Prevalence of Colistin-Resistant Bacteria among Retail Meats in Japan.

Colistin (CST) is considered the last resort for the treatment of infectious diseases due to multidrug-resistant bacteria. Since the mcr-1 gene has been reported in Enterobacteriaceae isolated from food, animals, and humans in China, the prevalence of CST-resistant bacteria has been of great concern. Here, we investigated the prevalence of CST resistance and plasmid-mediated colistin-resistance genes (mcr) in gram-negative bacteria isolated among retail meats in Japan. CST-resistant bacteria were isolated from 310 domestic retail meats (103 chicken meat, 103 pork, and 104 beef) purchased between May 2017 and July 2018 from retail shops in Japan using CST-containing media and antimicrobial susceptibility testing. The mcr gene was investigated in isolates with a CST minimum inhibitory concentration of ≥1 μg/mL. Excluding the intrinsically CST-resistant isolates, CST-resistant bacteria were isolated from 39 of the total chicken meats (37.9%), 19 of the pork samples (18.4%), and 18 of the beef samples (17.3%). A total of 459 isolates were identified, out of which 99 were CST-resistant. CST resistance (resistance breakpoints: Aeromonas, >4 μg/mL; others, >2 μg/mL) was found in Aeromonas spp. (48/206, 23.3%), Yersinia spp. (5/112, 4.5%), Escherichia coli (23/39, 59%), Citrobacter spp. (4/26, 15.4%), Klebsiella spp. (2/23, 8.7%), Raoultella spp. (2/16, 12.5%), Enterobacter spp. (7/14, 50%), Pseudomonas spp. (1/8, 12.5%), Pantoea spp. (5/7, 71.4%), Ewingella spp. (1/4, 25%), and Kluyvera spp. (1/2, 50%). The mcr gene was detected in 16 isolates: mcr-1 in 14 isolates of E. coli from 10 chicken samples (9.7%), and mcr-3 in two isolates of Aeromonas sobria from pork and chicken samples (each 1.0%). The findings of this study highlight the necessity of surveillance of CST resistance and resistance genes in bacteria that contaminate retail meats. ©2021 Food Safety Commission, Cabinet Office, Government of Japan.

1. Introduction

Colistin (CST), a critically important antimicrobial agent, is considered the last resort for the treatment of infections due to multidrug-resistant bacteria[1]). CST acts on gram-negative bacteria by targeting lipopolysaccharides (LPS) on the outer membrane, leading to cell wall disintegration and subsequent bacterial death[2]). The resistance mechanism against CST is likely due to chromosomal mutations in the two-component system, such as PhoPQ and PmrAB, located on the chromosome of gram-negative bacteria, leading to the modification of LPS through the reduction of the negative charge of the lipid A moiety[3]). An alternate mechanism involves the plasmid-mediated CST-resistance gene (mcr). The mcr-1 gene found in Enterobacteriaceae isolated from food animals and humans was recently reported in China[4]). Thereafter, a surveillance study conducted in Denmark identified the mcr-1 gene in 0.2% of Escherichia coli (E. coli) implicated in human bloodstream infections[5]). In addition to mcr-1, other mcr genes including mcr-2[6]), mcr-3[7]), mcr-4[8]), mcr-5[9]), mcr-6[10]), mcr-7[11]), mcr-8[12]), mcr-9[13]), and mcr-10[14]) have been found in different plasmids in gram-negative bacteria.

