production of enzymes that hydrolyze antibiotics is one of the ways bacteria circumvent the effect of some antibiotics.[3] An example of these enzymes is beta-lactamases. There are two general classification schemes for beta-lactamases: The Bush-Jacoby-Medeiros functional classification system and the Ambler molecular classification scheme.[4-6] Based on the Ambler molecular classification scheme, beta-lactamases are divided into four major classes (A–D), and the protein homology (amino acid similarity), and not phenotypic characteristics are observed to be the basis of this classification scheme. Beta-lactamases A, C, and D are the serine beta-lactamases in the Ambler classification, and the Class B enzymes are metallo-beta-lactamases. However, the Bush-Jacoby-Medeiros classification scheme used functional similarities (substrate and inhibitor profile) to classify beta-lactamases, with four major groups and various subgroups in this system. As a result of consideration of beta-lactamase inhibitors and beta-lactam substrates that are clinically relevant, Paterson and Bonomo noted that the classification of Bush-Jacoby-Medeiros scheme is of much more direct importance to the microbiologist or physician in a diagnostic laboratory.[7]

One of the major breakthrough in the battle against beta-lactamase-mediated bacterial resistance to antibiotics is the introduction of the third-generation cephalosporins into clinical practice in the early 1980s.[7,8] In addition to effective activity against most beta-lactamase-producing organisms, the third-generation cephalosporins had significant advantage of lessened nephrotoxic effects compared to polymyxins and aminoglycosides.[7] However, the new generation of antibiotics also gave rise to new strains of antibiotic-resistant bacteria.

Extended-spectrum beta-lactamases (ESBLs) are a collection of plasmid-encoded enzymes that confer resistance to the third-generation cephalosporins.[9] Bush noted that there is no unanimity on the exact definition of ESBLs.[10] A generally used description is that ESBLs are beta-lactamases that have the capacity of conferring bacterial resistance to the penicillins, first-, second-, and third-generation cephalosporins, and aztreonam (but not the cephamycins or carbapenems) by hydrolyzing these antibiotics. They are inhibited by beta-lactamase inhibitors such as clavulanic acid.[10] However, the worldwide exponential increase in beta-lactamase resistance which is attended by a substantial increase in the prevalence of ESBL-producing Enterobacteriaceae has been reported over the past two decades.[9]

The first step in the curtailment of infectious disease is the rapid and accurate identification of pathogens involved in the infection.[11,12] Phenotypic methods such as biotyping, phage typing, serotyping, and antibiotic resistance patterns have been used in the past to characterize bacterial strains. The traditional techniques were followed by molecular approaches to evaluate genotypic differences directly in prokaryotes. Restriction fragment investigation of genomic DNA by conventional agarose gel electrophoresis or pulsed-field gel electrophoresis as well as plasmid profiles generation are some of the molecular methods used.[13] Apart from genomic fingerprinting, polymerase chain reaction (PCR) methodologies employing enterobacterial repetitive intergenic consensus (ERIC) sequences as PCR primer binding sites could be used to study the distribution of repetitive sequences in different genomes.[13] It implies that bacteria vary in repeated sequences. Hence, this technique yields diverse configurations of definite primers in electrophoresis. Therefore, the ERIC-PCR technique made it so that primers are designed to attach to repetitive sequences and increase the reserves among those attached primers to repetitive arrangements.[13,14] Despite being defined as intergenic repetitive units, ERIC sequences vary from other bacterial repeats. Such sequences are spread through a wide variety of species and have remained only in intergenic regions, seemingly only inside copied or transcribed regions.[15] The basis for comparison between species and strains as reported is copy number differences, which implies that orthologous intergenic regions may have an ERIC arrangement in one species but not the other.[15]

Globally, ERIC-PCR has also been indicated for typing epidemic organisms and other enteric bacteria for epidemic detection.[16] The sequences have been used as the basis of a method for fingerprinting bacterial genomes.[13] Therefore, the working principle of ERIC-PCR is PCR primers which were made to amplify between duplicates of the ERIC sequence at neighboring positions in the bacterial genome,[15] and the technique was found to give results in a wide range of bacterial species.[13] Bakshi et al. observed that the previous studies focused on the ability of ERIC-PCR to distinguish between strains of same or closely related species.[16] Their findings have shown that ERIC-PCR could discriminate between different species members.[17] ERIC-PCR fingerprints are considerably easier to examine, rapid and practical, suited to most modern, well-equipped laboratories. As a result of its accuracy, simplicity, rapidity, and economy, this technology was generally used in strain identification and traceability investigation of disease causing organisms.[18] In addition to strain identification and traceability, the technique has been used for studying other bacterial species and even eukaryotes.[15] ERIC-PCR provides more sensitive typing results than basic phenotyping methods, hence its suitability to its traditional role in distinguishing closely associated members of a genus or typing within species rather than overall identification of bacteria.[17] It has been noted that one of the pitfalls of ERIC-PCR is the inability of dendrogram analysis to provide flawless information on relatedness in some cases.[17]

