Abstract
The aim of the study was to detect the prevalence of CTX-M-type extended-spectrum β-lactamases (ESBL) in Escherichia coli strains isolated in healthy chickens at poultry farms in Tenerife, Spain. From November 2012 to February 2013, 260 live chickens were screened. Samples were cultured in chromogenic media. Suspect strains were identified by Vitek 2 system and ESBL production was confirmed by the double-disk synergy test. Pulsed-field gel electrophoresis (PFGE) was performed with XbaI (Promega, Madison, WI) to ESBL–E. coli isolates. The presence of CTX-M-type was detected by real-time polymerase chain reaction. Of 260 rectal swabs, 237 (91.1%) were ESBL–E. coli, 196 (75.38%) strains were characterized by PFGE, and CTX-M-type was detected in 116 (59.1%) of these strains. With respect to the susceptibility patterns of E. coli bla CTX-M strains, 7.8% showed resistance to more than two non-β-lactam antibiotics. In our area, the prevalence of CTX-M-type in E. coli isolated in chicken was even higher than those found in other countries. The impact of food animals as a possible reservoir for ESBL-E. coli, especially CTX-M-type ESBL, and the dissemination of such strains into the food production chain need to be assessed.
Introduction
E
Though the first descriptions of extended-spectrum β-lactamases (ESBL) strains producers in clinical isolates date back to the end of the 1980s and beginning of the 1990s (Knothe et al., 1983; Baquero et al., 1988; Fernández-Rodríguez, 1992), it has not been until the last decades that an important and “explosive” increase of this type of enzyme has taken place, as indicated by diverse studies (Pitout and Laupland, 2008; Peirano and Pitout, 2010; Pitout, 2010; Enoch et al., 2012; Lowe et al., 2013; Korzeniewsk and Harnisz, 2013). Whereas early ESBLs from humans mainly evolved from TEM and SHV β-lactamases, the significance of CTX-M-type enzymes has increased over the last decade (Ewers et al., 2012).
The first isolation of an ESBL-producing microorganism (SHV-12 β-lactamase producing E. coli) from an animal occurred in 1998. Since then, there has been an alarming increase in the detection of ESBLs, mainly of the CTX-M group, in E. coli strains in healthy animals destined for human consumption and, to a lesser extent, in pets and even in wild animals (Torres and Zarazaga, 2007; Dierikx et al., 2010; Ewers et al., 2012; Dierikx et al., 2012).
Studies performed confirm detection of ESBL-producing isolates in different animals and meats with varied percentages of positivity (Horton et al., 2011; Tamang et al., 2012; Trott, 2013; Friese et al., 2013; Seifert et al., 2013; Donati et al., 2014). Moreover, a high prevalence of ESBL-producing isolates is found in broilers and in broiler meat (Dierikx et al., 2010; Li et al., 2010; Gregova et al., 2012; Mnif et al., 2012; Zheng et al., 2012; Dierikx et al., 2013a; Reich et al., 2013; Ferreira et al., 2014). Dierikx et al. (2013b) demonstrated that ESBL/AmpC-producing isolates are found at every level of the broiler production pyramid. In broiler production farms these isolates spread very quickly, leading to a high prevalence.
ESBLs enzymes, hydrolyzed third- and fourth-generation cephalosporins, as well as monobactams are inhibited by clavulanic acid and cephamycins, such as cefoxitin. In contrast with other multidrug-resistant bacteria, it is suspected that ESBL-mediated resistance has been spreading mainly throughout the community and not primarily within healthcare-related institutions (Leistner et al., 2013). Although the source of the colonization of ESBL-producing bacteria in humans is not completely understood, circumstantial evidence also points to a foodborne source.
Some studies found a similar distribution of ESBL drug-resistant genes among isolates from retail chicken meat and poultry, and isolates from colonized as well as infected humans (Leverstein-van Hall et al., 2011; Overdevest et al., 2011).
When the uptake of these isolates occurs (through consumption or handling of contaminated food), they are able to share their ESBL genes with other bacteria in the gastrointestinal tract by plasmid-transfer, especially when selecting compounds such as β-lactam antibiotics are administered (Dierikx et al., 2013a).