The mcr genes have been frequently found in Enterobacteriaceae isolated from chickens[15]), pigs[16]), and turkeys[17]). Additionally, the mcr gene has been identified in other bacteria, including Aeromonas spp.[18]), Acinetobacter spp., and Pseudomonas spp.[19]). In Japan, the mcr-1 gene has been detected in E. coli isolated from livestock and humans[20],[21],[22]). The prevalence of mcr-1 (30%), mcr-3 (8.3%), and mcr-5 (28.3%) in E. coli was reported among diseased pigs[23]). Moreover, the prevalence of the mcr-1 gene in E. coli isolated among retail meats was also evaluated[24]). These reports prompted the Food Safety Commission of Japan (FSCJ) to undertake a risk assessment of CST use in food-producing animals in 2017[25]). Results of this assessment revealed a medium risk, and as a result, the Ministry of Agriculture, Forestry and Fisheries of Japan issued a directive for the withdrawal of CST as a growth promoter and relegated the drug as a second-choice antimicrobial agent for therapeutic use in food-producing animals. In the assessment, the FSCJ also recommended that important information on the prevalence of mcr genes in bacteria be regularly updated to guide necessary interventions for infection control. Therefore, in this study, we investigated the prevalence of CST resistance and mcr genes in gram-negative bacteria isolated among retail meats in Japan.

2. Materials and Methods

Sample Collection of Retail Meats

A total of 310 domestic retail meats comprising 103 chicken meats, 103 pork samples, and 104 beef samples were purchased between May 2017 and July 2018 from retail shops in Hokkaido (33 samples), Iwate (27 samples), Tokyo (46 samples), Chiba (24 samples), Gifu (60 samples), Shiga (24 samples), Osaka (36 samples), Fukuoka (30 samples), and Kumamoto (30 samples) prefectures located in Japan.

Isolation and Identification of Bacteria among Retail Meats

Using sterile forceps, each meat sample was gently pressed against the surface of deoxycholate hydrogen sulfide lactose (DHL) agar supplemented with 0.1 μg/mL of CST (CST-DHL medium) and incubated at 37°C overnight. For enrichment isolation, 5 g of each meat sample was aseptically cut and first cultured in 45 mL of tryptic soy broth (enrichment medium) at 37°C overnight. Afterward, the obtained bacterial suspension was streaked onto the CST-DHL medium and incubated at 37°C overnight. A maximum of three distinct red colonies formed per sample were preferentially picked at random. The properties of the bacteria were examined using triple sugar iron medium, lysine indole motility medium (Eiken Chemical Co., Ltd., Tokyo, Japan), and cytochrome oxidase test filter paper (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan). The isolates were analyzed using VITEK® 2 GN identification card (Sysmex BioMérieux, Tokyo, Japan) and PCR to further identify Aeromonas spp., as described previously[26]).

Antimicrobial Susceptibility Testing

The minimum inhibitory concentration (MIC) of antimicrobials for bacteria isolated among retail meats were determined by the broth microdilution method using a frozen plate (Eiken Chemical Co., Ltd.,) following the manufacturer’s instructions. The following 15 antimicrobial agents were tested: CST (0.25–8 μg/mL), ampicillin (AMP, 2–64 μg/mL), cefazolin (CFZ, 1–32 μg/mL), cefotaxime (CTX, 0.5–16 μg/mL), ceftazidime (CAZ, 1–32 μg/mL), meropenem (MEM, 0.5–16 μg/mL), tetracycline (TET, 2–64 μg/mL), gentamicin (GEN, 2–64 μg/mL), kanamycin (KAN, 4–128 μg/mL), amikacin (4–128 μg/mL), nalidixic acid (NAL, 4–128 μg/mL), ciprofloxacin (CIP, 0.12–4 μg/mL), levofloxacin (0.25–8 μg/mL), chloramphenicol (CHL, 4–128 μg/mL), and sulfamethoxazole/trimethoprim (SXT, 9.5/0.5–152/8 μg/mL). The antimicrobial resistance breakpoints, except that for CST, were interpreted following the Clinical and Laboratory Standard Institute guidelines[27]). The European Committee on Antimicrobial Susceptibility Testing resistance breakpoints were used for CST[28]). Isolates resistant to three or more antimicrobial classes were identified as multi-drug resistant[29]).

Detection of the mcr gene

The presence of mcr-1 through mcr-5 genes was investigated in 365 isolates (except in intrinsically CST-resistant bacterial isolates) among retail meats with CST MICs of ≥1 μg/mL using primers, as described previously[6],[7],[8],[9],[30]).

