Abstract
The bacterial contamination of raw and processed meat products with resistant pathogens was studied. The raw samples included sheep (40), goat (40), pork (120), beef (80), and chicken (19) meat, and the processed samples included turkey filets (33), salami (8), readymade mincemeat (16), stuffing (22), and roast-beef (50). The samples were collected from retail shops in Northwestern Greece over a period of 3 years. The isolated pathogens were evaluated for susceptibilities to 19 antimicrobial agents used in humans. Out of 428 samples, 157 strains of Escherichia coli, 25 of Yersinia enterocolitica, 57 of Staphylococcus aureus, 57 of Enterococcus spp., 4 of Salmonella spp., and 3 of Campylobacter jejuni were isolated. Among the isolates 14.6% of the E. coli, 10.5% of S. aureus, 4% of Y. enterocolitica, 25% of Salmonella spp., and 42.1% of Enterococcus spp. were susceptible to antibiotics. E. coli from chicken exhibited high rates of resistance to ciprofloxacin (62.5%) followed by lamb/goat (10.9%), pork (15.7%), and beef (27.9%) meat. Resistance to nitrofurantoin dominated in the lamb/goat isolates (60%). Resistance to tetracycline predominated in pork (68.2%) and chicken (62.5%), and resistance to aminoglycosides dominated in lamb/goat meat isolates. S. aureus resistance to clindamycin predominated in lamb/goat isolates (50%), whereas resistance to ciprofloxacin predominated in the pork strains, but no resistance to methicillin was observed. Of the enterococci isolates 21.1% were resistant to vancomycin. High resistance to ampicillin (96%) was observed in Y. enterocolitica and all of the C. jejuni isolates were resistant to ampicillin, cephalothin, and cefuroxime. These results indicate that meat can be a source of resistant bacteria, which could potentially be spread to the community through the food chain.
Introduction
Although there is an ongoing controversy about the causes of the development of resistance, there is growing evidence based on research data suggesting the potential link between the antimicrobial resistance and the veterinary practices (Engberg et al., 2001; Delsol et al., 2004; Mathew et al., 2005; Funk et al., 2006; Rosengren et al., 2009). In addition, there are reports by international public health authorities (EMEA, 2009a,b; WHO, 2009; EFSA, 2010) dealing with the potential link and risk between the overuse of antibiotics in veterinary practice and the emergence of resistant human pathogens. It is generally agreed that the administration of antimicrobials to food animals for growth promotion, prophylaxis, and treatment can lead either to the selection of resistant bacteria, which can be transmitted through the food chain (Witte, 2000; Mayrhofer et al., 2004), or to the horizontal transfer of resistance genes to human pathogenic or commensal microflora (McDermott et al., 2002). The antimicrobials administered to food-producing animals are frequently the same or belong to the same classes as those used in human medicine (Aarestrup et al., 2008). Moreover, bacteria may be resistant to several classes of antimicrobials, whereby the use of one class of antimicrobial may result in the selection of resistance against another, unrelated class (coresistance) (EFSA, 2010).
Meat is a major source of foodborne infections and resistant bacteria in meat products have been implicated in outbreaks and sporadic cases (Mølbak et al., 1999; Smith et al., 1999; Marano et al., 2000; Helms et al., 2002a). In Greece, there is a significant increase in intensive animal production, regretfully coupled with inappropriate prescription of antimicrobials. Frequently, antibiotics are administered without professional veterinary consultation and, as a consequence, inappropriate doses and combinations of drugs are used in animals; even the official withdrawal periods for veterinary medicinal products are not followed by some farmers (Athanasiou et al., 2008a). Empiric treatment based on clinical findings, rather than isolation of the causative pathogen, is the common practice, while the submission of clinical specimens from sick animals is uncommon due to the associated cost and the limited number of veterinary diagnostic laboratories. Residual antimicrobials were detected in foods from productive animals in Greece, despite the ban of certain antibiotics by the European Union (EU) (O'Keeffe et al., 2004; Athanasiou et al., 2008b).
