Fecal Colonization with Extended-Spectrum Beta-Lactamase and AmpC-Producing Escherichia coli

Background. Extended-spectrum β-lactamases (ESβLs) and AmpC β-lactamases cause β-lactam resistance in Escherichia coli. Fecal colonization by ESβL- and/or AmpC-positive E. coli is a source of nosocomial infections. Methods. In order to investigate inpatient fecal colonization by ESβLs and AmpC, antibiotic sensitivity tests were conducted and minimum inhibitory concentrations (MICs) were determined using the disk diffusion method and E-test, respectively. Characterization of ESβL and AmpC was performed using E-test strips, and a set of PCRs and DNA sequence analyses were used to characterize the ESβL and AmpC genes. Results. The whole collection of E. coli isolates (n = 50) was sensitive to imipenem, tigecycline, colistin, and fosfomycin, while 26% of the isolates showed reduced susceptibility to ceftazidime (MIC ≥ 4 μg/mL). ESβL was phenotypically identified in 26% (13/50) of cases, while AmpC activity was detected in two ESβL-producing E. coli isolates. All ESβL-producing E. coli were positive for the CTX-M gene, eleven isolates carried bla CTX-M-15, and two isolates carried bla CTX-M-14 gene. Two CTX-M-positive E. coli isolates carried bla CMY-2. Conclusions. The alimentary tract is a significant reservoir for ESβL- and/or AmpC-producing E. coli, which may lead to nosocomial infection.


Introduction
A remarkable increase in fecal colonization rates with extended-spectrum beta-lactamase-(ES L-), AmpC-plasmid mediated-, and/or carbapenemases-producing Enterobacteriaceae has been reported in many regions worldwide [1,2]. Infections caused by Enterobacteriaceae, which are resistant to -lactams, are coupled with the inappropriate use of antibiotics and/or a prolonged period of hospital admission. The rising use of carbapenems for empirical treatment of nosocomial infections has led to fast global dissemination of carbapenemase-positive enterobacterial strains [3]. ES Ls arise through point mutations in TEM-1/TEM-2 and SHV-1. However, over the last three decades, non-TEM and non-SHV ES Ls strains have been detected, primarily CTX-M. Enterobacteriaceae that produce CTX-M enzymes have shown rapid and concerning dissemination and have been documented as the most prevalent etiological infectious agents [4]. ES L confers resistance to penicillins, cephalosporins, and monobactam (aztreonam), but they are susceptible to cephamycins (cefoxitin and cefotetan) and carbapenems (imipenem, meropenem, and doripenem) and are typically reserved by inhibitors of Ambler class A -lactamase (clavulanic acid, tazobactam, or sulbactam). Most ES Ls can hydrolyze fourth-generation cephalosporins. While AmpC -lactamases confer resistance to penicillins, third-generation cephalosporins, monobactam, and cephamycins, they are sensitive to carbapenem and are not inhibited by -lactamase inhibitors; however, they are inhibited by cloxacillin [5][6][7].
Numerous studies in Saudi Arabia have focused on the identification of ES L-producing strains from clinical specimens [4], but there are few reports on the fecal colonization of ES L-producing isolates in Saudi Arabia. Therefore, in the present study, we determine the incidence of ES Land/or AmpC cephalosporinase-producing Escherichia coli isolates in human fecal flora and investigated the genes encoding the corresponding enzymes.

Bacterial Identification.
Fifty different E. coli isolates were isolated from 50 stool samples of different inpatients carriers, under nonoutbreak conditions, at a hospital in Riyadh, Saudi Arabia, from April 2014 to June 2014. Briefly, fresh stool specimens were aseptically collected and transported to the microbiology laboratory. Stool samples were suspended in sterile phosphate-buffered saline, pH 7. A 100 L volume was directly inoculated onto blood agar and Eosin Methylene Blue agar (Oxoid Microbiology Products, Hampshire, UK). After 48 h incubation at 37 ∘ C, the isolated organisms were identified by conventional procedures and automated identification systems with the API20E identification kit (bioMerieux, Marcy l'Etoile, France). These isolates were preserved in brain heart infusion broth containing 20% glycerol at −70 ∘ C.