Whole-genome Sequencing

The isolates harboring the mcr gene were analyzed by whole-genome sequencing. Whole-cell DNA was purified from each isolate using a QIAquick PCR purification Kit (QIAGEN, Hilden, Germany). A DNA sequencing library (insert size: 750–1,000 bp) was prepared using a QIAseq FX DNA Library Kit (QIAGEN) for paired-end sequencing on an Illumina MiSeq or iSeq sequencer (Illumina, Inc., San Diego, CA, USA) according to the manufacturer’s instructions. The whole-genome sequence data of each isolate was analyzed using whole-genome and plasmid sequence databases from GenEpid-J (https://gph.niid.go.jp/genepid-j/) and PlasmidFinder version 2.0.1 (https://cge.cbs.dtu.dk/services/PlasmidFinder/) to identify plasmid replicon type, and ResFinder version 4.1 to identify acquired antimicrobial resistance genes, which is available in the homepage of the Center for Genomic Epidemiology website (http://www.genomicepidemiology.org/). Briefly, to identify resistance genes on plasmid, the plasmid sequence data was extracted from the whole genome sequence with PlasmidFinder and analyzed using the nucleotide Basic Local Alignment Sequencing Tool (BLASTn; National Center for Biotechnology Information) against reference sequences for confirmation. By using the ResFinder version 4.1, we then searched for resistance genes in the plasmid sequence data as previously described[31]).

3. Results

Isolation of Bacteria among Retail Meats

A total of 1,125 bacteria were isolated from 281 of 310 retail meat samples (90.6%) purchased. Four hundred and fifty-four isolates were obtained from chicken samples (98/103, 95.1%), 348 isolates from pork samples (95/103, 92.2%), and 323 isolates from beef samples (88/104, 84.6%). In total, 18 bacterial genera, including six genera (Cedecea spp., Hafnia spp., Morganella spp., Proteus spp., Providencia spp., and Serratia spp.) that are intrinsically CST-resistant were isolated (data on intrinsically CST-resistant isolates not shown). Excluding the intrinsically CST-resistant isolates, the predominant genus isolated was Aeromonas spp. (206/459, 44.9%), followed by Yersinia spp. (112/459, 24.4%), and Escherichia spp. (39/459, 8.5%). The remaining bacteria belonged to Citrobacter spp., Klebsiella spp., Raoultella spp., Enterobacter spp., Pseudomonas spp., Pantoea spp., Ewingella spp., Acinetobacter spp., and Kluyvera spp. (Table 1).

Table 1.

