Prevalence and antimicrobial resistance of Enterococcus species of food animal origin from Beijing and Shandong Province, China

Authors

  • Y. Liu,

    1. Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing, China
    Search for more papers by this author
  • K. Liu,

    1. Department of Pharmaceuticals, China Institute of Veterinary Drugs Control, Beijing, China
    Search for more papers by this author
  • J. Lai,

    1. Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing, China
    Search for more papers by this author
  • C. Wu,

    1. Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing, China
    Search for more papers by this author
  • J. Shen,

    1. Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Y. Wang

    Corresponding author
    • Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing, China
    Search for more papers by this author

Correspondence

Yang Wang, Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety, Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing 100193, China.

E-mail: wangyang@cau.edu.cn

Abstract

Aims

To evaluate the prevalence and antimicrobial resistance of Enterococcus species from chickens and pigs in Beijing and Shandong Province, China.

Methods and Results

Swab samples were collected from four farms in Beijing and two in Shandong Province in 2009 and tested for Enterococcus. Minimum inhibitory concentrations of antimicrobial agents were determined using broth microdilution or agar screening methods. A total of 453 Enterococcus isolates were recovered, belonging to six different Enterococcus species. All isolates were sensitive to vancomycin. Resistance to tetracycline (92·5%), amikacin (89·4%), erythromycin (72·8%) and rifampin (58·1%), and high-level streptomycin resistance (HLSR, 50·3%) were prevalent, while resistance to penicillins (7·9% to penicillin and 4·2% to ampicillin) was rare. The resistance rates to phenicols (chloramphenicol and florfenicol) and enrofloxacin, and high-level gentamicin resistance (HLGR) were approximately 30%. The vast majority of the Enterococcus isolates were classified as multidrug-resistant organisms.

Conclusions

Resistance of Enterococcus sp. to most antimicrobials was more prevalent in China than in European or other Asian countries.

Significance and Impact of the study

Our findings reveal a high level of antimicrobial resistance in Enterococcus isolates from food animals in China and underline the need for prudent use of antibiotics in chicken and pig production to minimize the spread of antibiotic-resistant enterococci.

Introduction

Enterococci are Gram-positive, facultative anaerobes that are present as part of the natural microflora of the intestinal tracts of animals and humans. For many years, Enterococcus species were believed to be harmless to humans and were considered medically unimportant (Fisher and Phillips 2009). Because they produce bacteriocins, Enterococcus species have been used widely over the last decade in the food industry as probiotics, as starter cultures for fermented food products (Foulquie Moreno et al. 2006; Ogier and Serror 2008) or as a preservation method to control emergent pathogenic bacteria (Callwaert et al. 2000; Franz et al. 2007). However, recently, enterococci have rapidly emerged as leading nosocomial pathogens that frequently present a considerable therapeutic challenge. The nature of the challenge rests in the mechanisms by which enterococci cause disease, including infections of the urinary tract and bile ducts, wound infections and life-threatening infections such as bacteremia or endocarditis. In addition, there is increasing resistance of enterococci to most of the antimicrobial agents currently approved to treat infections, including the emergence of vancomycin-resistant Enterococcus (VRE).

Enterococci are intrinsically resistant to a number of antimicrobial agents, including β-lactams (particularly cephalosporins and penicillinase-resistant penicillins) and clindamycin, and low concentrations of aminoglycosides. Therefore, treatment of enterococcal infections may be difficult. Furthermore, enterococci can acquire resistance to all currently available antibiotics, either by mutation or by receipt of foreign genetic material through the transfer of plasmids and transposons (Clewell 1990). Many studies have addressed the importance of enterococci as a reservoir of antibiotic resistance genes in animals and the environment (Kayser 2003; Klein 2003; Kuhn et al. 2003). Because of the risk of potentially harmful enterococcal strains being transmitted through the food chain and the contribution of enterococci to the spread of antimicrobial resistance, it is important to evaluate the prevalence and antibiotic resistance of strains found in food animals, which play a relevant role in the transmission of enterococci. Many studies have reported the prevalence of antimicrobial-resistant enterococci in animal reservoirs in different countries (Aarestrup et al. 2000, 2002; Butaye et al. 2001; van den Bogaard et al. 2002; Hayes et al. 2004; Hwang et al. 2009). However, less information is available about enterococci from food animals in China (Zou et al. 2011). In this study, enterococci from food animal samples (chickens and pigs) collected from Beijing and Shandong Province were isolated, identified and screened for their resistance to 12 antibiotics, which were chosen for their importance in human and veterinary medicine as well as in animal husbandry.

