High incidence of multidrug-resistant Escherichia coli coharboring mcr-1 and bla_CTX-M-15 recovered from pigs

Abstract

Purpose

The coexistence of mobile colistin (COL)-resistant gene mcr-1 with extended-spectrum beta-lactamase (ESBL) gene in Escherichia coli has become a serious threat globally. The aim of this study was to investigate the increasing resistance to COL and in particular its coexistence with ESBL-producing E. coli recovered from pig farms in China.

Materials and methods

E. coli were isolated from 14 pig farms in Jiangsu China. Susceptibility testing was identified by micro-dilution method. PCR assay and nucleotide sequencing were used to detect COL-resistant genes, mcr-1 to −5, as well as ESBL genes, bla_CTX-M, bla_SHV and bla_TEM. Conjugation experiment, plasmid replicon typing of the multidrug resistance (MDR), S1-PFGE and DNA southern hybridization were performed to study the transferability of these genes.

Results

Overall, 275 E. coli isolates were recovered from a total of 432 cloacal and nasal swabs. More than 90% of the isolates were MDR, of which 70.18% were resistant to COL. Of these 275 isolates, mcr-1 was identified as the most predominant gene carried by 71.63% (197/275) of isolates, 39.59% (78/197) of the isolates were harboring both mcr-1 and ESBL genes (bla_CTX-M, bla_SHV and bla_TEM). ESBL genotyping showed that bla_CTX-M was the most predominant ESBL (68.49%) followed by bla_SHV (16.4%) and bla_TEM (15%). Sequencing revealed that the most common variants of bla_CTX-M identified were, bla_CTX-M-15 (69%), bla_CTX-M-55 (29%) and bla_CTX-M-1 (1.8%). IncHI2, IncFIB, IncFIC, IncN and IncX4 were found to be the most common Inc-types found both in donors and in transconjugants and were associated with the transfer of the mcr-1 and ESBL encoding genes. Six strains carried a total of five different plasmids: approximately 97-, 130-, 160-, 227- and 242-kb plasmids.

Conclusion

The coexistence of the mcr-1- and bla_CTX-M-15-carrying isolates displaying high MDR, recovered from E. coli of pig origin, is a major concern for both humans and veterinary medicine.

Keywords:

• E. coli
• colistin
• mcr-1
• ESBL
• coexistence

Introduction

Antimicrobial resistance (AMR) has now been widely recognized as a crucial threat to human and animal health as the extensive use of antimicrobials in humans as well as in food-producing animals.¹ Global consumption of antimicrobials in animal settings may rise up to 67% by 2030, determined predominantly by BRICS (Brazil, Russia, India, China and South Africa) countries, as large-scale and intensive farming operations are greatly in demand with the upsurge in revenue and animal protein consumption.^2,3 This heavy antimicrobial practice creates a selective pressure that contributes to the emergence and spread of bacterial resistance. One of the major concerns is the rapid increase of the multidrug-resistant (MDR) Escherichia coli in animal settings and clinical medicine.^4–6 This is not only because of the lessened number of useful antimicrobials for curing MDR E. coli infections, but also due to the potential transfer of MDR E. coli strains from animals to humans, especially which producing extended-spectrum β-lactamases (ESBLs) and carbapenemases, and display resistance to colistin (COL).⁷

ESBLs are β-lactamases that confer resistance to oxyimino “second- and third-generation” cephalosporin’s (eg, cefotaxime (CTX), ceftriaxone and ceftazidime) and aztreonam.^8,9 ESBL-producing bacteria were first reported in 1980, soon after the introduction of the third-generation cephalosporin’s (CTX and ceftiofur (CEF) into clinical settings.¹⁰ Currently, there are more than 350 ESBL genes that have been reported, and these genes are commonly developed through point mutations of the classical SHV-1 and TEM-1 β-lactamases and more increasingly prominent the CTX-M types.^11–13 Among the CTX-M enzymes, bla_CTX-M-55 has become the leading CTX-M type in ESBL-producing E. coli isolates of animal origin during the last decade.^14,15 In contrast, bla_CTX-M-15 seems to be the most extensive types in isolates of human origin.¹⁶ ESBL-producing E. coli are highly linked with multiple plasmids and studies have reported that ESBL genes are often carried on IncF, IncI1, IncN, IncHI1 and IncHI2 in food-producing animals worldwide.^13,17–19 There is potential for ESBL genes/plasmid spreading between E. coli from animals, food and humans.^20,21

The co-occurrence of ESBL genes and mcr-1 in E. coli was reported from China in 2016.²² Rhouma and Letellier assumed that a historic relation existed between ESBL genes, carbapenemase genes and mcr-1.²³ A recent study proposed that cephalosporin resistance is commonly spread in animals and humans through distinct plasmids.²⁴ It is highly expected that food-producing animals have become the most significant reservoirs in disseminating these resistance genes in the community through horizontal gene transfer. To assess the co-occurrence and emergence of mcr and ESBL genes in E. coli of pig source, we examined 14 pig farms of Jiangsu province in China to evaluate the current scenario of these resistant genes in pigs and further clarified the predominant genotype and plasmids diversity of mcr and ESBL genes.