 Colistin-resistant bacteria isolated among retail chicken meat, pork, and beef

Bacteria		Total pos. sample (Total isolates)	Chicken			Pork			Beef
Genus	Species		Pos. sample (Isolate No.)	Pos. CST-res sample (CST-res Isolate No.)	% CST-res Isolates	Pos. sample (Isolate No.)	Pos. CST-res sample (CST-res Isolate No.)	% CST-res Isolates	Pos. sample (Isolate No.)	Pos. CST-res sample (CST-res Isolate No.)	% CST-res Isolates
Aeromonas	sobria	63(117)	50(101)	5(7)	6.9	10(13)	2(2)	15.4	3(3)	1(1)	33.3
	hydrophila	59(74)	38(44)	20(23)	52.2	14(18)	6(6)	33.3	7(12)	4(6)	50
	veronii	12(15)	12(12)	2(2)	16.7	2(2)	1(1)	50	1(1)	0	0
	Subtotal	98(206)	65(157)	27(32)	20	23(33)	8(9)	27.3	10(16)	5(7)	43.8
Yersinia	enterocolitica	57(100)	6(10)	0	0	22(34)	1(2)	5.9	29(56)	2(3)	5.4
	frederiksenii	7(9)	2(2)	0	0	4(6)	0	0	1(1)	0	0
	intermedia	3(3)	0	0	0	0	0	0	3(3)	0	0
	Subtotal	63(112)	8(12)	0	0	24(40)	1(2)	5	31(60)	2(3)	5
Escherichia	coli	27(39)	19(27)	13(19)	70.4	4(5)	2(2)	40	4(7)	1(2)	28.6
Citrobacter	freundii	14(15)	7(7)	0	0	3(4)	0	0	4(4)	2(2)	50
	braakii	6(9)	1(1)	0	0	3(5)	1(1)	20	2(3)	1(1)	33.3
	amalonaticus	1(2)	0	0	0	0	0	0	1(2)	0	0
	Subtotal	21(26)	8(8)	0	0	6(9)	1(1)	11.1	7(9)	3(3)	33.3
Klebsiella	pneumoniae	11(17)	1(1)	0	0	6(9)	1(1)	11.1	4(7)	0	0
	oxytoca	5(6)	2(2)	0	0	3(4)	1(1)	25	0	0	0
	Subtotal	16(23)	3(3)	0	0	9(13)	2(2)	15.4	4(7)	0	0
Raoultella	planticola	11(13)	7(9)	1(1)	11.1	2(2)	0	0	2(2)	0	0
	ornithinolytica	3(3)	1(1)	1(1)	100	2(2)	0	0	0	0	0
	Subtotal	14(16)	8(10)	2(2)	20.0	4(4)	0	0	2(2)	0	0
Enterobacter	cloacae	7(9)	2(2)	0	0	2(3)	1(1)	33.3	3(4)	2(3)	75
	aerogenes	2(2)	2(2)	0	0	0	0	0	0	0	0
	asburiae	2(2)	0	0	0	1(1)	1(1)	100	1(1)	1(1)	100
	amnigenus	1(1)	0	0	0	0	0	0	1(1)	1(1)	100
	Subtotal	12(14)	4(4)	0	0	3(4)	2(2)	50	5(6)	4(5)	83.3
Pseudomonas	aeruginosa	3(3)	0	0	0	3(3)	1(1)	33.3	0	0	0
	fluorescens	2(2)	0	0	0	2(2)	0	0	0	0	0
	luteola	1(1)	0	0	0	0	0	0	1(1)	0	0
	putida	1(1)	1(1)	0	0	0	0	0	0	0	0
	stutzeri	1(1)	1(1)	0	0	0	0	0	0	0	0
	Subtotal	8(8)	2(2)	0	0	5(5)	1(1)	20	1(1)	0	0
Pantoea	Pantoea spp.	6(7)	1(1)	1(1)	14.3	2(3)	1(1)	33.3	3(3)	3(3)	100
Ewingella	americana	3(4)	0	0	0	2(2)	1(1)	50	1(2)	0	0
Acinetobacter	baumannii	2(2)	0	0	0	0	0	0	2(2)	0	0
Kluyvera	intermedia	2(2)	0	0	0	2(2)	1(1)	0	0	0	0
Total		190(459)	79(224)	39(54)	24.1	60(120)	19(22)	18.3	51(115)	18(23)	20

Resistance breakpoints: Aeromonas and Pseudomonas, > 4μg/mL; others, > 2μg/mL.

Of the 206 isolates of Aeromonas spp., 117 Aeromonas sobria (A. sobria) isolates and 74 Aeromonas hydrophila (A. hydrophila) isolates were identified in 63 (20.3%) and 59 meat samples (19.0%), respectively (Table 1). In contrast, 100 Yersinia enterocolitica (Y. enterocolitica) isolates were identified in 57 meat samples (18.4%) and 39 E. coli isolates were identified in 27 meat samples (8.7%). In chicken meat samples, the isolation rates were as follows: A. sobria, 48.5%; A. hydrophila, 36.9%; E. coli, 18.4%; and Aeromonas veronii (A. veronii), 11.6%. In pork samples, the isolation rates were as follows: Y. enterocolitica, 21.4%; A. hydrophila, 13.6%; and A. sobria, 9.7%. In beef samples, the isolation rate of Y. enterocolitica was 27.9% and the isolation rates of other bacteria were less than 10%.