Materials and methods

Sample collection, isolation and identification of Enterococcus species

A total of 1034 faecal samples derived from food animals were collected from Beijing and Shandong Province during 2009. Of these samples, 502 were obtained from two intensive chicken farms located in the central (Xintai) and south-east (Linyi) regions of Shandong Province. Another 532 samples, including pigs (n = 238, 70 days of age) and chickens (n = 294, 20–25 days of age), were collected from four free-range mixed poultry-pig farms located in the north (Huairou), north-west (Changpin), north-east (Miyun) and south-east (Tongzhou) regions of Beijing (Table 1). All farms were sampled once. Animal sampling practices complied with the principles of the Beijing Municipality Review of Welfare and Ethics of Laboratory Animals (BAOLA 2005). Sterile cotton swabs were used to collect faecal samples from the rectums of pigs and the cloacae of chickens and were immediately placed into sterile collection containers and transported on ice to the laboratory within 6 h of collection. Swabs were then transferred into 1 ml nutrient broth supplemented with 6·5% NaCl (Luqiao, Beijing, China) and incubated at 45 °C for 24 h for primary isolation. A loop of the resulting culture was streaked onto Slanetz and Bartley medium (Oxoid, Basingstoke, UK) and incubated at 37 °C for 24 h. One putative Enterococcus colony was isolated from each swab sample. Isolates were stored in 20% glycerol at −80 °C. Each swab corresponded to an individual animal. The following tests were conducted for putative identification of the isolates: observation of colony characteristics, determination of cell morphology by Gram staining and assessment of growth and aesculin hydrolysis on bile-aesculin agar (Luqiao, China), which is an Enterococcus medium base supplemented with ammonium ferric citrate.

Table 1. Enterococcus isolates obtained from chickens and pigs
RegionAreaOriginNumber (%)
SamplesEnterococcus spp. Enterococcus faecalis Enterococcus faecium Enterococcus casseliflavus Enterococcus gallinarum Enterococcus hirae Enterococcus durans
BeijingHuairouChickens10061 (61·0)39 (63·9)18 (29·5)4 (6·6)000
Pigs3920 (51·3)15 (75·0)3 (15·0)02 (10·0)00
ChangpinChickens9437 (39·4)26 (70·3)11 (29·7)0000
Pigs5012 (24·0)9 (75·0)3 (25·0)0000
TongzhouChickens5012 (24·0)8 (66·7)3 (25·0)1 (8·3)000
Pigs9953 (53·5)7 (13·2)19 (35·8)01 (1·9)26 (49·0)0
MiyunChickens500000000
Pigs5033 (66·0)32 (97·0)1 (3·0)0000
TotalChickens294110 (37·4)73 (66·4)32 (29·1)5 (4·5)000
Pigs238118 (49·6)63 (53·4)26 (22·0)03 (2·5)26 (22·0)0
ShandongXintaiChickens284155 (54·6)145 (93·5)10 (6·5)0000
LinyiChickens21870 (32·1)57 (81·4)9 (12·9)03 (4·3)01 (1·4)
TotalChickens502225 (58·6)202 (89·8)19 (8·4)03 (1·3)01 (0·4)
Total  1034453 (43·8)338 (74·6)77 (17·0)5 (1·1)6 (1·3)26 (5·7)1 (0·2)

PCR was used to identify four common species, Enterococcus faecalis, Enterococcus faecium, Enterococcus casseliflavus and Enterococcus gallinarum, using previously described primers (Macovei and Zurek 2006). Isolates that could not be identified by species-specific PCR were subjected to 16S rRNA gene sequencing. For this, a 1466-bp amplicon obtained with the universal prokaryotic primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGCTACCTTGTTACGACTT-3′) was analysed. The identification results were further confirmed using the Rapid ID 32 Strep system (bioMérieux, Craponne, France).

