Prevalence and molecular epidemiology of extended spectrum β-lactamase producing Escherichia coli from hospital and community settings in Egypt

Article history: Received on: 13/11/2015 Revised on: 20/12/2015 Accepted on: 04/01/2016 Available online: 26/01/2016 The prevalence and molecular epidemiology of Escherichia coli that produce extended spectrum β-lactamase (ESBL) in Cairo, Egypt was investigated. Ninety E. coli isolates were collected along the period of September to November 2012 from hospital and community settings. Antibiotic susceptibility of the E. coli isolates was determined by disk diffusion method. All isolates were screened phenotypically for ESBL production by combination disk method. The presence of bla CTX-M-I, bla CTX-M-IV, bla TEM and bla SHV genes in ESBLproducing E. coli was examined by PCR and sequencing experiments. The results showed high prevalence of ESBL-producing E. coli, 52% of the collected isolates were ESBL producers. The ESBL-producing isolates significantly (P < 0.05) had increased resistance compared with non–ESBL producers to cefuroxime, cefotaxime, ceftazidime, cefepime, ciprofloxacin, and co-trimoxazole. Imipenem was the most effective drug against ESBL producing isolates. All ESBL producing E. coli isolates were multi drug resistant (MDR) to eight antibiotics or more. Detection of ESBL genes in selected MDR-ESBL producing E. coli revealed that bla CTX-M-I was the most prevalent ESBL type. It is clear that the prevalence of ESBL producing E. coli in Cairo, Egypt is alarming high. This study is useful for clinician in order to improve the empiric treatment.


INTRODUCTION
In the early 1980s, third-generation cephalosporins were introduced to the clinical practice as β-lactam antibiotics able to overcome resistance caused by the common β-lactamases produced by Escherichia coli (Livermore, 2012).However, within few years; E. coli produced mutated versions of these βlactamases called extended spectrum β-lactamases (ESBLs) which enable them to neutralize the activity of expandedspectrum cephalosporins, and monobactams (Livermore, 2012).ESBLs are enzymes capable of conferring bacteria resistance to penicillins, 1 st , 2 nd , 3 rd , 4 th generation cephalosporins, and monobactams by hydrolysis of these antibiotics.ESBLs do not hydrolyze cephamycins (e.g., cefoxitin or cefotetan) or carbapenems (e.g., imipenem and meropenem), and they are inhibited by β-lactamase inhibitors such as clavulanic acid (Paterson and Bonomo, 2005).The majority of ESBLs belongs to class A Ambler classification which includes the bla SHV, bla TEM, and bla CTX-M types (Paterson and Bonomo, 2005).Since 2000, .

* Corresponding Author
Email:metwally@ymail.com E. coli producing bla CTX-M have emerged worldwide as an important cause of community-onset UTIs (Lahlaoui et al., 2014).ESBLs are mostly encoded by large plasmids (up to 100 kb and even more) that are transferable from strain to strain and between bacterial species (AitMhand et al., 2002).Genes encoding ESBLs are frequently found on the same plasmid as genes encoding resistance for other classes of antibiotics such as aminoglycosides, tetracyclines, and sulfonamides.Taking into the account, many of E. coli strains possess chromosomal changes that confer resistance to fluoroquinolones (Livermore, 2012).As a result, ESBL producing E. coli is frequently multidrug resistant (MDR), posing particular difficulties in the treatment of infections, especially in critically ill patients.ESBL-producing E. coli have recently reported in Egypt (Abdul Rahman and El-Sherif, 2011).A more comprehensive survey of ESBLs from Egypt is urgently needed not only for the hospital setting but also for the community.This study was carried out to determine the prevalence of ESBLproducing E. coli in hospital and community settings in Cairo, Egypt and to assess antibiotic susceptibility patterns of these isolates in order to define appropriate antimicrobial therapy.We are also aimed to reveal basic aspects of the molecular epidemiology of these isolates in Cairo, Egypt.