Colistin Resistance in the Isolated Bacteria

Of the 459 isolates, 99 (21.6%) isolates with MIC > 2 μg/mL and MIC > 4 μg/mL for Enterobacteriaceae and non- Enterobacteriaceae isolates respectively were resistant to CST (Table 1). These isolates included Aeromonas spp. (48/206, 23.3%), Yersinia spp. (5/112, 4.5%), E. coli (23/39, 59%), Citrobacter spp. (4/26, 15.4%), Klebsiella spp. (2/23, 8.7%), Raoultella spp. (2/16, 12.5%), Enterobacter spp. (7/14, 50%), Pseudomonas spp. (1/8, 12.5%), Pantoea spp. (5/7, 71.4%), Ewingella spp. (1/4, 25%), and Kluyvera spp. (1/2, 50%). CST-resistant bacteria were isolated from 39 of 103 chicken meat samples (37.9%), 19 of 103 pork samples (18.4%), and 18 of 104 beef samples (17.3%).

Antimicrobial Resistance Profiles of the Isolated Bacteria

The antimicrobial resistance profiles of the dominant isolates are shown in Table 2. In 117 isolates of A. sobria, resistance to TET (29 isolates, 23.5%) was predominantly observed, followed by resistance to NAL (15 isolates, 12.8%), CHL (4 isolates, 3.4%), and CTX (2 isolates, 1.7%). In 74 isolates of A. hydrophila, resistance to CTX (1 isolate, 1.4%), TET (13 isolates, 17.6%), GEN (1 isolate, 1.4%), KAN (1 isolate, 1.4%), NAL (7 isolates, 9.5%), CHL (1 isolate, 1.4%), and SXT (1 isolate, 1.4%) was observed. In 100 isolates of Y. enterocolitica, resistance to TET was observed in 2 isolates (2%) and resistance to NAL and SXT was observed in 1 isolate (1%) each. In contrast, E. coli isolates from chicken meat samples showed resistance to AMP (17 isolates, 63.0%), CFZ (9 isolates, 33.3%), CTX (3 isolates, 11.1%), CAZ (2 isolates, 7.4%), TET (18 isolates, 66.7%), GEN (2 isolates, 7.4%), KAN (10 isolates, 37%), NAL (14 isolates, 51.8%), CIP (5 isolates, 18.5%), LVF (5 isolates, 18.5%), CHL (4 isolates, 14.8%), and SXT (7 isolates, 25.9%). In addition, E. coli isolates from pork samples were resistant to AMP (2 isolates, 40%), CFZ (2 isolates, 40%), CAZ (2 isolates, 20%), TET (2 isolates, 40%), NAL (1 isolate, 20%), CIP (1 isolate, 20%), LVF (1 isolate, 20%), and CHL (2 isolates, 40%), and E. coli isolates from beef samples were resistant to AMP (4 isolates, 57.1%), CFZ (2 isolates, 28.6%), TET (2 isolates, 28.6%), CHL (2 isolates, 28.6%), and SXT (2 isolates, 28.6%). All bacterial isolates were susceptible to MEM and AMK.

Table 2.

 Antimicrobial resistance profile of bacteria species to other tested antimicrobial agents