Although our study is the first reported work examining the genetic relatedness of ESBL-producing Escherichia coli using ERIC-PCR in Northwest Nigeria, a related work was done in Nigeria on ERIC-PCR as a tool for genetic classification of bacterial isolates.[17] Therefore, the present study was carried out as an important pioneer of a study to understand the genetic similarity of ESBL-producing E. coli from clinical samples using the ERIC-PCR technique.

Methods

ESBL genes detection by PCR

Amplification of resistant DNA fragments was carried out using Nexpro™ DNA polymerase (Genelabs, Korea). After thawing, Nexpro PCR master mix (2X) was vortexed and centrifuged for 30 s at 6000 rpm. A singleplex PCR assay targeting four ESBL genes of CTX-M, OXA, SHV, and TEM were carried out. The thermal cycling conditions have initial denaturation at 95°C for 3 minutes, then 35 cycles at 95°C for 30 seconds. The annealing temperature was 53°C for 30 seconds, extension was 72°C for 30 seconds and then 72°C for 5 minutes for final extension. The final hold was 4°C [Table 1].

Table 1 Oligonucleotides sequence for the detection of ESBL genes and ERIC-PCR

Results

Isolation and phenotypic detection of ESBL production in E. coli

During the study period, 576 clinical samples were analyzed. A total of 339 were sourced from male patients accounting for 58.8%, while female patients had 237 with 41.1%. The confirmed E. coli isolates from the clinical samples were 23, and their distribution was only from vaginal swab (4%), stool (35%), and urine (61%), as shown in Figure 1.

Close

Figure 1

Distribution of E. coli isolates in clinical samples

Export to PPT

Out of the 23 E. coli isolates, all 23 were ESBL producers by ESBL screening and DDST.

Antibiotic susceptibility of ESBL-producing E. coli isolates

Figure 2 shows the antimicrobial resistance (AMR) in the ESBL-producing E. coli isolates. About 95.6% and 91.3% of E. coli isolates were resistant to cefixime and cefotaxime, respectively. High resistance to ampicillin/sulbactam, ciprofloxacin, and chloramphenicol with 82.6% each, 73.9% for ceftazidime, 69.6% for gentamicin, 56.5% for nitrofurantoin, 47.8% for imipenem, and 43.5% for meropenem was observed. About 82.6% of the isolates were susceptible to amikacin.

Close

Figure 2

Antibiotic susceptibility testing result of the extended-spectrum beta-lactamases-producing E. coli using modified Kirby–Bauer disc diffusion method. Key: SAM = Ampicillin/sulbactam, MEM – Meropenem, IPM = Imipenem, F = Nitrofurantoin, CTX = Cefotaxime, CN = Gentamycin, CIP = Ciprofloxacin, CFM = Cefixime, CAZ = Ceftazidime, C = Chloramphenicol, AK = Amikacin

Export to PPT

ERIC-PCR analysis of E. coli-producing ESBL

The ERIC-PCR gel analysis of ESBL-producing E. coli revealed 1–6 bands. According to ERIC-PCR fingerprinting, the 23 E. coli isolates’ genotyping profile of the fingerprint patterns involved is above 200 bp–1500 bp. Predominant bands included 250 bp and 500 bp.

The ERIC-PCR profiles allowed differentiating the 23 E. coli isolates into four ERIC types, which are grouped into four clusters (CL1–CL4) based on the dendrogram analysis. A bulk of E. coli isolates are found in CL2 (47.8%), which has three isolates that all produced blaCTX-M, blaOXA, blaSHV, and blaTEM, followed by CL3 (26.1%), then CL1 (17.4%), and CL4 (8.7%). The cluster analysis and genetic diversity of these organisms are shown in Table 2 and Figure 3. ERIC data obtained also depicted little connection between the clinical source of the isolates and the fingerprint profile. In the case of CL1, for example, three isolates out of the four are from urine, while in CL3, 50% are from urine and 50% from the stool. The fingerprint profile in CL2 depicted that 36.4% of the isolates are from the stool, while 63.6% from urine.