Antimicrobial susceptibility testing of Enterococcus isolates

Antimicrobial susceptibilities of all Enterococcus isolates were determined for the following antibiotics: penicillin, ampicillin, chloramphenicol, florfenicol, erythromycin, amikacin, tetracycline, enrofloxacin, rifampin and vancomycin, using broth microdilution methods as recommended by the Clinical and Laboratory Standards Institute (CLSI) documents M31-A3 (2008) and M100-S21 (2011). High-level streptomycin resistance (HLSR) and high-level gentamicin resistance (HLGR) were detected using an agar screening method (CLSI, M100-S21, 2011). All antimicrobial agents were obtained from the China Institute of Veterinary Drug Control (Beijing, China). To ensure the accuracy of the test, the minimum inhibitory concentrations (MICs) of all the strains were repeated three times. Test concentrations, breakpoint for each antimicrobial agent and quality control (QC) MIC ranges with Ent. faecalis ATCC 29212 are presented in Table 2. Multidrug-resistant (MDR) isolates were defined as those with resistance to compounds from at least three different classes of the antimicrobial agents tested. HLSR and HLGR were considered resistant to two classes of antimicrobials.

Table 2. MIC quality control (QC) ranges, antimicrobial test ranges and breakpoints used for antimicrobial susceptibility testing by broth microdilution or agar screening method
Antimicrobial agentMIC QC range (μg ml−1)aTest range (μg ml−1)MIC breakpoint (μg ml−1)b
S I R
  1. HLGR, high-level gentamicin resistance; HLSR, high-level streptomycin resistance; MIC, minimum inhibitory concentration; S, susceptible isolates; I, intermediate isolates; R, resistant isolates.

  2. a

    Broth microdilution QC ranges of Enterococcus faecalis ATCC 29212 approved by CLSI (2008) M31-A3.

  3. b

    MIC breakpoints for penicillin, ampicillin, chloramphenicol, erythromycin, tetracycline, rifampin, vancomycin, HLSR and HLGR of Enterococcus spp. are those recommended by the CLSI M100-S21 (2011), while amikacin (with no organism restriction) breakpoints are those recommended by M31-A3 (2008). Because standardized MIC breakpoints for florfenicol and enrofloxacin are not available for Enterococcus spp., we used the breakpoint for Streptococcus suis (swine) for florfenicol from CLSI (2008) M31-A3 and the breakpoint for Enterococcus spp. for ciprofloxacin as the breakpoint of enrofloxacin from CLSI (2011) M100-S21.

  4. c

    Tested by an agar screening method.

Penicillin1–44–64≤8≥16
Ampicillin0·5–24–64≤8≥16
Chloramphenicol4–161–64≤816≥32
Florfenicol2–82–1024≤24≥8
Erythromycin1–44–64≤0·51–4≥8
Amikacin64–25616–256≤1632≥64
Tetracycline8–324–128≤48≥16
Enrofloxacin0·12–10·25–8≤12≥4
Rifampin0·5–41–16≤12≥4
Vancomycin1–40·5–256≤48–16≥32
HLSRc≤1 colony2000≤1 colony>1 colony
HLGRc≤1 colony500≤1 colony>1 colony

Statistical analysis

The frequency of antimicrobial resistance profiles of Enterococcus isolates was compared between chickens from Beijing and Shandong, between chickens and pigs from Beijing and among different species of Enterococcus isolates, using the chi-squared test with a significance level of P < 0·05.