Phenotypic Detection of ES L.
The isolates showing reduced susceptibility to ceftazidime (CAZ), cefotaxime (CTX), or aztreonam (ATM) (minimum inhibitory concentration (MIC) ≥ 1 g/mL or zone diameter ≤ 22 mm) were selected for screening of ES L production (Clinical and Laboratory Standards Institute (CLSI), 2014). E-test ES L strips were used in accordance with the manufacturer's instructions to evaluate ES L production. The CAZ/ceftazidime + clavulanate-(CAZ/CAL-) ES L E-test strip was used to detect ES L production. The test is considered positive if the ratio of MIC of CAZ/CAL is ≥8. To inhibit AmpC -lactamase, the CAZ/CAL-ES L E-test was carried out on cloxacillin Mueller-Hinton agar, and the results were interpreted in a similar manner.

Phenotypic Detection of AmpC.
The isolates showing reduced susceptibility to cefoxitin (FOX) or cefotetan (CTT) (zone diameter of 18 or 16 mm, resp.) were selected for screening of AmpC enzyme production [9]. The phenotypic detection test consists of a strip containing CTT on one end and CTT-cloxacillin (CTT/CXT) on the other end. Ratios of the MICs of CTT/CXT ≥ 8 are considered to indicate positive AmpC -lactamase production.

Screening for the Presence of -Lactamase Genes.
The isolate was cultured in 2 mL of Tryptic Soy Broth (Difco, Franklin Lakes, NJ, USA). A 200 L volume of overnight culture was heated at 99 ∘ C in a heat block for 10 min. The obtained DNA was used in polymerase chain reaction (PCR) assays on a Techne Flexigene Thermal Cycler (Techne, Duxford, Cambridge, UK). Positive and negative controls were included in all PCR assays. All PCR products were analyzed on 0.8% agarose gels (incorporated with 0.5 mg/L ethidium bromide) and then visualized under UV light (Pharmacia LKB, Biotechnology AB, Gothenburg, Sweden) and photographed using a documentation system (CE, DP-CF-011.C, European Union).
The PCR primers used are listed in Table 1. The primers were used to search for class A -lactamase genes ( TEM , SHV , OXA-1 , and CTX-M families) and class Clactamase genes ( CMY , MOX , FOX , DHA , ACC , ACT , MIR , EBC , CIT , and BIL ). PCR assays were conducted as previously described [8].

Sequencing of -Lactamase Genes.
Purification of PCR amplicons was performed using a PCR purification kit (Qiagen, Hilden, Germany). PCR products of bla genes were sequenced on both strands using PCR primers to determine their molecular types. DNA sequences were analyzed using the ABI 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's recommendations.

Bacterial Identification.
Fifty fecal E. coli samples were isolated randomly from hospitalized patients in Riyadh, Saudi Arabia. The patients were treated for noninfectious diseases under nonoutbreak conditions. Escherichia coli isolates were identified manually and according to the API20E identification kit (bioMerieux, Marcy l'Etoile, France).

Characterization of ES L and AmpC .
Thirteen of the 50 E. coli isolates, which showed reduced susceptibility to CAZ, CTX, or ATM (MIC ≥ 1 g/mL or inhibition zone ≤ 22 mm), were selected for screening of ES L and AmpC enzyme production using CAZ/CAL-ES L and CTT/CXT-AmpC E-test strips. Thirteen isolates were positive for ES L and two ES L-positive isolates produced AmpC -lactamase.
PCR was used to detect ES L genes and AmpC plasmidmediated genes in ES L-and AmpC-positive E. coli isolates ( = 13). The results of PCR and DNA sequencing of bla genes are shown in Table 2.

