Antimicrobial resistance in Enterococcus faecium and Enterococcus faecalis isolates of swine origin from eighteen provinces in China.

Enterococcus faecium and E. faecalis are important human pathogens and also served as sentinel organisms for monitoring systems of antimicrobial resistance in both animals and humans. In this study, 106 E. faecium and 56 E. faecalis isolates were collected from 61 pig farms in 18 proveinces of China. Antimicrobial susceptibility was determined for 9 clinically important antibiotics and 3 antimicrobial growth promoters. The Enterococcus isolates showed high prevalence of resistance to medically important antibiotics, such as ampicillin (50.9% for E. faecium and 19.6% for E. faecalis), chloramphenicol (24.5% for E. faecium and 41.1% for E. faecalis), erythromycin (83.0% for E. faecium and 91.1% for E. faecalis), tetracycline (79.2% for E. faecium and 100% for E. faecalis), quinupristin/dalfopristin (26.4% for E. faecium) and ciprofloxacin (73.6% for E. faecium and 66.1% for E. faecalis). Resistance to tigecycline, linezolid and vancomycin was very rare. The resistance status of three representative in-feed antibiotics bacitracin, nosiheptide and enramycin was firstly investigated with Enterococcus as indicator bacteria. The Enterococcus isolates showed extremely high frequency of bacitracin resistance (96.7% for E. faecium and 87.8% for E. faecalis), while no nosiheptide and enramycin resistance was observed. Pulsed-field gel electrophoresis (PFGE) analysis showed that a majority of E. faecium and E. faecalis strains showed unrelated profiles, indicating high heterogeneity among the Enterococcus isolates. Our study provided basic data on the antimicrobial resistance of E. faecium and E. faecalis isolates.

Enterococcus faecium and E. faecalis are opportunistic pathogens responsible for several human infectious diseases, including urinary and bloodstream infections and endocarditis [2]. Multiple-drug resistant E. faecium and E. faecalis have been a major public health threat for last two decades, and vancomycin-resistant E. faecium is an antimicrobial-resistant pathogen regarded by World Health Organization (WHO) as a global priority for research and development of new antibiotics. In addition, E. faecium and E. faecalis are commensal bacteria present in the gut microbiota of humans and animals, and consequently, serve as Gram-positive indicator bacteria in animal-origin antimicrobial resistance (AMR) surveillance programs in several countries and areas, such as The European Antimicrobial Susceptibility Surveillance in Animals (EASSA) in the European Union, The Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) in Canada, The Japanese veterinary antimicrobial resistance monitoring systems (JVARM) in Japan, and The National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS) in the United States [6]. Few studies involving the AMR surveillance of enterococci of animal origin have been reported in China. Although China has been running the AMR surveillance Network for Bacteria of Animal Origin since 2008, no published data are available.

Antibiotic growth promoters (AGPs) have been widely and extensively used in food-animal productions for many years [15]. This poses risks to human health due to the selection of antibiotic-resistant bacteria and the potential transmission of AMR bacteria/genes to humans through consumption chains of animal food products. For this reason, many countries have banned antibiotics as feed addictives for animal growth promotion. For example, the European Union banned all AGPs in 2006 and, in China, all antibiotics were formally forbidden to be used as feed addictives since 2020 [9, 20]. AMR monitoring of the AGPs could provide useful information for evaluating the effects of the antibiotic withdrawal and policy making. However, few studies have been conducted to investigate the resistance status of AGPs, especially those exclusively used as feed addictives.

In the present study, antimicrobial resistance profiles of clinically important antibiotics and representative AGPs and genetic relationships were determined for E. faecium and E. faecalis isolates from pig farms in 18 provinces of China.

MATERIALS AND METHODS

Sample collection and bacteria isolation

Between October 2017 and January 2019, a total of 843 faecal samples were collected from 61 swine farms in 18 provinces of China, including Xinjiang, Qinghai, Sichuan, Yunnan, Guizhou, Hainan, Jiangxi, Fujian, Zhejiang, Shandong, Beijing, Liaoning, Hebei, Henan, Shaanxi, Shanxi, Jilin and Heilongjiang. Rectal swabs were collected from individual pig using the ESwab Liquid Amies transport system (Copan Diagnostic Inc., Murrieta, CA, USA) and transported to laboratory for further processing.

For Enterococcus isolation, 10 μl liquid samples were firstly transferred into 1 ml nutrient broth with 6.5% NaCl and incubated at 45°C for 24 hr. These cultures were then streaked onto Slanetz and Bartley medium (Oxoid, Basingstoke, UK) and incubated at 37°C for 24 hr [16]. One presumptive Enterococcus colony per sample was picked and sub-cultured for preservation and further testing. Species identification was performed by MALDI-TOF MS (VITEK MS, bioMerieux, Marcy-lʼEtoile, France).