Table 2 Distribution of ESBL genes by PCR among E. coli isolates with their corresponding resistance pattern using modified Kirby–Bauer disc method and their cluster

Close

Figure 3

Dendrogram of enterobacterial repetitive intergenic consensus sequences using simple match similarity matrix clustered by the UPGMA

Export to PPT

Discussion

The study aimed at determining the genetic relatedness of ESBL-producing E. coli using the ERIC-PCR technique. There are no reported studies on ERIC-PCR to determine the relationship among ESBL-producing isolates in Nigeria. As such, the study will be the first in Sokoto and Nigeria.

E. coli belong to the family of Enterobacteriaceae, which is a Gram-negative bacterium known to be a general commensal flora in humans as well as in many animal species.[25,26] E. coli was considered for this study because it is one of the essential opportunistic bacteria. They are predominantly facultative anaerobe and commensal microbiota,[27] known for many infectious diseases such as abdominal pain, vomiting, diarrhea, hemorrhagic colitis, hemolytic uremic syndrome, and thrombotic thrombocytopenic purpura.[26] E. coli is also a well-studied pathogen among the Gram-negative bacteria that play a vital position or role in the community, and nosocomial infections.[28] Specimens from which E. coli isolates were obtained may contain commensal or pathogenic E. coli. This study did not consider the pathogenicity of E. coli. Our aim was only on ESBL-producing E. coli. Additional investigation may be conducted to ascertain the pathogenic and commensal E. coli and serve as a basis for comparison.

Recently, there have been significant reports on the prevalence of ESBL-producing Gram-negative bacteria in Nigeria,[12,29] with most studies on E. coli[30,31] sourced from the hospital community. The detection of ESBLs using DDST method revealed that 100% of E. coli isolates are ESBL producers. Although, several studies have reported the prevalence of ESBL can differ from one geographic setting or another, or even from one health-care location to another, and for a specific place over a period of time.[12,32,33] This finding is the highest so far from studies in Nigeria. However, there are studies in which high rates were recorded in Nigeria. For example, in Oyo State, Southwest Nigeria, a prevalence of 79.6% was recorded[34] using DDST, while a 67.7% prevalence was recorded in a study in Enugu State, Southeast Nigeria, recorded a prevalence of 67.7%.[35] However, a study from Bauchi State, Northeast Nigeria, reported the highest prevalence of 82.3% of the isolates harbored ESBL genes in E. coli-producing ESBL.[36] With these data and the absence of newer antibiotics in the pipeline to treat Gram-negative bacteria producing ESBL[37] and the upsurge in ESBL-producing E. coli in Nigeria, therefore, a necessity for antibiotic stewardship and control use of antibiotics can never be overemphasized. Thus as noted by Dhillon and Clark, ESBLs are predominantly produced by Gram-negative Enterobacteriaceae with prominence in E. coli and Klebsiella pneumoniae, thereby harboring the ESBL enzymes chromosomes or mobile genetic elements (plasmids) [37]. Therefore, infections as a result of ESBL-producing E. coli can be a serious threat due to its inherent failure to manage diseases and ease of spread.

Another implication of these organisms harboring ESBLs genes on the plasmids is the ability of gene transfer from one organism to another, invariably leading to a more significant number of antibiotic-resistant pathogens. ESBL genes are known to carry additional resistant determinants for tetracyclines, aminoglycosides, chloramphenicol, trimethoprim quinolones, and sulfamides.[38]

Studies have shown that the effective usage of an antimicrobial drug is rendered of less use as a result of possible tolerance or development of resistance from the 1^st time this antimicrobial is used. From our study, amikacin, an aminoglycoside, is effective on most E. coli-producing ESBL isolates. The rate at which these organisms are resistant to amikacin in this study is as low as 17.4%. Many studies have reported a similar rate of resistance both within and outside Nigeria. For example, Iliyasu et al. reported ESBL-producing E. coli to be resistant to the action of amikacin by 28.3% in Nigeria in 2018.[36] Lower resistance was observed against amikacin in studies from Ethiopia (13.1%),[39] Nigeria (4%),[40] and Burkina Faso (12.3%).[41] Studies from Nigeria[42] and Qatar[9] reported 100% sensitivity to amikacin of the isolates producing ESBL. The low resistance of E. coli-producing ESBL to amikacin from our study might be attributed to the low level of usage to manage infections as a result of these organisms. Furthermore, a wide clinical practice with amikacin in the management of infections due to microbes producing ESBL is lacking. If amikacin is to be use, extra caution should be taken to prevent resistance.