Materials and methods

Collection of samples

A total of 432 samples (400 from healthy and 32 from dead pigs) were collected from 14 commercial pig farms in Jiangsu, China (Figure S1), during the period of August 2016 until December 2017. From each farm, samples were randomly collected. The anal swabs were collected by inserting the swab into the rectum and being rotated. To collect nasal swabs from swine, the nose was wiped with a piece of paper and a sterile swab was inserted into the nasal cavity and rotated for 3 s at 90°. From 32 dead pigs, all samples were aseptically obtained from different organs. All collected samples were immediately transported at 4°C to the laboratory for microbial examination and processed within 4 hrs.

Isolation and screening of ESBL-producing and COL-resistant E. coli

All samples were directly streaked onto MacConkey agar (Binhe Microorganism Reagent Co. Ltd., Hangzhou, China) supplemented with CTX (1 µg/mL) and COL (2 µg/mL) for the screening of possible ESBL-producing and COL nonsensitive E. coli as previously described.²⁵ Plates were incubated at 37°C for 18–24 hrs. Presumptive E. coli colonies with dark pink to red colors were confirmed microscopically and further verified by species-specific PCR as described previously.²⁶ Confirmed E. coli strains were stored in Luria–Bertani medium (Oxoid, United Kingdom) containing 40% (vol/vol) glycerol in aliquots at −80°C until further use.

ESBL-producing E. coli were further confirmed by double-disk synergy (DDS) testing as recommended by the Clinical and Laboratory Standards Institute (CLSI) guideline, using antibiotic discs of ceftazidime (30 µg), ceftazidime plus clavulanic acid (30/10 µg), CTX (30 µg) and CTX plus clavulanic acid (30/10 µg). DDS test was performed for phenotypic detection of ESBLs. The test result is considered as positive if the zone of inhibition is ≥5 mm larger with clavulanic acid than without.²⁷

Antimicrobial susceptibility testing

Antimicrobial susceptibility was performed by minimum inhibitory concentration (MIC) determination using broth micro-dilution method against 11 antibiotics for all 275 isolates and 17 antibiotics for transconjugants. The MIC data was interpreted according to the CLSI recommendations.²⁷ Antibiotics used in this study, comprised of 5 β-lactams – ampicillin (AMP), CTX, cefoxitin (CFX), CEF and meropenem (MEM) – and 12 non-β-lactams – COL, ciprofloxacin (CIP), chloramphenicol (CHL), enrofloxacin (ENR), gentamycin (GEN), kanamycin, nalidixic acid, polymyxin-B (POL-B), tetracycline (TET), trimethoprim, streptomycin and sulfamethoxazole. The MIC of COL was determined by broth micro-dilution method recommended by the joint CLSI-EUCAST polymyxin breakpoints working group (www.EUCAST.org), CLSI VET01-A4 is used for CEF and ENR which are missing in the human CLSI M100-S27. E. coli ATCC 25922 was used as a quality control in antimicrobial susceptibility testing. Isolates that exhibited resistance to more than 3 antimicrobial agents were classified as MDR.²⁸

PCR assays for detection of mcr and ESBL genes

PCR assay was used to detect COL-resistant genes mcr-1 to mcr-5 as well as ESBL genes (bla_CTX-M, bla_SHV and bla_TEM). Total DNA was isolated by conventional boiling method. All these resistant genes were screened via PCR-based diagnostics with specific primers, as previously described. All the primers and PCR conditions used in this study are listed in Table 1. All PCR positive amplicons of these targeted genes were sequenced by Sanger sequencing in TSINGKE Corporation (Nanjing, PR China).