Antimicrobial susceptibility testing

The minimal inhibitory concentrations (MICs) of 12 antimicrobials were tested, including ampicillin, chloramphenicol, erythromycin, tetracycline, quinupristin/dalfopristin, tigecycline, linezolid, ciprofloxacin, vancomycin and three representative AGPs bacitracin, nosiheptide and enramycin. MIC test was performed with agar dilution method or using MIC Test Strips (only for quinupristin/dalfopristin; Liofilchem, Roseto degli Abruzzi, Italy) in accordance with CLSI recommendations [4]. The resistance breakpoints of all antibiotics were interpreted according to the CLSI-M100-S28 document, except for tigecycline and bacitracin, for which the EUCAST breakpoint and epidemiological cut-off (ECOFF) value was used, respectively ( http://www.eucast.org). Due to the absence of resistance breakpoints and MIC data of nosiheptide and enramycin, Only MIC50 and MIC90 values were exhibited to reflect the MIC distributions of the two antibiotics (Tables 2and 3). E. faecalis ATCC 29212 and Staphylococcus aureus ATCC29213 served as quality control strains.

Table 2

 Resistance profile of Enterococcus faecium isolates in swine farms from 18 provinces of China

Table 3

 Resistance profile of Enterococcus faecalis isolates in swine farms from 18 provinces of China

Pulsed-field gel electrophoresis (PFGE)

The DNA fingerprinting profiles of the 106 E. faecium and 56 E. faecalis isolates were determined by SmaI-PFGE typing, as described previously [13]. Salmonella Braenderup strain H9812 (ATCC BAA 664) digested by XbaI restriction enzyme was used as a standard size marker. The fingerprinting profiles were analyzed using the BioNumerics 7.1 software (Applied Maths, Kortrijk, Belgium). The unweighted-pair group method using average linkages (UPGMA) was used to construct dendrograms for E. faecium and E. faecalis isolates on the basis of Dice coefficient with 1.0% band-position tolerance and 1.5% optimization. Strains with ≥80% similarity were considered as genetically related [17].

RESULTS

Enterococcus isolation and identification

Among the 843 faecal samples, a total of 225 Enterococcus isolates were identified, including 106 E. faecium strains, 56 E. faecalis strains, 34 E. hirae strains, 17 E. gallinarum strains, 7 E. casseliflavus strains, 4 E. durans strains and 1 E. thailandicus strain (Table 1). Since E. faecium and E. faecalis are commonly used as indicator bacteria in AMR monitoring system [6], the E. faecium (47.1%, 106/225) and E. faecalis (24.9%, 56/225) isolates were subjected to further susceptibility testing and genotyping.

Table 1.

The isolation of Enterococcus spp. of pig origin from 18 provinces in China

Province	Farm numbers	Sample numbers	No. of E. faecuim isolates	No. of E. faecalis isolates	No. of E. gallinarum isolates	No. of E. casseliflavus isolates	No. of E. hirae isolates	No. of E. durans isolates	No. of E. thailandicus isolates
Guizhou	4	53	10	4	-	-	4	1	-
Sichuan	5	55	4	9	1	-	6	-	1
Beijing	3	50	5	10	-	-	-	-	-
Yunnan	2	26	6	1	-	-	4	-	-
Shanxi	3	40	1	1	-	-	1	-	-
Zhejiang	4	73	6	0	-	-	3	-	-
Liaoning	3	36	3	1	-	-	-	-	-
Fujian	3	38	2	4	2	-	1	-	-
Hainan	3	44	2	0	-	-	3	-	-
Heilongjian	3	41	4	0	-	-	-	-	-
Hebei	5	55	10	3	2	-	3	2	-
Jilin	2	34	2	2	-	-	-	1	-
Qinghai	4	62	24	2	-	-	1	-	-
Shanghai	3	65	4	7	8	7	1	-	-
Jiangxi	4	47	8	6	2	-	1	-	-
Shandong	4	45	9	0	-	-	3	-	-
Henan	3	36	3	0	-	-	2	-	-
Xinjiang	3	43	3	6	2	-	1	-	-
Total	61	843	106	56	17	7	34	4	1

Antimicrobial susceptibility

The 106 E. faecium and 56 E. faecalis isolates showed high rates of resistance to erythromycin (83.0% for E. faecium and 91.1% for E. faecalis), tetracycline (79.2% for E. faecium and 100% for E. faecalis), and ciprofloxacin (73.6% for E. faecium and 66.1% for E. faecalis). The present study found low rates of resistance to tigecycline (1.9% for E. faecium and 1.8% for E. faecalis), linezolid (2.8% for E. faecium and 5.4% for E. faecalis), and vancomycin (0% for both species). Two E. faecium isolates and one E. faecalis isolate were resistant to tigecycline (MIC, 0.5 μg/ml). Three E. faecium isolates and three E. faecalis isolates showed low-level resistance to linezolid (MIC, 8 μg/ml). Furthermore, the resistance rate of E. faecium and E. faecalis isolates to ampicillin was 50.9% and 19.6%, respectively. The resistance rate of E. faecium isolates to quinupristin/dalfopristin was 26.4%.