This study aimed to assess the prevalence of antimicrobial resistance of major foodborne pathogens (Salmonella, Campylobacter, E. coli, Yersinia, Enterococcus, and S. aureus), isolated from raw and processed meat produced from food animals in Greece. To our knowledge, there are limited and fragmented published research data pertinent to the antimicrobial resistance for some of the considered food pathogens in the specified food types. A search of the existing literature retrieved only four relevant studies reporting antibiotic resistance of Salmonella and Listeria from sausages (Abrahim et al., 1998); resistance of Salmonellae from chicken (Arvanitidou et al., 1998); resistance of Yersinia spp. from pigs, chickens, sheep, and cows at slaughter (Kechagia et al., 2007); and the E. coli resistance profile for isolates from poultry (Vasilakopoulou et al., 2009). According to the official country report on Zoonosis Monitoring, including information on foodborne outbreaks and antimicrobial resistance in zoonotic agents, no official national program was in force in Greece for antimicrobial resistance monitoring by the year 2007 (EFSA, 2007).
Materials and Methods
Sample collection
Over a period of 3 years (2004–2007), 428 samples of raw and processed meat products were collected. The raw meat samples included sheep (40), goat (40), pork (120), beef (80), and chicken (19) meat (Table 1). The processed samples included turkey filets (33), salami (8), readymade mincemeat (16), stuffing (22), and roast-beef (50). The samples were collected from retail shops in the proximity of the laboratory (Epirus prefecture-Northwestern Greece). Only retail shops selling meat products produced in Greece were selected. Each retail shop was visited at least once a month and the sampling was performed every month throughout each year of the study. The samples were transported to the laboratory into portable coolers and were processed within 4 h after collection.
B, bovine; C, chicken; CI, confidence interval (95% probability); CO, caprine/ovine; n, number of isolates; N 0, number of samples; N 1, number of positive samples; P, pork; R = 1, resistant to 1 antimicrobial; R = 2, resistant to 2 antimicrobials; R ≥3, resistant to ≥3 antimicrobials.
Sample examination
All samples were examined for the presence of E. coli, S. aureus, Enterococcus spp., Salmonella spp., Yersinia spp., and Campylobacter spp. From each sample, a subsample of 25 g was aseptically removed, weighed on an analytical balance (Kern KB), and homogenized in a stomacher for 2 min (BagMixer; Interscience) with the addition of 225 mL Buffered Peptone Water (Merck KGaA). E. coli was isolated by plating on Violet Red Bile agar (Merck KGaA) and Chromocult agar (Merck KGaA). The inoculated media were incubated aerobically at 37°C for 24 h. The Salmonella strains were isolated by pre-enrichment in Buffered Peptone Water, enrichment in Rappaport-Vassiliadis broth (Merck KGaA), and Selenite cystine broth (Merck KGaA) and plated on Salmonella-Shiggella agar (Oxoid Ltd.) and Xylose Lysine Desoxycholate agar (Merck KGaA). S. aureus was isolated by plating onto Baird Parker agar (Scharlau Chemie S.A.). Campylobacter spp. were isolated according to the ISO 10272-1 method (ISO, 2006), and Yersinia spp., according to the ISO method 10273 (ISO, 2003).
Characteristic colonies were picked from the selective media, and then they were pure cultured and identified to the genus level using the API20E system (BioMérieux) for Enterobacteriaceae, the API 20 STAPH system (BioMérieux) and the catalase and coagulase test (Staphytect plus; Oxoid Ltd.) for Staphylococci, the API20 STREP system (BioMérieux), and the catalase test for enterococci. Salmonella isolates were identified by the API 20E system and also by serotyping.