Bacteria	Meat type	Total isolate	Antimicrobial agents: Isolate No (% resistant isolates)
			AMP	CFZ	CTX	CAZ	MEM	TET	GEN	KAN	AMK	NAL	CIP	LVF	CHL	SXT
A. sobria	C	101			2(2.0)	0	0	23(22.8)	0	0	0	14(13.7)	0	0	4(4.0)	0
	P	13			0	0	0	4(30.8)	0	0	0	1(7.7)	0	0	0	0
	B	3			0	0	0	2(66.7)	0	0	0	0	0	0	0	0
	Subtotal	117			2(1.7)	0	0	29(23.5)	0	0	0	15(12.8)	0	0	4(3.4)	0
A. hydrophila	C	44			0	0	0	9(20.5)	1(2.3)	1(2.3)	0	6	0	0	1(2.3)	1(2.3)
	P	18			1(5.6)	0	0	3(16.7)	0	0	0	0	0	0	0	0
	B	12			0	0	0	1(8.3)	0	0	0	1(8.3)	0	0	0	0
	Subtotal	74			1(1.4)	0	0	13(17.6)	1(1.4)	1(1.4)	0	7(9.5)	0	0	1(1.4)	1(1.4)
Y. enterocolitica	C	10			0	0	0	0	0	0	0	0	0	0	0	0
	P	34			0	0	0	2(5.9)	0	0	0	1(2.9)	0	0	0	1(2.9)
	B	56			0	0	0	0	0	0	0	0	0	0	0	0
	Subtotal	100			0	0	0	2(2)	0	0	0	1(1)	0	0	0	1(1)
E. coli	C	27	17(63.0)	9(33.3)	3(11.1)	2(7.4)	0	18(66.7)	2(7.4)	10 (37.0)	0	14(51.8)	5(18.5)	5(18.5)	4(14.8)	7(25.9)
	P	5	2(40)	2(40)	0	0	0	2(40)	0	0	0	1(20)	1(20)	1(20)	2(40)	0
	B	7	4(57.1)	2(28.6)	0	0	0	2(28.6)	0	0	0	0	0	0	2(28.6)	2(28.6)
	Subtotal	39	23(59.0)	13(33.3)	3(7.7)	3(7.7)	0	22(56.4)	2(5.1)	10 (25.6)	0	15(38.5)	6(15.4)	6(15.4)	8(20.5)	9(23.1)

Those printed in boldface are bacteria with intrinsic resistance to AMP and CFZ.

C, chicken; P, pork; and B, beef.

Characteristics of mcr-harboring Isolates

Of the 365 isolates with CST MICs of ≥1 μg/mL, 16 harbored the mcr gene. In chicken meat samples, one A. sobria isolate from 1 sample (1.0%) and 14 E. coli isolates from 10 samples (9.7%) were positive for the mcr-3 and mcr-1 genes, respectively. In pork samples, one A. sobria isolate was positive for the mcr-3 gene. The CST MIC of two A. sobria isolates carrying mcr-3 was 1 μg/mL, indicating that the isolates were susceptible to CST.

Of the 16 isolates carrying the mcr gene, 12 were selected for next-generation sequencing (NGS) because only one isolate was selected when two were isolated from a sample. The serotypes, multilocus sequence typing results, bacterial subtypes, and the locations of the mcr gene and other resistance genes are shown in Table 3. NGS analysis revealed that plasmidic mcr-1.1 and mcr-1.12 were found in nine and one E. coli isolates, respectively, from chicken meat samples and chromosomal mcr-3.25 in two A. sobria isolates from pork and chicken meat samples. In addition, several acquired resistance genes to β-lactams (blaTEM-1B), tetracyclines (tet(A) and tet(B)), aminoglycosides (aadA5, aph(3”)-Ib, aph(3’)-Ia, and aph(6)-Id), phenicols (catA1), quinolone (qnrS13), sulfonamides (sul1 and sul2), trimethoprim (dfrA1 and dfrA17), and macrolides (mdf(A)) were found in mcr-harboring E. coli. In addition, β-lactam resistance genes (ampS and blaFOX-4) and the TET resistance gene, tet (E), was found in one A. sobria isolate from pork samples.

Table 3.