Here, we used E. faecium and E. faecalis isolates to investigate the resistance status of three in-feed antibiotics, bacitracin, nosiheptide, and enramycin (Tables 2 and 3). According to the EUCUST ECOFF values of E. faecium and E. faecalis, 96.7% of E. faecium isolates and 87.8% E. faecalis isolates exhibited bacitracin resistance. The MIC50 and MIC90 for both species are >256 μg/ml. The MIC50 and MIC90 values of nosiheptide for E. faecium isolates were 0.004 μg/ml and 0.16 μg/ml, respectively, and those for E. faecalis isolates were 0.008 μg/ml and 0.008 μg/ml, respectively. The MIC50 and MIC90 values of enramycin for both E. faecium and E. faecalis isolates were 4 μg/ml and 4 μg/ml, respectively.

PFGE typing

The genetic relatedness of the 106 E. faecium and 56 E. faecalis isolates was analyzed by PFGE (Figs. 1 and 2). In general, highly diverse profiles were observed for both E. faecium and E. faecalis isolates, especially for the strains from different regions. The result revealed that there are no predominant E. faecium and E. faecalis clones in pig industry in China. A small proportion of the collected strains, most of which are from same provinces, showed phylogenetic linkage (≥85% similarity). Nevertheless, interregional transmissions of some genotypes were also observed. For example, seventeen E. faecium strains obtained from four provinces (Qinghai, Sichuan, Hebei and Xinjiang) showed ≥90% pulsotype similarity (Fig. 1, black box).

Fig. 1.

SmaI-pulsed-field gel electrophoresis (PFGE) profiles of the 106 Enterococcus faecuim isolates in this study. The dotted line on the dendrogram indicates 80% similarity. The strains in the black box are isolated from different regions and showed ≥90% similar PFGE profile.

Fig. 2.

SmaI-pulsed-field gel electrophoresis (PFGE) profiles of the 56 Enterococcus faecalis isolates in this study. The dotted line on the dendrogram indicates 80% similarity.

DISCUSSION

This study revealed that the resistance rates of enterococcus isolates of pig origin in China to erythromycin, tetracycline and ciprofloxacin was higher than those in Europe and the United States [5, 6, 19]. Macrolides (tilmicosin and tylosin), tetracyclines (tetracycline), and fluoroquinolones (enrofloxacin) are widely used in pig production in China, which may result in the severe resistance condition for these drugs. Similar to the results of other large-scale investigations in European countries and the United States, rare resistance to tigecycline, linezolid and vancomycin were observed in this study [8, 11]. Tigecycline, linezolid and vancomycin are critically important for the treatment of Enterococcus infections and not used in food-producing animals in China. Our results, together with reports in other areas [1, 10, 18], demonstrated that resistance to the three last-line antibiotics are infrequent in enterococci of food-animal origin.

Quinupristin/Dalfopristin is a streptogramin combination and an important treatment option for vancomycin-resistant E. faecium infections in humans [7]. The streptogramin mixture virginiamycin has been commonly used in animal feed as a growth promoter for many years, which may be the reason for the high prevalence of resistance to quinupristin/dalfopristin in China. Previous studies have shown that resistance to ampicillin mainly occurs in E. faecium, but is very rare in E. faecalis [14, 19]. However, 19.6% E. faecalis isolates exhibited ampicillin resistance in this study. Further studies are necessary to investigate the molecular mechanisms underlying this phenomenon.

Bacitracin, nosiheptide, and enramycin are active against Gram-positive bacteria and have been licensed as feed additives in food-animal production for decades in China. However, few studies have evaluated resistance to these antibiotics. Resistance to bacitracin in Enterococcus is mostly attributed to the presence of bcrABDR cluster, which encodes a putative ATP-binding cassette (ABC) transporter [12]. Previous studies have shown that the plasmid-carrying bcrABDR gene is highly prevalent in Enterococcus of animal origin in China [3, 21]. The continuous selection pressure given by in-feed use of bacitracin may promote the dissemination of bcrABDR gene and led to the extremely high resistance frequency. Our study evaluated antimicrobial susceptibility of nosiheptide and enramycin with Enterococcus as indicator bacteria. Unlike the high-level resistance observed for bacitracin, the MIC50 and MIC90 values of nosiheptide and enramycin for both Enterococcus spp. were close to those for the wild-type E. faecalis strain ATCC29212 (nosiheptide MIC, 0.008 μg/ml; enramycin MIC, 2 μg/ml). Besides, none of the analyzed strains presented high MIC values. Although there are no available resistance breakpoints for the two growth-promoting antibiotics, the MIC distributions observed in this study indicated that resistance to the two drugs is infrequent, even though they have been using as feed additive in pig industry for decades.

In summary, this study gave an overview of the antimicrobial resistance of E. faecium and E. faecalis isolates in pig production in China. Resistance to medically important antimicrobials was high, except for tigecycline, linezolid, and vancomycin. The resistance prevalence of in-feed antibiotics was also investigated. The higher rate of resistance to bacitracin and absence of resistance to nosiheptide and enramycin may provide useful information in the policy-making for the use of antibiotics in pig farms in China.

CONFLICT OF INTEREST

The authors have nothing to disclose.