The investigation of ESBL-producing E. coli in food-producing animals will help us understand the distribution status of these isolates and establish proper prevention protocols. Additionally, surveillance data of ESBL-producing E. coli isolates from animals and resistance mechanisms can also be used to guide the application of antimicrobials to food animal production and infection chemotherapy (Li et al., 2010).
The aim of the study was to detect the prevalence of CTX-M-type β-lactamases ESBL in E. coli strains isolated in healthy chicken at poultry farms in Tenerife, Canary Islands, Spain, as it is geographically remote from the European and African continents and the animals are more isolated than those of the continent.
Methods
Collection of samples
A cross-sectional prevalence study was conducted. During the period November 2012 to February 2013, a total of 260 live chickens were screened with a rectal swab. A randomized selection of 26 animals was taken from 10 farms (out of 25) for local consumption. Each farm was visited once during the study.
The farms are dedicated to chicken consumption production, with an average of 18,100±3800 per farm (max 26,000 and min 15,000). All were closed wards, with controlled atmosphere (temperature, humidity, and air) and devices to avoid entrance of pathogens. The age of the studied chickens was <45 d. Chicken samples were taken by an expert veterinarian. All sampling was done with cotton-tipped swabs that were placed in Amies Rayon (Deltalab, Barcelona, Spain) and were subsequently stored at 4°C and transported directly to the laboratory.
Microbiological analysis
All samples were cultured in chromogenic media (chromID TM ESBL, bioMérieux, Marcy l'Etoile, France). The culture was incubated at 37°C during 24–48 h. The suspect strains of E. coli present a spontaneous coloration (rose to violet) due to the production of β-glucuronidase. Oxidase (Fluka Analytical, Buchs, Switzerland) and indole (Panreac, Barcelona, Spain) tests were performed on all suspect strains. These isolates were also identified by the Vitek 2 system (bioMérieux) and ESBL production was confirmed by the disc diffusion method (Oxoid, Hants, UK) on Mueller Hinton agar according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI, 2012).
ESBL producers were confirmed by the double-disk synergy test using both cefotaxime and ceftazidime alone and in combination with clavulanic disks. E. coli ATCC 25922 was used as reference strain.
Antimicrobial susceptibility was determined with AST-N243 by Vitek 2 system (bioMérieux). The isolates were tested for susceptibility to aminoglycosides: gentamicin, amikacin; ß-lactams: penicillin, ampicillin, cefuroxime, cefuroxime axetil, cefoxitin, cefotaxime (CTX), ceftazidime, cefepime; β-lactams/β-lactamase inhibitor combinations: amoxicillin–clavulanic acid, piperacillin–tazobactam; carbapenems: ertapenem, imipenem; quinolone: nalidixic acid, ciprofloxacin; bacteriostatics: trimethoprim–sulfamethoxazole; glycylcycline: tigecycline. The breakpoints used were the same established by the Clinical and Laboratory Standards Institute Guidelines (CLSI, 2012). Pulsed-field gel electrophoresis (PFGE) was performed on ESBL–E. coli isolates with XbaI (Promega) to define their epidemiological relation, as described previously (Fernández-Baca et al., 2001).
The presence of CTX-M-type ESBL was detected by real-time polymerase chain reaction (RT-PCR) using the Kit RealCycler BLACTX-U (RealCycler, Progenie Molecular, Valencia, Spain) to detect CTX-M-1 Group (includes: CTX-M-1, CTX-M-3, CTX-M-10, CTX-M-12, CTX-M-15, and FEC-1, CTX-M-22, CTX-M-23, CTX-M-28); CTX-M-2 Group (includes: CTX-M-2, CTX-M-4, CTX-M-4L, CTX-M-5, CTX-M-6, CTX-M-7, CTX-M-20, and Toho-1) and CTX-M-9 Group (includes CTX-M-9, CTX-M-13, CTX-M-14, CTX-M-16, CTX-M-17, CTX-M-19, CTX-M-21, CTX-M-27, Toho-2, and CTX-M-24). Nucleic acid purification was performed with Maxwell TM 16 viral total nucleic acid purification kit (Promega).