Results

Prevalence of Enterococcus species

Overall, 453 (43·8%) of the samples examined contained Enterococcus isolates (Beijing, n = 228; Shandong Province, n = 225). Among Beijing isolates, 118 and 110 were collected from pigs and chickens, respectively. The isolation rates from chickens and pigs in Beijing were 37·4% and 49·6%, respectively. However, the isolation rate from chickens from Shandong Province was 58·6%, which was somewhat higher than that from Beijing. The details of the Enterococcus identification are summarized in Table 1. Two most predominant species were Ent. faecalis (Beijing, n = 136; Shandong Province, n = 202) and Ent. faecium (Beijing, n = 58; Shandong Province, n = 19), which accounted for 91·6% of the total Enterococcus strains isolated. Only 38 isolates were identified as other Enterococcus species and consisted of five Ent. casseliflavus isolates (four from chickens in Huairou and one from a chicken sample from Tongzhou), six Ent. gallinarum isolates (two from pig samples from Huairou, one from a pig sample from Tongzhou and three from chicken samples from Linyi), 26 Enterococcus hirae isolates (all from pigs in Tongzhou) and one Enterococcus durans isolate (from a chicken sample from Linyi). In all six areas, Ent. faecalis was the dominant species, except in isolates from a pig farm in Tongzhou. At this location, isolates were determined to be Ent. hirae (26/53, 49·1%), followed by Ent. faecium (19/53, 35·8%) and Ent. faecalis (7/53, 13·2%).

Antimicrobial resistance of the Enterococcus isolates

All 453 isolates were susceptible to vancomycin, except only four Ent. faecium isolates from chickens from Huairou and two Ent. faecalis isolates from pigs from Changping, with a vancomycin MIC value (8 μg ml−1) approaching the lower intermediate threshold (8–16 μg ml−1) in the guidelines set out by the CLSI (M100-S21). Although <8% of isolates from Beijing were resistant to penicillin and ampicillin (7·9% and 4·2%, respectively), the majority were resistant to tetracycline (92·5%), amikacin (89·4%) and erythromycin (72·8%). Notably, 50·3% and 27·6% of Enterococcus isolates were indicated to have high-level resistance to streptomycin and gentamycin. For phenicols (chloramphenicol and florfenicol) and quinolones (enrofloxacin), nearly 30% of isolates were resistant (33·8%, 32·9% and 29·6%, respectively). Over 90% of the Enterococci were resistant to multiple antimicrobial agents.

The comparison of resistance profiles between total chicken isolates from Beijing and Shandong showed that only erythromycin and rifampin were not statistically significant (> 0·05), while rates of resistance to the majority of antibiotics varied significantly (< 0·05). Among 305 MDR isolates (91·0% of all chicken isolates), no statistical significance (= 0·04) was observed between isolates from Beijing (95, 86·4%) and from Shandong Province (210, 50·8%) (Table 3). Although Enterococcus strains from Beijing in this study were isolated from two distinct animal origins, statistical differences of antimicrobial resistance rates (< 0·05) between pigs and chickens in Beijing were only observed for chloramphenicol, tetracycline, HLSR and HLGR (Table 3). The resistance rates of all of these agents, except amikacin, enrofloxacin and rifampin, were relatively lower in chickens than in pigs. The MDR rates were 86·4% and 91·5% in chickens and pigs, respectively, which was not found to be significantly different (> 0·05) (Table 3).

Table 3. Resistance of 453 Enterococcus spp. isolates from different origins to antimicrobial agents
Antimicrobial agentPrevalence of resistant/MDR isolates (%) per originTotal (= 453)
ChickensPigs
BeijingShandongTotal (= 335)BeijingTotal (= 118)
HuairouChangpingTongzhouTotalXintaiLinyiTotalHuairouChangpingTongzhouMiyun
(= 61)(= 37)(= 12)(= 110)(= 155)(= 70)(= 225)(= 20)(= 12)(= 53)(= 33)
  1. MDR, multidrug-resistant; NC, not calculated.