Bacterial isolates
Bacterial isolates were collected along the period of September to November 2012 from three hospitals (University of Ain Shams Hospital, El-Salam International Hospital, National Cancer Institute) and two community clinical labs (Cairo Lab -Helwan branch, Shaker Lab -Misr El-Gedida branch).The isolates were non-consecutive (Only a single positive culture per patient).Isolates were streaked on Tryptic soy agar medium to ensure purity and viability.
Isolates that were gram-negative, lactose-fermenting, non-swarming, indole positive, oxidase negative, producing acid slant/acid butt reaction with or without gas on triple sugar iron medium test, citrate negative and urease negative identified as E. coli (Engelkirk and Duben-Engelkirk, 2007).Isolates identified as E. coli, were stored on glycerol Tryptic soy broth at -20° C freezer.When a fresh seed-stock vial is required, it was removed and used to inoculate a series of working cultures.Quality control strains; E. coli ATCC 25922 and Klebsiella pneumoniae ATCC 700603 were kindly provided by the U.S. naval medical research unit no. 3 (NAMRU-3) in Egypt.
The diameters of the inhibition zones were interpreted by referring to CLSI document M100-S21 (CLSI, 2011), and the examined isolates were reported as susceptible, intermediate, or resistant to the agents under test.Control strain used to validate susceptibility tests was E. coli ATCC 25922.Quality control strain was tested daily to ensure the test system is working and gives results within specified limits.
ESBL testing requires use of both cefotaxime and ceftazidime, alone and in combination with clavulinic acid, ≥ 5mm increase in a zone diameter for either antibiotic (ceftazidime or cefotaxime) tested in combination with clavulinic acid versus its zone when tested alone was determined to be ESBL producer.E. coli ATCC 25922 (negative control) and Klebsiella pneumoniae ATCC 700603 (positive control) were used as quality control strains.

Statistical analysis
Fisher exact test (Agresti, 1992) was used as a statistical tool to determine the significance of difference between the resistance level to various antibiotics in ESBL and non ESBL producing isolates, P < 0.05 was considered significant.

Detection of ESBL genes PCR amplification
Boiling lyses method was used for extraction of DNA from isolates under test (Moore et al., 2004).Klebsiella pneumoniae ATCC 700603 was used as positive ESBL producer ( bla SHV-type).Primers (Biosearch Technologies, USA) were purchased in lyophilized form.DreamTaq Green PCR Master Mix (2X) (Thermo Scientific, USA) containing (DreamTaq DNA polymerase, 2X DreamTaq green buffer, dATP, dCTP, dGTP, dTTP, 0.4mM each, and 4mM MgCl 2 ) was used.PCR amplification was used to identify the presence of bla CTX-M genes using specific primers that targeted bla CTX-M group I and IV (Pitout et al., 2004).The presence of genes encoding bla TEM and bla SHV were also analyzed by PCR (Hanson et al., 1999).The primers sequence, sizes of the expected amplification product and nucleotide positions are listed in Table 1.
Amplifications of ESBL genes were carried out on a DNA Thermal Cycler instrument (PTC-100; MJ Resrach Inc, USA).The composition of the reaction mixture was as follows: 12.5 µl of DreamTaq Green PCR Master Mix (2X), 3 µl of bacterial lysate, 1 µl of each primer (10 pM/ µl), and nuclease-free water was added to complete a final PCR reaction volume to 25 μl; the mixture was mixed gently before running the PCR program.The PCR program consisted of an initial denaturation step at 96°C for 4 minutes, followed by 24 cycles of; DNA denaturation at 96°C for 30 seconds, primer annealing at 55°C for 30 seconds, and primer extension at 72°C for 2 min, and final elongation step at 72°C for 10 minutes (Pitout et al., 1998).Aliquots of 10 µl of PCR products were analyzed by gel electrophoresis with 1.6 % agarose gels in 1× TAE buffer for 30 minutes at 90 V. Gels were stained with 2 µl ethidium bromide (10 mg/L), visualized and compared to 100 bp DNA ladder (fisher bioreagents, Canada) by UV transilluminator.