 Microbiological and resistance profiles of bacteria carrying the mcr genes

Location	Strain No.		Bacteria	Serotype	Multilocus sequence typing	Resistance profile	mcr gene
							Subtype	Location
Kanto	CL-266	C	E. coli	O13/O135:H48	ST 10 (10-11-4-8-8-8-2)	CST	mcr-1.1	IncI2 Plasmid
	CL-276	C	E. coli	H34	ST2614(31-276-83-140-1-187-2)	TET-CST-NAL-SXT	mcr-1.1	IncI2 Plasmid
	CL-304	C	E. coli	H31	ST101 (43-41-15-18-11-7-6)	AMP-CFZ-TET-CST-KAN-NAL-CHL-SXT	mcr-1.1	IncI2 Plasmid
Chubu	CL-21, CL-22	C	E. coli	H27	ST1112 (10-11-5-10-8-1-2)	TET-CST-KAN-NAL	mcr-1.12	IncI2 Plasmid
	CL-25, CL-26	C	E. coli	O91:H28	ST135(13-39-50-13-16-37-25)	TET-CST	mcr-1.1	IncI2 Plasmid
	CL-480	C	E. coli	O81:H7	ST5826 (80-57-18-55-8-6)	AMP-TET-CST-NAL-SXT	mcr-1.1	IncI2 Plasmid
Kansai	CL-230, CL-231	C	E. coli	H52	novel (6-5-188-8-24-8-6)	TET-CST	mcr-1.1	IncI2 Plasmid
	CL-235	P	A. sobria			TET	mcr-3.25	Chromosome
	CL-184	C	E. coli	O91:H28	ST1196 (6-6-33-26-11-8-2)	AMP-CST-NAL-CPFX-LVFX	mcr-1.1	IncI2 Plasmid
Kyusyu	CL-859	C	E. coli	H5	ST206(6-7-5-1-818-2)	AMP-CST	mcr-1.1	IncI2 Plasmid
	CL-931, CL-933	C	E. coli	H5	ST206(6-7-5-1-818-2)	AMP-CST	mcr-1.1	IncI2 Plasmid
	CL-1133	C	A. sobria				mcr-3.25	Chromosome

Underlined strains were subjected to NGS analysis.

4. Discussion

The present study showed that domestic retail meats were contaminated with CST-resistant bacteria with or without mcr genes. The mcr-1 gene was detected in E. coli isolated from chicken meat samples (10/103, 9.7%) but not in E. coli isolated from pork or beef samples. In addition, chromosomal mcr-3 was found in A. sobria isolated from chicken meat and pork samples. In a previous study, an E. coli strain carrying the mcr-1 gene was found in domestic retail chicken meat (8/154, 5.2%) and imported retail pork (1/55, 1.8%) samples in Japan[24]). The use of CST as a feed additive in food-producing animals, including poultry, has been prohibited since July 2018 in Japan; however, the therapeutic use of CST is approved in cattle and pigs. In the current study, the presence of the mcr-1 gene was observed in CST-resistant E. coli only from chicken meat samples. In Japan, the mcr-1 gene was detected in E. coli from healthy cattle (5/3,134, 0.2%), pigs (20/2,052, 1.0%), and broiler chickens (14/2,017, 0.7%) in 2000–2014[20]). Despite the low prevalence of the mcr-1 gene, the frequent contamination of poultry carcasses with intestinal contents during the slaughter process has been reported[24]). A study conducted in Bangladesh found high prevalence (25%) of the mcr-1 gene in E. coli isolated from broiler chickens, and this finding was associated with high CST usage for treatment and prophylaxis purposes[32]). In addition, a survey in the Netherlands detected the mcr-1 gene (8%) in chicken[15]). In the current study, more than 60% of the mcr-1 gene-containing bacteria isolated from chicken meat samples exhibited multidrug resistance. The contamination of meat with multidrug-resistant bacteria that can be transmitted to humans through meat consumption is a potential public health risk.

Furthermore, A. sobria, A. hydrophila, and A. caviae are some of the most important pathogens that cause gastroenteritis and wound infections in humans. In Japan, foodborne disease outbreaks caused by A. hydrophila or A. sobria are regulated by the Food Sanitation Act (Act No. 233 of 1947). CST resistance was observed in 48 Aeromonas spp. isolated from 32 chicken meat, 9 pork, and 7 beef samples, but these isolates did not carry the mcr genes tested. The mcr-3 gene was detected in two A. sobria isolates, from chicken and pork samples, with a CST MIC of 1 μg/mL. The mcr-3 gene was first reported in E. coli isolated from pigs[7]) and has been reported to be responsible for CST resistance in Aeromonas spp. isolated from meat and environmental water samples[18],[33]). In Japan, current reports on the detection of the mcr gene from livestock and meat are limited to E. coli and Salmonella[20],[21],[22],[23],[24],[34]). However, because the mcr gene has been found in other gram-negative bacteria, further investigations on the prevalence of mcr genes in different bacteria are necessary.