Table 1 PCR primers and conditions used in this study

Primer name	PCR target	Sequence (5′- 3′)	Annealing temperature(°C)	Product size (bp)	Reference
E. coli-specific
UAL UAR	uidA	TGGTAATTACCGACGAAAACG GC ACG CGT GGT TAC AGT CTT GCG	62	147	^Citation58
ESBL genes
CTX-MA CTX-MB	blaCTX-M	CGC TTT GCG ATG TGC AG ACC GCG ATA TCG TTG GT	54	550	^Citation58
SHV-F SHV-R	blaSHV	GGG TTA TTC TTA TTT GTC GC TTA GCG TTG CCA GTG CTC	58	930	^Citation58
TEM-F TEM-R	blaTEM	ATA AAA TTC TTG AAG ACG AAA GAC AGT TAC CAA TGC TTA ATC	56	1086	^Citation58
MCR genes
CLR5-F CLR5-R	mcr-1	CGG TCA GTC CGT TTG TTC CTT GGT CGG TCT GTA GGG	58	309	^Citation41
MCR-2-F MCR-2-R	mcr-2	TGT TGC TTG TGC CGA TTG GA AGA TGG TAT TGT TGG TTG CTG	58	567	^Citation42
MCR-3-F MCR-3-R	mcr-3	TTG GCA CTG TAT TTT GCA TTT TTA ACG AAA TTG GCT GGA ACA	50	542	^Citation43
MCR-4F MCR-4R	mcr-4	ATT GGG ATA GTC GCC TTT TT TTA CAG CCA GAA TCA TTA TCA	58	487	^Citation44
MCR-5-F MCR-5-R	mcr-5	ATG CGG TTG TCT GCA TTT ATC TCA TTG TGG TTG TCC TTT TCT G	50	1644	^Citation45

Conjugation experiment

To determine the transferability of resistance genes, 15 COL-resistant E. coli isolates were selected as donors for conjugation. E. coli EC-600 (Nal^R, Rif^R) was used as recipient bacteria. Conjugation experiments were performed as previously described.²⁹ These putative transconjugants were further confirmed using antibiotic susceptibility testing, PCR detection and plasmid incompatibility (Inc) groups typing carried by the transconjugants.

Plasmid replicon typing

Plasmid DNA was extracted from both donors and transconjugants using the Wizard Genomic DNA Purification kit (Promega) and was characterized by PCR-based replicon typing method (PBRT). Eighteen pairs of primers were designed to perform 5 multiplex and 3 simples PCRs targeting the FIA, FIB, FIC, HI1, HI2, IncI1, L/M, N, P, W, T, A/C, K, B/O, X, Y, F and FIIA replicons as previously mentioned.³⁰ While primers for two other plasmids IncI2 and IncX4 were designed separately which were missing in the previous replicon typing. All the PBRT primers and PCR conditions used in this study are listed in the Supplementary Table S1. PCR amplicons of plasmids were sequenced by Sanger sequencing (TSINGKE Corporation, Nanjing, PR China) and retrieved sequences were used to confirm replicon types by using BLAST tool available at NCBI web (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

Pulsed field gel electrophoresis (PFGE) and Southern hybridization

To determine the genetic relatedness and location of transmissible mcr-1-positive elements, the six conjugative E. coli strains were characterized by S1-PFGE and Southern hybridization using a probe specific for mcr-1. Genomic DNA from each of the isolate was digested with S1 nuclease (Thermo scientific) and was examined by PFGE as previously described.³¹

Southern hybridizations of plasmid DNA were performed with a digoxin-labeled mcr-1-specific probe according to the manufacturer’s instructions (Roche Diagnostics, Mannheim, Germany) as previously described.³²

Statistical analysis

Differences in the AMR profiles of E. coli isolates with or without mcr-1 were assessed by a two-tailed Chi-square test or Fisher’s exact test using the Statistical Packages of Social Sciences software for Windows, version 20.0 (IBM Corp., Armonk, NY), with P<0.05 set as the level of significant differences.

Results

Bacterial isolation and antimicrobial susceptibility

Overall, 275 E. coli isolates (243 from healthy and 32 from dead pigs) were recovered from 432 samples of 14 different pig farms. All isolates were observed to be ESBL-producing and COL nonsensitive as determined by phenotypic approaches, giving a carriage rate of 63.6% (275/432). Of these 275 E. coli isolates, 174 (63.27%) were from feces, 69 (25.09%) were from nasal and remaining 32 (11.63%) were from diseased and dead pigs.

The MICs were obtained from the antibiotic susceptibility testing for all isolates. To determine the resistance profiles of 275 representative E. coli strains, susceptibility of 11 antibiotics were used (Table 2). Of the 275 E. coli isolates, the resistant rate to TET was 97.81%, followed by AMP (96.72%), CHL (94.54%), CFX (86.18%), CTX (78.18%), CEF (77.81%), CIP (73.81%), POL-B (71.27%), GEN (70.54%), and COL (70.18%). In contrast, the most effective antibiotic against these isolates was MEM with 99.6% susceptibility. Majority of the E. coli strains showed considerable MDR to β-lactams and several non-β-lactams groups, including polypeptides group, fluoroquinolones, aminoglycosides, amphenicol, quinolone group, sulfonamides and TET.