Sequencing of ESBL genes
Randomly selected products from the CTX-M group I and CTX-M group IV PCRs were sequenced on an Applied Biosystems 3100 DNA Analyser to identify the specific ESBL type.The primers used for sequencing were the same as those used for PCR.The PCR products were purified with high pure PCR product purification kits (Roche Molecular Biochemicals, Espoo, Finland).ABI BigDyeTM terminator cycle sequencing kit version 3.1 (Applied Biosystems, Espoo, Finland) was used.The DNA sequences were analysed and translated into amino acid sequences with DNA sequencing analysis software, version 5.3.1 (Applied Biosystems, Espoo, Finland).Amino acid sequences were compared to the known CTX-M variants by BLAST at website search (http://www.ncbi.nlm.nih.gov/blast).

RESULTS AND DISCUSSION
ESBLs are now commonly found in E. coli isolates from patients in nursing homes and long term-care facilities, and even in patients with community-acquired infections (Lahlaoui et al., 2014).In this study; during the 3 months study period; a total of 90 E. coli isolates were collected from three hospitals included University of Ain Shams Hospital (32 isolates), El Salam International Hospital (32 isolates), National Cancer Institute (6 isolates), and two clinical community labs included Cairo Lab -Helwan branch (10 isolates), Shaker Lab -Misr El Gedida branch (10 isolates).The majority of isolates were from urine samples (67 isolates).There were 7 isolates from stool samples, 5 from sputum samples, 4 from pus samples, 2 from groin samples, 2 from drainage samples, and 3 isolates from miscellaneous sites.
High prevalence level of ESBL producing E. coli was recorded in the present study, as of the 90 E. coli isolates collected; 47 (52%) were ESBL producers (Table 2).Many studies from Egypt also recorded high percentage of ESBL production among E. coli isolates.In a study carried out at Cairo University Hospitals, 400 bacterial isolates from 632 stool samples were found to be ESBL producers.Out of these 400 isolates, 285 (71.25%) were identified as E. coli (Abdul Rahman and El-Sherif, 2011).In another study from Egypt, which included only blood stream infections in patients admitted to intensive care units (ICUs), the proportion of ESBL-producing E. coli was 39% (Saied et al., 2011).In contrast to our results, low prevalence level of ESBL-producing E. coli was reported in many African countries such as Morocco (1.3%) (Barguigua, et al., 2011), Nigeria (12.8%) (Aibinu et al., 2012), and South Africa (7.6%) (Brink et al., 2012) (Agrawal et al., 2008).On the other hand; in Taiwan; Hsieh et al. (2010) conducted a study on hospitalized patients with E. coli bacteremia, and found that the frequency of ESBL producers was 4.7%.As it was expected, all ESBL producing E. coli isolates in this study were resistant to all cephalosporins tested (Table 2).
Co-resistance to non-β-lactam antibiotics is common among ESBL producers, especially for fluoroquinolones, cotrimoxazole and/or aminoglycosides.In this study, the obtained results revealed that; resistance level to ciprofloxacin and cotrimoxazole was highly significant in ESBL producing E. coli in comparison with non ESBL producing isolates, however; there was no significant difference in the resistance level to gentamicin between ESBL producing and non ESBL producing E. coli (Table 2).
Other studies reported high degree of resistance to fluoroquinolones and co-trimoxazole and less resistance degree to gentamicin among ESBL producing E. coli when compared to non-ESBL producing E. coli (Chander andShrestha, 2013, Somily, et al., 2014).In contrast to our results; some studies reported significant difference of resistance level to gentamicin between ESBL producing and non ESBL producing E. coli (Cagan Aktas et al., 2014).
All ESBL producing E. coli isolates in this study were susceptible to imipenem (Table 2), indicating that it could be the proper drug for treating serious infections caused by ESBL producing E. coli.Many studies reported high susceptibility rate to imipenem (100%) among ESBL producing E. coli (Hawser et al., 2011, El-Bouamri et al., 2014).In this study, E. coli isolates expressed multi drug resistance (MDR) phenotype at high level.Eighty six (95%) isolates were found to be resistant to 3 or more antibiotic classes (Table 3).
MDR level was higher among ESBL producers than in non-ESBL producers; all ESBL producing E. coli were resistant to 8 antibiotics or more.Other workers also found high proportions of multidrug resistance among ESBL producers (Serefhanoglu et al., 2009;Chander and Shrestha, 2013).High resistance level to all antibiotics used in this study with the exception of imipenem and high prevalence of MDR-ESBL producing E. coli detected in our study as well as in the previous reports from Egypt compared to other countries, may be attributed to the uncontrolled consumption of large amount of antibiotics by patients in Egypt.
A patient's previous exposure to an antibiotic, especially to extended spectrum cephalosporins, and fluoroquinolones has been widely reported as a risk factor for infection with ESBLproducing bacteria (Hsieh et al., 2010;Goulenok et al., 2013).Twelve ESBL-producer E. coli isolates; which showed resistance to 10 antibiotics in the test panel (Table 3); were further subjected to PCR assay for the detection of ESBL genes and the results revealed that bla CTX-M-I was the most prevalent gene type (Table 4).
Co-production of bla CTX-M ( bla CTX-M-I or bla CTX-M-IV) with bla TEM enzymes was observed in 8 isolates.Sequencing of one amplicon from bla CTX-M-I group, was characterized as bla CTX-M-15 enzyme, the sequence obtained was submitted to GenBank under KP325147.1 accession number.Sequencing of one amplicon from bla CTX-M-IV group was characterized as bla CTX-M-14 enzyme, the sequences obtained was submitted to GenBank under NG_041766.1 accession number.Many studies also reported that bla CTX-M-I was the most prevalent type among ESBL producing E. coli in Egypt.For example, Fam et al. (2011) collected 47 E. coli isolates from all specimens' types at Theodor Bilharz research institute in Egypt, and examined them for presence of ESBL genes.The obtained results showed that the all 47 isolatess produced bla CTX-M-15 which belongs to bla CTX-M-I group.In another study, also carried out at Theodor Bilharz research institute on 44 E. coli isolates collected from stool samples; bla CTX-M-I group was the most prevalent enzyme; detected in 29 isolates, whereas bla CTX-M-IV group genes were detected in 14 isolates (Fam et al., 2014).Coproduction of bla CTX-M and bla TEM in ESBL producing E. coli was previously reported in other studies from Egypt (Al-Agamy et al., 2006) and Turkey (Gorgec et al., 2015).Worldwide; several studies reported that bla CTX-M was the most prevalent ESBL type among ESBL producing E. coli (Hoban et al., 2014;Wang et al., 2014, Gorgec et al., 2015).