Previously, the carbapenems are the preferred drugs in the management of infections as a result of ESBL-producing organisms. Carbapenems are active against ESBL-producing organisms and show tremendous in vitro activity on Enterobacteriaceae organisms producing ESBLs.[32,39] Interestingly, our study showed that the organisms’ resistance to meropenem was 43.5% and imipenem was 47.8%. The emergence of carbapenem-resistant Enterobacteriaceae may be due to rise in the use of carbapenems to treat organisms producing ESBL. It is believed that the level of this resistance might result to a likely treatment failure or impasse.

Beta-lactamase inhibitors (clavulanic acid, sulbactam, or tazobactam) usually inhibit the hydrolyzing effect ESBL-producing organisms. These combinations have been considered for the management of infections for organisms producing ESBL.[43] This study showed that the beta-lactam/beta-lactamase inhibitor was also highly resisted by the microorganisms. The combination of ampicillin/sulbactam used was resisted in 82.6% of the strains. A related study in Ghana, similar to high resistance to ampicillin/sulbactam, was recorded.[44] Based on our findings, there is limited clinical experience using ampicillin/sulbactam in managing severe infections with organisms producing ESBL. As a result, the beta-lactam/beta-lactamase combination should not be generally considered as a first-line option.

Wang et al. stated that worldwide antibiotic resistance appears to show no signs of reduction. However, it may perhaps change direction, the etiology of antibiotic resistance is multidimensional, and its after effect has impact globally.[1] In this study, the highest rate of resistance by the ESBL E. coli was observed for the following; cefixime 95.6% and cefotaxime 91.3%. Cefixime and cefotaxime have been reported in many studies indicating low sensitivity against ESBL producers.[45,46]

The blaTEM represented the dominant family of the enzyme identified in this study. Several studies have reported varied dominance at different times and locations. As reported in our study, the high prevalence is consistent with other studies in Nigeria.[35,47] However, Egwuatu et al. (2019) reported non-production of TEM type ESBL in their study.[48] In Central India, Bajpai et al. (2017) reported TEM as the most dominant enzyme family followed by CTX-M and SHV.[49] In contrast to our study, many reports in the past decade found that Enterobacteriaceae producing CTX-M-type ESBLs have increasingly become a major reason of MDR UTIs and bloodstream infections.[2,50] Our study recorded SHV as the second most prevalent with 38.9%, while other studies reported CTX-M as the dominant genes.[51,52]

There was an incidence of some isolates harboring greater than 1 gene from our study. The coexistence of ESBL genes in a single isolate further explains the growing complexity of AMR problems, and more so, a reason for further investigation. Our study revealed six isolates possessed four genes (blaCTX-M, blaOXA, blaSHV, and blaTEM) which accounted for 26.1%, one isolate possessed three genes (blaCTX-M, blaOXA, and blaSHV). Two isolates possessed two genes (blaSHV and blaTEM). Our findings with regard to possession of more than 1 beta-lactamase by an isolate are aligned to studies from Egypt,[53] South Ecuador,[54] and Port Sudan.[55] The detection of ESBL producers may be hindered in an isolate if it possesses more than 1 gene and this may complicate treatment strategy for the clinicians. Another challenge in ESBL detection is that many microbiology laboratories of hospitals or private laboratories in Nigeria do not carry out routine test for ESBL production. Hence, the necessity for laboratory testing for the detection of ESBL among bacteria isolates should be emphasized.