Penicillin9 (14·8)7 (18·9)1 (8·3)17 (15·5)00017 (5·1)2 (10·0)2 (16·7)15 (28·3)019 (16·1)36 (7·9)
Ampicillin6 (9·8)1 (2·7)07 (6·4)0007 (2·1)2 (10·0)1 (8·3)9 (17·0)012 (10·2)19 (4·2)
Chloramphenicol10 (16·4)8 (21·6)2 (16·7)20 (18·2)31 (20·0)41 (58·6)72 (32·0)92 (27·5)15 (75·0)12 (100·0)1 (1·89)33 (100·0)61 (51·7)153 (33·8)
Florfenicol28 (45·9)13 (35·1)5 (41·7)46 (41·8)12 (7·7)32 (45·7)44 (19·6)90 (26·9)5 (25·0)11 (91·7)13 (24·5)30 (90·9)59 (50·0)149 (32·9)
Erythromycin43 (70·5)21 (56·8)12 (100·0)76 (69·1)107 (69·0)60 (85·7)167 (74·2)243 (72·5)20 (100·0)12 (100·0)22 (41·5)33 (100·0)87 (73·7)330 (72·8)
Amikacin55 (90·2)31 (83·8)9 (75·0)95 (86·4)155 (100·0)64 (91·4)219 (97·3)314 (93·7)18 (90·0)10 (83·3)35 (66·0)28 (84·9)91 (77·1)405 (89·4)
Tetracycline53 (86·9)24 (64·9)12 (100·0)89 (80·9)152 (98·1)64 (91·4)216 (96·0)305 (91·0)18 (90·0)11 (91·7)52 (98·1)33 (100·0)114 (96·6)419 (92·5)
Enrofloxacin30 (49·2)9 (24·3)3 (25·0)42 (38·2)46 (29·7)15 (21·4)61 (27·1)103 (30·7)8 (40·0)11 (91·7)2 (3·8)10 (30·3)31 (26·3)134 (29·6)
Rifampin37 (60·7)28 (75·7)5 (41·7)70 (63·6)83 (53·5)37 (52·9)120 (53·3)190 (56·7)11 (55·0)3 (25·0)32 (60·4)27 (81·8)73 (61·9)263 (58·1)
Vancomycin00000000000000
High-level streptomycin resistance23 (37·7)22 (59·5)11 (91·7)56 (50·9)51 (32·9)30 (42·9)81 (36·0)137 (40·9)16 (80·0)12 (100·0)34 (64·2)29 (87·9)91 (77·1)228 (50·3)
High-level gentamicin resistance24 (39·3)14 (37·8)1 (8·3)39 (35·5)16 (10·3)10 (14·3)26 (11·6)65 (19·4)16 (80·0)10 (83·3)8 (15·1)26 (78·8)60 (50·8)125 (27·6)
MDR52 (85·3)31 (83·8)12 (100·0)95 (86·4)145 (93·6)65 (92·9)210 (93·3)305 (91·0)18 (90·0)12 (100·0)45 (84·9)33 (100·0)108 (91·5)413 (91·2)

In this study, six species of Enterococcus were identified from 453 isolates. The antimicrobial resistance rates of isolates from the different species, regardless of the animal origin, are presented in Table 4. The vast majority of the isolates were resistant to tetracycline (>91%), which had the highest resistance rates in all species except Ent. faecalis. Among Ent. faecalis isolates, the highest rate of resistance was to amikacin (97·3%). Overall, the most similar resistance profiles were observed between Ent. faecalis and Ent. faecium, while penicillin, ampicillin, florfenicol, amikacin, enrofloxacin and HLSR rates were significantly different (< 0·05) (Table 4). Compared with these antimicrobial agents, the resistance rates to β-lactams (penicillin and ampicillin), enrofloxacin and HLSR were relatively low in Ent. faecalis (2·1%, 1·8%, 27·2% and 45·6%, respectively), but were much higher in Ent. faecium isolates (27·3%, 10·4%, 46·8% and 63·6%, respectively) (Table 4). Interestingly, the prevalence of amikacin resistance was significantly higher in Ent. faecalis (97·3%) than that in Ent. faecium (71·4%) (Table 4). Only one Ent. durans isolate was resistant to HLSR. The other three species of Enterococcus were resistant to the majority of antibiotics, with varied resistance rates, but none of the 26 Ent. hirae or six Ent. gallinarum isolates were resistant to chloramphenicol, penicillin or ampicillin. Among the six species, rates of resistance to the majority of antibiotics varied significantly (< 0·05), and only rifampin (P > 0·05) rate showed no significant difference (Table 4). The highest MDR rate was observed for Ent. casseliflavus (100·0%), followed by Ent. faecium (92·2%), Ent. faecalis (91·7%), Ent. hirae (84·6%) and Ent. gallinarum (83·3%) (Table 4).

Table 4. Resistance of different Enterococcus spp. to antimicrobial agents
Antimicrobial agentPrevalence of resistant/MDR isolates (%) per species
Enterococcus faecalis (= 338)Enterococcus faecium (= 77)Enterococcus hirae (= 26)Enterococcus gallinarum (= 6)Enterococcus casseliflavus (= 5)Enterococcus durans (= 1)
  1. MDR, multidrug-resistant; NC, not calculated.