Susceptibility tests
Susceptibility to penicillin, ampicillin, ampicillin/sulbactam, amoxicillin/clavulanic acid, methicillin, carbenicillin, ticarcillin, ticarcillin/clavulanic acid, cefuroxime, ceftriaxone, imipenem, ciprofloxacin, nitrofurantoin, gentamicin, amikacin, tetracycline, erythromycin, clindamycin, and vancomycin was tested, using the disc diffusion method according to the recommendations of the Clinical and Laboratory Standards Institute (CLSI, 2007). One or two colonies were picked from an overnight culture and were suspended in 10 mL of sterile normal saline. The turbidity of the suspension was adjusted to 0.5 of the McFarland scale; subsequently, it was streaked over Mueller Hinton agar plates (51075; Biomerieux). The antibiotic discs were applied with a dispenser and the plates were incubated overnight at 37°C and for Yersinia spp. at 30°C. For Campylobacter spp., the method described by Hein et al. (2003) was followed. All antibiotic susceptibility discs were supplied by Oxoid. E. coli ATCC 25922 and S. aureus ATCC 25923 were used as quality control strains. The inhibition zones were measured to the nearest millimeter using an electronic digital caliper (Powerfix, Model Z22855) and were characterized as sensitive, intermediate resistant, and resistant according to CLSI breakpoints (CLSI, 2007). For Campylobacter spp. isolates, the inhibition zones proposed by the CLSI for Enterobacteriaceae were used, since there are no currently available breakpoints specific for Campylobacter. This may lead to some interpretation problems, which also have been faced by other researchers, but the majority is following the CLSI criteria for Enterobacteriaceae (Smith et al., 1999; Ge et al., 2002; Luangtongkum et al., 2007; Ioannidis et al., 2009; McGill et al., 2009; Rosengren et al., 2009). The antimicrobials used for the susceptibility tests are listed in Tables 2 and 3; the antimicrobials used in the relevant food animal species, from which the meat samples were collected, are presented in Table 4.
AK 30, amikacin (30 μg); AMC 30, amoxicillin/clavulanic acid (20/10 μg); AMP 10, ampicillin (10 μg); CAR 100, carbenicillin (0.1 mg); CIP 5, ciprofloxacin (5 μg); CN 10, gentamicin (10 μg); CRO 30, ceftriaxone (30 μg); CXM 30, cefuroxime (30 μg); F 300, nitrofurantoin (300 μg); IPM 10, imipenem (10 μg); N, number of isolates; n, number of resistant isolates; SAM 20, ampicillin/sulbactam (10/10 μg); TE 30, tetracycline (30 μg); TIC 75, ticarcillin (75 μg); TIM 85, ticarcillin/clavulanic acid (75/10 μg).
N, number of isolates; n, number of resistant isolates.
Also used in sheep without approval by the National Drug Organization.
The shaded X indicates the use of the specific antimicrobial per animal category.
Results
During the study, from the raw meat samples 157 E. coli, 57 S. aureus, 25 Yersinia enterocolitica, 27 Enterococcus faecalis, 18 Enterococcus faecium, 12 Enterococcus durans, 3 Campylobacter jejuni, and 4 Salmonella strains were recovered (Table 1). The Salmonella strains were identified by the National Salmonella Reference Center as Salmonella blockley, Salmonella typhimurium, Salmonella infantis, and Salmonella enteritidis. No pathogens were isolated from the thermally processed products. The most common contaminant found in the raw meat samples was E. coli, isolated from all meat types. S. aureus was isolated from beef, pork, and lamb/goat meat; Y. enterocolitica only from pork; and Salmonella and C. jejuni only from chicken. The prevalence of the bacterial pathogens found in the raw meat samples is presented in Table 1 and the isolate resistances are shown in Tables 1 –3. The frequency of the resistance profiles within the bacterial species and sample types are presented in Table 5.
N, number of isolates from each sample type; n, number of isolates possessing each antimicrobial resistance profile; KF, cephalothin; E, erythromycin; P, penicillin G; DA, clindamycin; VA, vancomycin.