Bakhshi et al. have reported that ERIC-PCR has been useful for routine epidemiological investigations due to its rapid nature, relative easiness, and demonstrated discriminatory power.[16] There are few molecular typing studies carried out on E. coli-producing ESBL in Nigeria. None of these studies used ERIC-PCR on ESBL-producing E. coli, thereby making it difficult to compare our study with other studies. Our study is the first to investigate the conserved inverted repeat in E. coli-producing ESBL. These repeats were predominantly found in the second cluster (CL2). A possible reason for this is that these isolates may likely originate from the same environment. According to Waturangi et al., identical fingerprint profiles may also occur in samples of diverse origins. Therefore, ERIC-PCR has been chosen for intraspecies profiling of some bacteria.[56] From the dendrogram, some isolates showed similar ERIC profiles, two from CL1 (E11 and E12), two from CL2 (E15 and E17), and two from CL3 (E2 and E18). This indicated similarities between the isolates. A further look at the resistant patterns exhibited by these isolates tend to have the same resistant profile; E11 and E12 from CL1 were resistant to cefotaxime, cefixime, ciprofloxacin, and ceftazidime. Likewise, E15 and E17 from CL2 were highly resistant to some antibiotics (cefotaxime, cefixime, ciprofloxacin, gentamicin, ceftazidime, and imipenem). This is also an indicator these isolates might be similar in origin. On the isolates’ source showing similar ERIC profile, E11 and E12 are sourced from urine and same with E2 and E18. This shows that isolates from a particular source had similar ERIC profiles and clustered in the same group. This is also reported in a similar study on food products.[57] Bacterial isolates from this study are mainly from urine, stool, and a vaginal swab. Based on the ERIC-PCR, no significant association was found based on the source of the isolates from our study. Another possibility is that these isolates may share similar genetic material since they displayed or have the same resistant mechanism (ESBL producers). This is in line with Waturangi et al. that clustering is based on other aspects such as production of toxin, antibiotic resistance, or pathogenesis influence.[56] This can be linked to specific ESBL types produced by these organisms, as seen from the ESBL production. In CL1 and CL2, the isolates with similar ERIC profiles also have the same ESBL type (TEM). This is also a pointer to the earlier assertion that clustering can be based on other factors like antibiotic resistance. However, in CL3, despite the similarity in their ERIC profile, E18 possessed the only TEM, while E2 possessed both TEM and SHV. In CL4, where there are only two isolates, most of the antibiotics they resisted are similar (meropenem, cefotaxime, nitrofurantoin, cefixime, gentamicin, ciprofloxacin, and ampicillin/sulbactam). Therefore, comparing the antibiogram results with the dendrogram showed that they differ with similar antibiotic-resistant patterns despite E3 and E9 in the same cluster. In addition to that, CL2 has three isolates that all produced blaCTX-M, blaOXA, blaSHV, and blaTEM, and in the same cluster, majority of the isolates produced blaTEM. We observed a correlation of these isolates with the antibiotic-resistant pattern and the ESBL genes they produced. There is inconsistency in the genes’ distribution in other clusters (CL1, CL3, and CL4). In a related study on E. coli from the urinary tract, there exists difference in the number and position of ERIC sequences between strains of E. coli that are not related.The thermal cycling conditions have initial denaturation at 95°C for 3 minutes, then35 cycles at 95°C for 30 seconds. The annealing temperature was 53°C for 30 seconds,extension was 72°C for 30 seconds and then 72°C for 5 minutes for final extension.[58] Based on the banding patterns from the gel picture obtained from the ESBL-producing isolates, there seems to be a little correlation between the organisms and the ERIC print profile, bearing in mind the resistant pattern and the production of ESBL. The horizontal transfer of plasmids, transposons, and other mobile genetic elements might be responsible for this genetic relatedness. However, there was enormous inconsistency in a related study from the configurations obtained in three successive PCR runs carried out using similar conditions.[59] They further noted that the technique is based on the anticipation that complementary oligonucleotides will anneal to ERIC sequences. The DNA between the ERIC sequences may be amplified, so long as the space between ERIC sequences is 15 kilobases.[59] From the results of their study, they concluded a flaw in the ERIC-PCR technique due to its non-reproducibility. However, our study observed that the banding patterns vary slightly from each other in the three consecutive PCR ran under the same conditions. Some bands are found to be brighter than others from the gel pictures. This is aligned with a study earlier that reported bands that are bright are generally consistent. Still, lighter bands are often subject to interexperimental variability and are often ignored by human readers.[59] Computer software was used to account for bands’ presence or absence since variations in the bands’ intensity. As such, band gain or loss can pose a problem for computer-based analysis.[59] Light bands were removed from analysis by some researchers. This is to eliminate fingerprints’ irreproducibility, but light bands were considered in this study. Therefore, selecting which bands will be incorporated in a specific molecular fingerprint presents personal bias and decreases the fingerprint down to only a few bands, thereby considerably decreasing the technique’s discriminatory power.[60]

Conclusion

The majority of the ESBL-producing E. coli isolates are sensitive to amikacin and highly resistant to the third-generation cephalosporins. The blaTEM represented the dominant family of enzymes identified genes. Overall, ERIC obtained indicated some evidence in the genetic relatedness of the ESBL genes among E. coli isolates.

Authors’ Declaration Statements

Availability of Data and Material

The data that support the findings of this study are available from the corresponding author, NT, on request.

Competing Interest

The authors declare that there are no conflicts of interest.

Funding Statement

Nil.

Authors’ Contributions

BOO: Conceptualization, data collection, investigation, methodology, and writing – original draft preparation. NT: Formal analysis, methodology, data collection, writing – original draft preparation, and writing – review and editing. ROB: Conceptualization, supervision, methodology, validation, and writing – review and editing. ATO: Conceptualization, supervision, methodology, validation, and writing – review and editing. EBBO: Supervision, methodology, validation, and writing – review and editing. MY: Conceptualization and writing – review and editing.