Results
In our study, of 260 rectal swabs collected from farms, 237 (91.1%) were ESBL–E. coli and 3 (1.2%) were ESBL–Klebsiella pneumoniae.
Eight susceptibility patterns were obtained for all ESBL–E. coli isolates (Table 1). Analysis of the susceptibility status of isolates revealed that 179 isolates were resistant to 2 antibiotic groups, 46 isolates were resistant to 3 antibiotic groups, and 12 isolates were resistant to 4 or more antibiotic groups. The major resistances were to nalidixic acid (92.4%), ciprofloxacin (65.5%), and trimethoprim–sulfamethoxazole (24.1%). All strains were susceptible to amikacin and tigecycline.
AM, ampicillin; CXM, cefuroxime; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; GM, gentamicin; NA, nalidixic acid; CIP, ciprofloxacin; SXT, trimethoprim/sulfamethoxazole.
PFGE was performed in all isolates, and a great genetic variability was found according to criteria by Tenover (Tenover et al., 1995); 44 different band patterns were detected in 196 strains and 41 (17.3%) were nontypeable strains. RT-PCR of gene bla CTX-M was performed in these characterized strains. The bla CTX-M genes were detected in 116 (59.1%) ESBL–E. coli: 87 (75.0%) CTX-M-9 group, 16 (13.8%) CTX-M-1 group, 11 (9.5.0%) CTX-M-2 group, and in 2 (1.7%) cases CTX-M-1 and CTX-M-2 groups were detected in the same strain (Table 2). All these groups were detected in every farm screened.
E. coli bla CTX-M strains showed resistance to more than two non-β-lactam antibiotics in 7.8% of the cases. Seven different susceptibility patterns were obtained for all of the isolates of E. coli bla CTX-M analyzed (Table 3). The major resistances were to nalidixic acid (89.7%), ciprofloxacin (69.8%), and trimethoprim-sulfamethoxazole (31.9%).
AM, ampicillin; CXM, cefuroxime; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; GM, gentamicin; NA, nalidixic acid; CIP, ciprofloxacin; SXT, trimethoprim/sulfamethoxazole.
Discussion
In our study, a high number of chicken samples with ESBL–E. coli was found, all farms being positive for those strains because the chicken farms were intensive facilities that are more densely populated, and the reason could be the geographical characteristics of Tenerife: a steep island, with few plains and large areas considered nature reserve, making difficult the development of expansive facilities, instead of intensive facilities, were animals are more crammed.
The prevalence was higher than that of other studies conducted in different countries (Briñas et al., 2005; Smet et al., 2008; Costa et al., 2009; Yuan et al., 2009; Li et al., 2010; Randall et al., 2012; Geser et al., 2012), and there was a high prevalence of E. coli bla CTX-M strains. The majority group was the CTX-M-9 group, followed by CTX-M-1 group, similar to other studies such as Smet et al. (2008), who reported that CTX-M-carrying cloacal E. coli samples in Belgian broiler farms were the following: CTX-M-1 (27.4%), CTX-M-2 (7.8%), CTX-M-14 (5.9%), and CTX-M-15 (2%), or Li et al. (2010), who found in chickens that of 56 ESBL strain producers, 54 isolates contained CTX-M β-lactamase, concluding that the chickens had turned into a significant reservoir of CTX-M in China. Geser et al. (2012), in an ESBL-Enterobacteriaceae prevalence study conducted in Switzerland, found 63.4% prevalence in chickens and the PCR detection revealed that 71% were CTX-M-1. Also, Leverstein-Van Hall et al. (2011) in Holland carried out a study to relate strains found in patients to those of healthy chickens and chicken meat. They thought that the types of strains in the chickens were CTX-M-1 in 49% of the samples, TEM-52 in 29%, SHV-2 in 11%, CTX-M-2 in 9% and TEM-20 in 3%. These same β-lactamases were found in chicken meat and in humans but in different percentages. In Great Britain, Randall et al. (2011) reported that CTX-M-carrying E. coli were isolated from 54.5% of the broiler abattoirs and from 3.6% of individual broiler cecal samples and were CTX-M sequence types 1 (mainly), 3, and 15. Mnif et al. (2012) did a study in a poultry farm in Tunis where they found that all ESBLs belonged to CTX-M group 1: 39 CTX-M-1 and 4 CTX-M-15. There are some studies of ESBL type in E. coli isolated in chicken meat. Accordingly, Egea et al. (2012) conducted a study of prevalence and characteristics of ESBL–E. coli isolated in chicken meat in Spain, during two periods of time: 2007 and 2010. They found a high prevalence of E. coli. The most frequent types of ESBL were SHV-12, though there was a prevalence decrease from 2007 to 2010. Nevertheless, the CTX-M type had increased in the studied period of time. In chickens, they found a major prevalence of CTX-M-1. As a limitation of the study, we must emphasize that we only focused on the detection of CTX-M group strains, not studying the type of resistance of the rest of ESBL–E. coli isolates like other authors.