CONCLUSION
Our data pointed out that the prevalence of ESBLproducing E. coli is high in Cairo, Egypt.Carbapenems should be regarded as the drugs of choice for serious infections with ESBL producing E. coli.If the causative agent has been found as susceptible to other antibiotics such as aminoglycosides, they can be used in the treatment of non-life-threatening infections caused by ESBL producing E. coli to reduce carbapenem utilization.The bla CTX-M-I group was the most prevalent ESBL type, especially in combination with blaTEM enzymes.Continued surveillance, appropriate use of antibiotics, and implementation of strict infection control measures are recommended to decrease ESBL frequency.
. Prevalence of ESBL producing E. coli isolates in the United States and Europe was lower than that reported in Egypt.Bhusal, et al. (2011) reported that out of 443 E. coli isolated from cancer patients at a cancer center in USA, only 41 (9.2%) isolates were ESBL producers.Also, Hawser et al. (2011) examined 3160 isolates of E. coli collected from 44 hospitals in different European countries (i.e.France, Germany, Greece, Romania, Spain, Turkey, Estonia, Italy, Latvia, Lithuania, Portugal, and UK), and found that only 11% of isolates were ESBL producers.Variable percentage of ESBL-producing E. coli was reported in Asia by different workers, for example, a study from India reported that 30% of E. coli isolates obtained from different clinical samples were found to be ESBL producers

Table 1 :
Primers used for amplification of DNA of ESBL genes.

Table 2 :
The antimicrobial resistance percentages of ESBL-producing and non ESBL-producing E. coli isolates.
resistance level between ESBL and non ESBL isolates was significant at p <0.05.N/D: not determined

Table 3 :
Antibiotic resistance pattern of the multi drug resistant (MDR) E. coli isolates.

Table 4 :
Occurrence of different ESBL genes among the selected 12 MDR-ESBL E. coli isolates.