Bacterial contamination of retail meats is an important issue that depends on various factors, including the differences in intestinal flora of animals and treatment processes in slaughterhouses[35]). Aeromonas spp. and E. coli were frequently encountered in the present study, which is in accordance with food poisoning cases reported in Japan that were associated with Aeromonas spp. and E. coli isolated from chickens[36]). Among the 14 CST-resistant E. coli isolates harboring mcr-1, some exhibited multidrug resistance, which can be attributed to the acquisition of resistance genes (Table 3). The CST-resistant E. coli isolates with the mcr-1 genes were susceptible to 3rd generation cephalosporins (CTX and CAZ) and carbapenem (MEM), suggesting that these antimicrobial agents may be an appropriate treatment for infections caused by these CST-resistant E. coli strains. However, there is a potential risk of horizontal transfer of the mcr-1 plasmid to other relevant E. coli possibly carrying extended spectrum beta-lactamase and carbapenemase resistance genes in humans. Hence, contamination of meat with such multidrug-resistant bacteria possessing mcr genes highlights the risk of resistance gene transmission to humans through meat consumption. Colonization of the human intestinal tract with multidrug-resistant bacteria can result in infections with limited treatment options.

The other gram-negative bacteria belonging to the Enterobacteriaceae family isolated in this study are part of the normal flora of the intestinal tract or environment of animals and have been associated with opportunistic infections in humans[37]). In addition, Y. enterocolitica is a causative agent of food poisoning, is widely distributed in nature, and inhabits livestock, wild animals, water, and soil[38]). Moreover, Y. enterocolitica serogroups O3 and O9 are pathogenic to humans[39]), although the serogroups of the isolates were not investigated in this study. In this study, using CST-DHL media, we isolated CST-resistant bacteria among retail meat. However, we were not able to determine the mechanism underlying CST resistance in bacteria lacking the mcr gene. Therefore, further research is needed to elucidate the mechanism behind CST resistance in mcr-negative bacteria.

In conclusion, we detected CST-resistant bacteria in retail meats and mcr-1 and mcr-3 genes in E. coli and A. sobria, respectively. The findings of this study highlight the necessity of surveillance of CST resistance and resistance genes in bacteria that contaminate retail meats.

Table 3.

 (Continued)

Strain No.	Resistance genes against
	Aminoglycoside	β-lactam	Phenicol	Trimethoprim	Macrolide	Quinolone	Sulfonamide	Tetracycline
CL-266					mdf(A)
CL-276	aadA5, aph(3”)-Ib,aph(6)-Id			dfrA17	mdf(A)	gyrA(S83L)	sul2	tet(B)
CL-304	aph(3”)-Ib, aph(3’)-Ia, aph(6)-Id	blaTEM-1B	catA1	dfrA1	mdf(A)	gyrA(S83L)	sul1, sul2	tet(B)
CL-21, CL-22	aph(3”)-Ib, aph(3’)-Ia, aph(6)-Id				mdf(A)	gyrA(S83L)		tet(B)
CL-25, CL-26	aph(3”)-Ib, aph(6)-Id				mdf(A)			tet(A)
CL-480		blaTEM-1B		dfrA1	mdf(A)	gyrA(S83L)	sul1	tet(A)
CL-230, CL-231					mdf(A)	qnrS13	sul2	tet(A)
CL-235		ampS, blaFOX-4						tet(E)
CL-184		blaTEM-1B			mdf(A)	gyrA(S83L, D87N)		tet(A)
CL-859		blaTEM-1			mdf(A)
CL-931, CL-933		blaTEM-1			mdf(A)
CL-1133		ampS, blaFOX-7