Antimicrobial Resistance Pattern of ES L and AmpC
Enzyme-Positive Isolates. The MICs of 17 antimicrobial agents were determined for E. coli fecal isolates. The results of the susceptibility pattern for ES L and AmpC enzymeproducing E. coli isolates are illustrated in Table 2.

Discussion
The human and animal alimentary tracts are vital reservoirs for ES L-, carbapenemases-, and AmpC enzyme-producing Enterobacteriaceae. Patient-to-patient transmission of resistant microorganisms may occur in hospitals [10,11]. The overuse of antibiotics has recently been associated with the emergence of resistant intestinal bacteria, particularly ES L-, carbapenemases-, and AmpC enzyme-producing Enterobacteriaceae. Numerous studies have demonstrated that exposure to -lactam antibiotics is a risk factor for the selection of multidrug-resistant E. coli [12]. Therefore, the current study examined the antimicrobial resistance patterns of fecal E. coli isolates, as well as the molecular basis for their -lactam resistance mechanisms, using phenotypic and genotypic methods. The present study included 50 patients admitted to a hospital in Riyadh, Saudi Arabia, from April 2014 to June 2014. These patients were treated for noninfectious diseases under nonoutbreak conditions. Fifty fecal stool specimens collected from the 50 patients were cultured on blood agar and EMB agar as described in Materials and Methods. Fifty suspected E. coli isolates were selected and the isolates were identified by conventional procedures and using the API20E identification kit. All isolates were identified as E. coli. Production of -lactamases is the main mechanism of -lactam resistance in Gram-negative bacteria, including E. coli [6,7]. Several -lactamases (ES Ls, AmpC enzymes, and carbapenemases) have been previously reported in fecal E. coli isolates [13][14][15].
In the present study, 100% of 50 E. coli isolates were found to be sensitive to imipenem and 13 (26%) of 50 isolates were resistant or showed reduced susceptibility to CAZ, CTX-M, or ATM. The carbapenem susceptibility results indicated that our isolates did not harbor carbapenemase, while 26% (13/50) of the E. coli isolates harbored ES L and/or AmpClactamase. Two of 13 isolates exhibited reduced susceptibility to cephamycins (FOX and CTT).
Phenotypic screening for the presence of different types of -lactamases was conducted using E. coli isolates. Thirteen E. coli isolates were selected for screening of ES L and AmpC -lactamase production using CAZ/CAL-ES L, and CTT/CXT-AmpC E-test strips were used to detect ES L production. Using phenotypic detection methods, all isolates were found to produce ES L, while two isolates phenotypically produced AmpC enzyme. A battery of PCR assays was conducted to detect ES L genes and AmpC plasmidmediated genes in the 13 E. coli isolates. Therefore, class A and class C -lactamase genes were tested. The PCR-purified product was subjected to DNA sequencing to identify the gene variants. The results of molecular characterization of bla genes are shown in Table 2. PCR amplification and DNA sequencing analyses of the PCR products showed that all isolates possessed a CTX-M-type ES L and that CTX-M-15 was present in 1 isolate, while two isolates contained CTX-M-14 . Other CTX-M families were not detected. The gene encoding CMY-2 enzyme was detected in two E. coli isolates. CMY-2-positive isolates are concomitant with CTX-M-15. All E. coli isolates ( = 13) were positive for TEM-1 , while 61.5% and 23% of the isolates contained OXA-1 and SHV-1 , respectively. The increase in expression of the AmpC -lactamases may mask the recognition of ES Ls [16]. Therefore, in the present study, the genotypic methods revealed that all 13 strains were ES L CTX-Mpositive, while phenotypic methods showed that 11 strains were ES L-positive and two strains were AmpC enzymepositive. AmpC-producing strains producing CTX-M-15 may act as a dormant reservoir for ES Ls.