In our study, 47.3% of ESBL–E. coli were resistant to β-lactam antibiotics, except carbapenems, and quinolones (nalidixic acid and ciprofloxacin). However, Geser et al. (2012) found in 62 ESBL–E. coli strains isolated from rectal samples of chicken that 51 were resistant to tetracycline (82.3%), 37 to trimethoprim–sulfamethoxazole (59.7%), 24 to nalidixic acid (38.7%), 6 to aminoglycoside (9.7%), 2 to chloramphenicol (3.2%), 4 to ciprofloxacin (6.5%), and none of the analyzed strains was resistant to imipenem. We found higher percentages of resistance in the case of ciprofloxacin and nalidixic acid and lower in gentamicin and trimethoprim–sulfamethoxazole. The reason for the high percentages could be the antibiotics that were prescribed in case of infection, such as quinolone and fluoroquinolone. In China, Li et al. (2010) found that of 56 ESBL strains, 54 isolates contained CTX-M type β-lactamase of different groups. They thought that more than three fourths of ESBL strains in chicken were also resistant to ciprofloxacin, in a slightly higher percentage to the one we obtained (65.4%).
Johnson et al. (2006) in the United States investigated multidrug-resistant E. coli isolated in humans and in poultry farms (chickens and turkeys), in Minnesota and Wisconsin. They demonstrated by phylogenetic studies and virulence markers that some resistant E. coli isolated in humans could have their origin in birds, whereas the resistant E. coli strains isolated in birds were obtained from susceptible strains that had acquired resistance genes.
E. coli of animal origin may act as a donor of antimicrobial resistance genes for other pathogenic E. coli. Thus, the intensive use of antimicrobial agents in food animals may add to the burden of antimicrobial resistance in humans. Bacteria from the animal reservoir that carry resistance to antimicrobial agents regarded as highly or critically important in human therapy (e.g., aminoglycosides, fluoroquinolones, and third- and fourth-generation cephalosporins) are of especially great concern.
In our study, regarding the susceptibility patterns of E. coli bla CTX-M strains, 7.8% showed resistance to more than 2 non-β-lactam antibiotics and 89.7% of strains were resistant to fluoroquinolones. That high resistance of the strains was found by Mnif et al. (2012), who observed that all the 67 CTX-M resistant E. coli isolates from healthy broilers were multidrug resistant and showed resistance to more than 2 non-β-lactam antibiotics, including tetracycline (94%), nalidixic acid (89.5%), norfloxacin (71.6%), trimethoprim–sulfamethoxazole (73.1%), gentamicin (6%), and amikacin (6%). The great resistance to quinolone in E. coli CTX-M strains has been detected in human strains, particularly extraintestinal pathogenic E. coli isolates (Dalhoff, 2012; Pitout, 2012). Johnson et al. (2006, 2007) compared ciprofloxacin-resistant E. coli of human and poultry origin. The authors concluded that ciprofloxacin-resistant E. coli may arise in the intestine of poultry from susceptible E. coli ancestors and be subsequently transmitted to humans via the food supply.
Conclusions
In our area, the prevalence of CTX-M-type ESBLs in E. coli isolated from rectal samples of chicken is even higher than those found in other countries.
The impact of food animals as a possible reservoir for ESBL–E. coli, especially CTX-M-type ESBL, and the dissemination of such strains into the food production chain need to be assessed.
Footnotes
Disclosure Statement
No competing financial interests exist.