The Salmonella strains displayed resistance to ampicillin (75%), amoxicillin/clavulanic acid (50%), ticarcillin/clavulanic acid (50%), ciprofloxacin (50%), tetracycline (50%), and amikacin (25%). C. jejuni strains were all resistant to ampicillin, cephalothin, and cefuroxime, while 33.3% of the strains were resistant to the combination of ampicillin/sulbactam; nevertheless, all the C. jejuni strains were susceptible to amoxicillin/clavulanic acid, erythromycin, imipenem, clindamycin, ciprofloxacin, gentamicin, amikacin, and tetracycline.
Discussion
Ever since drug resistance in zoonotic pathogens has been interrelated with therapeutic interventions in humans, the antimicrobial resistance of foodborne pathogens has become a major public health issue (Swartz, 2002; WHO 2002a; Mathew et al., 2007; EFSA, 2008; SANCO, 2009; Webster, 2009; EFSA, 2010). The indiscriminate use of antibiotics in medical (Plachouras et al., 2010) and veterinary practices (Aarestrup et al., 2008) has been incriminated in increased antimicrobial resistance (WHO, 2002b). In our study the resistance of Salmonella, Campylobacter, E. coli, Yersinia, Enterococcus, and S. aureus isolates from beef, pork, lamb/goat, and chicken meat was examined.
The Salmonella isolates from chicken meat were resistant to broad-spectrum penicillins, tetracyclines, aminoglycosides, and quinolones, antimicrobials that are used in veterinary medicine in Greece (Table 4) and elsewhere (Gundogan et al., 2005); quinolones, specifically enrofloxacin, are widely used for preventive purposes in poultry (Teshager et al., 2000). Also, the Salmonella isolates were resistant to β-lactams combinations with clavulanic acid, a class of antimicrobials not approved for use in poultry in Greece (Table 4). Three out of four (75%) Salmonella strains were multiresistant, each with a different resistance profile (Table 5). Multiresistant Salmonella isolates have been reported in other countries as well (Kasimoglu-Dogru and Ayaz, 2010). The available published data on Salmonella resistance in foods in Greece is very limited. In two studies conducted in northern Greece on the antimicrobial resistance of salmonellae from chicken and pork (Tables 5 and 6), 58.1% of the chicken isolates were resistant to at least one antimicrobial with 18 different resistance profiles (Arvanitidou et al., 1998), while 15.4% of the pork sausage isolates were resistant to 3 antimicrobials with one resistance profile (Abrahim et al., 1998). According to a recent EFSA report (EFSA, 2010), resistances to tetracycline, ampicillin, sulfonamide, ciprofloxacin, and nalidixic acid were commonly reported among Salmonella spp. isolates from broiler meat, and the resistance levels in 2007 were found up to 37% in the reporting Member States.
AMP/Sulb, ampicillin-sulbactam; AZT, aztreonam; CEF, cefotaxime; CFA, cephalothin; CFI, cefixime; CFL, cefaclor; CFM, cefamandole; CFP, cefepime; CFT, ceftriaxone; CFX, cefoxitin; CFZ, cefprozil; CHL, chloramphenicol; COX, cotrimoxazole; ERY, erythromycin; GEN, gentamycin; GLA, glindamycin; NA, nalidixic acid; NOR, norfloxacin; OX, oxacillin; SPM, spectinomycin; STM, streptomycin; TET, tetracycline.