Penicillin7 (2·1)21 (27·3)6 (23·1)02 (40·0)0
Ampicillin6 (1·8)8 (10·4)4 (15·4)01 (20·0)0
Chloramphenicol126 (37·7)25 (32·5)01 (16·7)1 (20·0)0
Florfenicol104 (30·8)33 (42·9)5 (19·2)3 (50·0)4 (80·0)0
Erythromycin256 (75·7)55 (71·4)11 (42·3)4 (66·7)4 (80·0)0
Amikacin329 (97·3)55 (71·4)16 (61·5)3 (50·0)2 (40·0)0
Tetracycline309 (91·4)73 (94·8)26 (100·0)6 (100·0)5 (100·0)0
Enrofloxacin92 (27·2)36 (46·8)02 (33·3)4 (80·0)0
Rifampin192 (56·8)48 (62·3)16 (61·5)2 (33·3)5 (100·0)0
Vancomycin000000
High-level streptomycin resistance154 (45·6)49 (63·6)15 (57·7)4 (66·7)5 (100·0)1 (100·0)
High-level gentamicin resistance90 (26·6)26 (33·8)2 (7·7)3 (50·0)4 (80·0)0
MDR310 (91·7)71 (92·2)22 (84·6)5 (83·3)5 (100·0)0

Discussion

In this study, Enterococcus species were isolated from 43·8% of the faecal samples collected from food animals. The finding that the majority of isolates were determined to be Ent. faecalis, followed by Ent. faecium, is consistent with previous reports that these are the two predominant Enterococcus species isolated from humans and animals (Murray 1990; Cetinkaya et al. 2000). However, the isolation rates of Ent. faecalis and Ent. faecium varied among the six regions surveyed, especially in the samples collected from pig farms in Tongzhou, where most of the Enterococcus isolates were Ent. hirae (26/53, 49·1%). The reasons for this variation are unknown but could be attributed to differences in sampling practices, geographical location and the farm environment. Similarly, other investigators have reported wide variation in the prevalence of Enterococcus in other countries; for instance, isolation rates of Ent. hirae and Ent. faecium were reported to be 27·1% and 21·3%, respectively, from pigs in Korea (Hwang et al. 2009), and 39% and 29% in Denmark (Aarestrup et al. 2000).

Enterococci are intrinsically resistant to several antibiotics and also readily accumulate mutations and exogenous genes that confer additional resistance (Arias and Murray 2012). The glycopeptide antibiotics vancomycin and teicoplanin are important reserve antibiotics in case of resistance to penicillins and high-level resistance to aminoglycosides and are known as ‘key last-line bactericidal drugs for treating infections’ (Kos et al. 2012). Although the prevalence of VRE in China increased from 0 in 2005 to 4·9% in 2010, according to the Gram-Positive Cocci Resistance Surveillance Program (2005–2010) (Zhao et al. 2012), none of the 453 Enterococcus isolates identified in the current study were resistant to vancomycin (the vancomycin MIC values of only six isolates approached the lower intermediate threshold). It was previously reported that the antimicrobial resistance rates among animal Enterococcus strains are lower than those found among human strains isolated from hospital patients (Butaye et al. 2001). Such low resistance rates may be attributed to the fact that none of the glycopeptides, including avoparcin, which has been demonstrated to correlate well with the frequent isolation of VRE (McDonald et al. 1997; Wegener 1998; Wegener et al. 1999), are authorized for use as antimicrobial growth promoters in feed for farm animals in China.