Numerous studies have documented a remarkable increase in intestinal colonization rates with ES L-and AmpC enzyme-producing Enterobacteriaceae in many countries [2,10,11,[14][15][16][17]. The prevalence of ES L-producing E. coli fecal isolates varies widely from country to country, from region to region, and at different time periods. A high incidence of fecal carriage rate of ES L-producing E. coli has been observed in Asia, Africa, and South America [13,14,[18][19][20][21], while a significantly lower prevalence of ES L-producing E. coli fecal isolates was reported in most European countries [22,23]. In Argentina, the rate of fecal carriage of Enterobacteriaceae-resistant strains to thirdgeneration cephalosporins was 26.8% [20]. Villar et al. [20] reported that 20.22% and 6.7% of fecal strains were colonized by ES L-and AmpC enzyme-producing Enterobacteriaceae [20]. In Egypt, Al-Agamy et al. reported that 22.6% and 3.22% of hospitalized patients were colonized by ES Land AmpC-positive E. coli, respectively. The CTX-M -like gene was the predominant ES L gene, detected in 71.4% of ES L-producing E. coli isolates [21]. In a recent study in Egypt, Bassyouni et al. reported that 21% and 3% of patients were colonized by ES L-and AmpC-producing E. coli, respectively. They also found that SHV gene was the predominant ES L gene, detected in 81.8% of the resistant E. coli isolates [13]. In Korea, 20.3% of fecal Enterobacteriaceae members were ES Ls [19]. In India, the prevalence of ES L-positive E. coli isolates was 19% in healthy volunteers from the community [14]. In Libya, 13.4% and 6.7% of E. coli isolates were ES Ls-and AmpC-positive, respectively [18]. In a previous study in Saudi Arabia, 17.7% of strains were found to be ES L-positive [24]. A high (26.1%) prevalence was detected in inpatients, followed by outpatients (15.4%), and the lowest prevalence rate (13.1%) was detected in healthy individuals [24]. In the present study, the prevalence of ES Lproducing E. coli was 26%. Despite differences in the date and region of isolation, the prevalence of fecal carriage rate of ES L in the present study (26%) was in agreement with the prevalence (26.1%) reported by Kader et al. In contrast, the prevalence rate of ES L-producing Enterobacteriaceae was 2.9% among healthy Swedish children. Escherichia coli containing CTX-M -lactamase predominated, and only one E. coli isolate harbored genes encoding for CMY [22]. The carriage rate of ES L-and AmpC enzyme-producing E. coli was 3.57% and 2.38% among 84 Danish army recruits, 6 BioMed Research International respectively.
CTX-M-14 gene was the predominant ES L gene detected in three (100%) ES L-producing E. coli isolates, while CMY-2 was detected in two AmpC enzymeproducing E. coli isolates [23]. In Spain, the prevalence of ES L and AmpC enzyme carriers was 5.06% and 0.59%, respectively [25].
MICs were determined for ES L-and AmpC enzymeproducing E. coli ( = 13) isolates. The results of MIC are shown in Table 2.
In conclusion, a high incidence of carriage of ES Lpositive E. coli fecal isolates among hospitalized patients in Riyadh was detected, reaching 26%, with CTX-M-15 (84.6%) being the most predominant gene. The emergence of fecal carriage of CMY-2-producing E. coli among hospitalized patients has been reported to be 4%. These outcomes emphasize the importance of the intestinal tract as a reservoir for ES L-and AmpC enzyme-producing E. coli, which may lead to nosocomial infection. The admission of colonized fecal carriers of ES L-and AmpC-positive E. coli to the medical setting increases the possibility of other patients acquiring infection in the same hospital. Our results emphasize the necessity for continuous surveillance in hospitals to detect the ES L-, AmpC enzyme-, and carbapenemase-producing strains and multidrug strains as well applying effective strategies for antimicrobial therapy and infection control measures to decrease the abuse and misuse of antimicrobial agents against resistant strains and to prevent their spread.