The Campylobacter strains (isolated from chicken meat only) were resistant to broad-spectrum penicillins, tetracyclines, and aminoglycosides, antimicrobials that are used in veterinary medicine in Greece (Table 4). The C. jejuni isolates were resistant to cephalosporins, which are not approved for use in poultry in Greece, but they were susceptible to quinolones, a class of antimicrobials that are extensively used for preventive purposes in poultry (specifically enrofloxacin) in Greece (Table 4). In the relevant literature the extensive use of quinolones in veterinary medicine is incriminated in the rising quinolone resistance of Campylobacters from poultry (Sáenz et al., 2001; Praakle-Amin et al., 2007; Van et al., 2007; Nannapaneni et al., 2009; Soltan-Dallal et al., 2010). Published research data on Campylobacter antimicrobial profiles from food isolates in Greece is missing. In other European countries, the reported resistance levels to tetracycline, ciprofloxacin, nalidixic acid, and erythromycin were 37%, 39%, 36%, and 3%, respectively (EFSA, 2010).
The E. coli strains isolated from all meat types were resistant to broad-spectrum penicillins and cephalosporins, which are approved for use in cattle and swine in Greece (Table 4). Cephalosporin resistances in E. coli isolates from pork, beef, and chicken have been reported in other countries (Hornish and Kotarski, 2002; Forward et al., 2004; EFSA, 2009b). E. coli isolates resistant to tetracyclines and aminoglycosides were identified; both antimicrobial classes are widely used in Greece for therapeutic or prophylactic purposes in productive animals. Due to their low toxicity, tetracyclines have been extensively used as growth promoters in farm animals (Chopra and Roberts, 2001), and tetracycline resistance has been expansively studied and shown to be widespread in food animals and products (Papadopoulou et al., 1997; Schlegelova et al., 2004; Roberts, 2005; Kérouanton et al., 2007; Meyer et al., 2007; Van et al., 2007).
In the present study, resistance to nitrofurantoin, an antimicrobial banned for veterinary use, was detected in 73 (46.5%) E. coli isolates from bovine, caprine/ovine, and swine meat. In a previous study of ours (Papadopoulou et al., 1996) conducted in the same geographic area in hen eggs, E. coli strains resistant to broad-spectrum penicillins, cephalosporins, and macrolides were reported, while in another study in southern Greece (Vasilakopoulou et al., 2009) high rates of single antimicrobial resistances to broad-spectrum penicillins, cephalosporins, tetracyclines, and chloramphenicol (banned for veterinary use) were reported in chicken (Table 6). Relevant studies in other countries report that 40%–70% of veterinary E. coli isolates were resistant to penicillin, while 15%–50% of E. coli strains isolated from farm animals, particularly from swine, were resistant to ampicillin (Sáenz et al., 2001; Van et al., 2007; Knezevic and Petrovic, 2008; Miranda et al., 2008). Low rates (8%–11.8%) of resistance to imipenem, a β-lactam, strictly reserved for human use, were found in beef, lamb/goat, and pork samples (Table 2). The observed resistance to imipenem may be due to cross-resistance with other β-lactams (Pagani et al., 1990). Relevant studies have reported either no resistance or very low resistance rates (Teshager et al., 2000; Sáenz et al., 2001; Meyer et al., 2007; Lampang et al., 2008).
According to the 2010 EFSA report, among E. coli isolates from fowl (Gallus gallus), the occurrence of resistance to tetracycline, ampicillin, and sulfonamide ranged between 7% and 44% in the reporting countries. Resistance levels to ciprofloxacin and nalidixic acid were between 13% and 50%, ceftiofur resistance ranged between 0% and 3%, and cefotaxime resistance was 4%. Similarly, among E. coli isolates from pigs, the occurrence of resistance to tetracycline, ampicillin, and sulfonamide ranged from 17% to 68%. The occurrence of ceftiofur and cefotaxime resistance among E. coli isolates from cattle ranged from 0% to 2% (EFSA, 2010).