Our findings revealed the relatively high prevalence of high-level aminoglycoside-resistant Enterococcus species in chickens and pigs in China (HLSR, 50·3%; HLGR, 27·6%), especially HLSR. The prevalence of HLSR and HLGR Enterococcus varies greatly between different countries. A previous study (Han et al. 2011) from South Korea reported that 23·4% of Ent. faecalis and 14·8% of Ent. faecium isolates from both broiler chickens and pigs were HLGR, while 25·5% of Ent. faecalis and 42·6% of Ent. faecium were HLSR. In Canada, resistant Ent. faecalis and Ent. faecium isolates from broiler chicken and turkey flocks were 9·6% and 4·3% HLGR, respectively, and 46·7% and 38·5% HLSR, respectively (Tremblay et al. 2011), of which only the HLSR Ent. faecalis rate was higher than that observed in the current study. A major reason for the high-level aminoglycoside resistance rates of Enterococcus in our study may be that this class of antibiotics is one of the most commonly used antimicrobials in the sampled farms, where it is used as both a therapeutic and nontherapeutic antimicrobial in veterinary medicine. Similar to the aminoglycosides, the resistance to phenicols was also more prevalent in the Enterococcus isolates obtained in this study, compared with the United States (Hayes et al. 2004). The use of chloramphenicol in food animal production has been banned for more than 10 years in China. However, florfenicol, a fluorinated chloramphenicol derivative that was licensed in China in 1999 exclusively for the use in veterinary medicine, has been widely used in food animal production to prevent disease (Wang et al. 2012), including on the farms sampled here. But, until now, no florfenicol-specific breakpoints presented in the CLSI document are valid for Enterococcus spp. In this study, ≥8 μg ml−1 was the tentative breakpoint for both swine and poultry resistance. The high resistance rates to phenicols were likely caused by the long-term use of florfenicol for both disease prevention and treatment in China prior to 2006 (Wang et al. 2012). Resistance to tetracycline and erythromycin was also highly prevalent, which is consistent with previous studies on tetracycline- or erythromycin-resistant Enterococcus from other regions (Hwang et al. 2009; Tremblay et al. 2011).

In the current study, we found no statistically significant differences in the majority of antibiotic resistance profiles of isolates from chickens vs pigs in Beijing (P < 0·05). Another interesting finding was the more frequent occurrence of resistance among the Enterococcus isolates from Beijing chickens compared with those from Shandong Province. According to our investigation of antibiotic usage, tetracyclines, macrolides, fluoroquinolones and cephalosporins had been used to prevent disease at the sampled chicken farms from both Beijing and Shandong Province. However, the major differences in the occurrence of resistance between the two regions were most likely due to the differences in the feeding styles and the usage of therapeutic drugs. The two farms at Xintai and Linyi in Shandong Province are intensive and use an all-in all-out–based system of production, with the primary aim of reducing the transmission of infectious agents. However, in Beijing, four farms were free-range and farmed by individuals. The owners of the farms preferred to use a wide range of antibiotics, such as penicillin–aminoglycoside synergy or florfenicol, rather than relying on a clean environment to maintain their animals' health. Therefore, long-term use of these antimicrobials in veterinary practice in Beijing may be the primary reason for the presence of penicillin-resistant and ampicillin-resistant isolates and the higher resistance rates for florfenicol (Beijing, 46·1%; Shandong Province, 19·6%), HLSR (Beijing, 64·5%; Shandong Province, 36·0%) and HLGR (Beijing, 43·4%; Shandong Province, 11·6%). Although this kind of comparison might be rare in Enterococcus spp., this situation is in contrast to some previously reported studies on Escherichia coli in other countries. Miranda et al. (2008) reported lower levels of resistance in E. coli isolates from organic poultry meat, while Obeng et al. (2012) reported that there was no significant difference of antimicrobial resistance in E. coli between intensively farmed and free-range chickens.

In conclusion, antimicrobial resistance is prevalent in Enterococcus isolates from chickens and pigs from Beijing and Shandong Province, China, and many isolates are resistant to multiple antimicrobial agents. Although enterococci are part of the endogenous flora, they are capable of causing serious disease in humans. The prevalence of MDR enterococci in food animals is alarming, especially given the fact that very few antimicrobial agents can be used to control enterococcal infection because of its natural antibiotic resistance and the ability to acquire resistance from other bacteria.

Acknowledgements

This work was supported by grants from the National Natural Science Foundation of China (no. 31001087), Special Fund for Agro-scientific Research in the Public Interest (201203040) and Key Projects in the National Science & Technology Pillar Program during the Twelfth Five-year Plan Period (2012BAK01B02).

Conflict of interest

We certify that no actual or potential conflict of interest in relation to this article exists.

Ancillary