Y. enterocolitica was isolated from pork only and the isolates were resistant to broad-spectrum penicillins, cephalosporins, and aminoglycosides, antimicrobial classes that are approved for veterinary use in pig production in Greece (Table 4). The aminoglycosides apramycin and gentamicin have been widely used in pig production in several European countries, including Greece, since the early 1980s (Jensen et al., 2006; EFSA, 2009b). Gentamicin was included in our antimicrobials list, but apramycin was not, because it is authorized for animal use only and only antibiotics used in human medicine were selected for the present study. There are studies reporting that exposure to apramycin can cause the resistance to gentamicin (Jensen et al., 2006; Millar, 2007). Also, nine (36%) Y. enterocolitica pork isolates were found resistant to (the banned) nitrofurantoin. Researchers from southern Greece have reported similar resistances in various food animals, including resistance to nitrofurantoin (Table 6) (Kechagia et al., 2007). Similar findings were reported from other countries (Schlegelova et al., 2004; Bucher et al., 2008). Although since the early 1990s the EU, United States, and Australia prohibited the use of nitrofurans in food-producing animals, residues of nitrofuran metabolites and resistances to nitrofurans continue to be reported in various countries, including Greece (Rahman Khan and Malik, 2001; O'Keeffe et al., 2004).
Enterococci were isolated from pork samples only and the isolates were resistant to broad-spectrum penicillins, cephalosporins, tetracyclines, macrolides, and vancomycin (Tables 1 and 3). Five multiresistant isolates with three different resistance profiles were identified (Table 5). Studies on vancomycin-resistant enterococci in other countries report resistance rates ranging from 0% to 35% (Schlegelova et al., 2004; Messi et al., 2006; Valenzuela et al., 2008). At the European level the majority of data on antimicrobial resistance in E. faecium and E. faecalis were reported for 2007. In particular, tetracycline (31%–85%), streptomycin (16%–59%), and erythromycin resistance (23%–48%) occurred frequently in E. faecium and E. faecalis isolates from poultry, pigs, and cattle. Also, E. faecium isolates resistant to vancomycin were reported from poultry, pigs, and cattle (1%–2%), even though the use of avoparcin (a similar glycopeptide antibiotic) as a growth promoter was banned in the EU in 1997 (EFSA, 2010)
Fifty-one S. aureus isolates from beef, pork, and lamb/goat meat were resistant to one or more antimicrobials from the classes of penicillins (narrow and broad spectrum), macrolides (erythromycin), lincosamides (clindamycin), tetracyclines, and cephalosporins (ciprofloxacin), which are used in farm animals for therapeutic purposes. Lincomycin is extensively used in small ruminants for the treatment of foot rot, contagious agalactia, and mycoplasmal atypical pneumonia, which are enzootic in Greece. In a previous study conducted in the same geographic area (Papadopoulou et al., 1996), resistant S. aureus strains were detected in hen eggs, but the resistance rates to the same antimicrobial classes were lower (Table 6).
Summarizing the data available for zoonotic pathogens from farm animals and meat, resistance to antimicrobials is commonly found among the isolates (Wegener et al., 1999; WHO 2002a, 2002b; Mathew et al., 2007; USDA 2007; EFSA 2008, 2009b; EMEA, 2009a,b; Webster, 2009). The emergence of antimicrobial resistance among foodborne pathogens has been recognized for four decades now and has induced an EU-wide ban on the use of antibiotics as growth promoters from January 2006 (Regulation EC, 2003). Although a total ban might not be the solution, since it could lead to increased use of antibiotics later in an animal's life, it might assist to containment of resistance by shortening animal exposure to antibiotics, and thereby decreasing the quantity of antibiotics used (AAM, 2009). This study investigated the prevalence of antimicrobial resistance of foodborne pathogens in Northwestern Greece. It identified resistant pathogens in beef, pork, lamb/goat, and chicken meat, but could not conclusively state that only the antimicrobial use caused the observed resistances. The reasons behind the establishment and spread of antibiotic resistance are complex, mostly multifactorial, and mostly unknown. Continued research bridging medical, veterinary, and other biomedical disciplines may generate sufficient knowledge about this old problem.
Footnotes
Disclosure Statement
No competing financial interests exist.