Table 2 Distribution of MICs of 11 antibiotics for 275 MDR Escherichia coli isolates

Antibiotics	MIC (mg/L)
	0.008	0.016	0.032	0.063	0.125	0.25	0.5	1	2	4	8	16	32	64	128	256	>256	MIC50	MIC90	Resistance
Ampicillin	-	-	-	-	-	-	-	0	1	1	1	6	4	12	23	38	189	>256	>256	96.72%
Cefotaxime	-	-	-	0	8	9	11	14	18	15	14	19	31	45	44	33	14	32	256	78.18%
Cefoxitin	-	-	-	-	0	3	2	11	3	1	18	72	51	56	30	9	19	32	128	86.18%
Ceftiofur	-	-	-	-	0	3	21	23	14	12	26	19	28	33	56	31	9	32	256	77.81%
Chloramphenicol	-	-	-	-	-	-	0	0	2	1	7	5	17	34	35	61	113	256	>256	94.54%
Ciprofloxacin	-	0	15	19	5	6	3	14	10	25	26	42	47	38	23	2	-	16	64	73.81%
Colistin	-	-	0	9	29	8	14	22	29	110	41	7	4	2	0	-	-	4	8	70.18%
Gentamycin	-	-	-	0	12	14	9	17	13	4	12	25	54	56	27	16	16	32	256	70.54%
Meropenem	0	16	158	75	10	4	3	2	3	1	0	0	0	-	-	-	-	0.032	0.0625	0.36%
Polymyxin-B	-	-	-	3	12	12	15	37	41	87	49	16	2	1	0	-	-	4	8	71.27%
Tetracycline	-	-	-	-	-	-	0	3	0	3	0	1	16	41	47	69	96	256	>256	97.81%

Notes: Red vertical lines indicate the breakpoints between intermediate and resistant values. White areas indicate range of tested dilutions for each antibiotic; the MIC₅₀ and MIC₉₀ values are concentrations at which ≥50% and ≥90% of isolates are inhibited.

Abbreviations: MIC, minimum inhibitory concentration.

mcr-1 and ESBL genes are prevalent among E. coli strains

Although 70.18% (193 of 275) swine E. coli isolates conferred resistance to COL, the mcr-1 carriage rate was 71.63% (197/275) (Table 3) and only mcr-1 gene was detected in these COL-resistant E. coli isolates. No other colistin-resistant gene (mcr-2 to mcr-5) could be detected in the study population of E. coli isolates. The mcr-1 gene was detected in all farms, and the prevalence rate was enormously high 71.6%, ranging from 47.8% to 100% in different farms, while 40.6% in diseased isolates (Table 3).

Table 3 Prevalence of mcr-1 & or ESBL-producing E. coli in swine samples collected from different farms of Jiangsu China

Farm numbers	No. of positive E. coli samples	No. of mcr-1 positive E. coli isolates (%)	No. of mcr-1 & ESBL producing E. coli isolates (%)
Farm 1	32	22 (68.7)	10 (31.2)
Farm 2	11	7 (63.6)	3 (27.2)
Farm 3	10	8 (80.0)	4 (40.0)
Farm 4	26	18 (69.2)	4 (15.3)
Farm 5	14	14 (100)	2 (14.2)
Farm 6	14	13 (92.8)	2 (14.2)
Farm 7	7	5 (71.4)	N.D.
Farm 8	14	12 (85.7)	6 (42.8)
Farm 9	16	14 (87.5)	9 (56.2)
Farm 10	43	37 (86.0)	11 (25.5)
Farm 11	23	11 (47.8)	5 (21.7)
Farm 12	12	9 (75.0)	6 (50)
Farm 13	14	7 (50.0)	4 (28.5)
Farm 14	7	7 (100.)	5 (71.4)
Diseased	32	13 (40.6)	7 (21.8)
Total	275	197	78

Abbreviations: ESBL, extended-spectrum β-lactamase; N.D., not determined.

We recovered a total of 146 (53.09%) ESBL-producing E. coli strains from 275 samples collected from 14 different farms of pigs (Figure S2 and Table 5). Among the 146 ESBL-producing isolates, 68.49% (100/146) harbored bla_CTX-M genes, 22.60% (33/146) harbored bla_SHV, while 18.49% (27/146) were carrying bla_TEM genes. Among them, 6 isolates carried bla_CTX-M and bla_SHV, 3 isolates contained bla_CTX-M and bla_TEM, and one isolate had three genes of bla_CTX-M, bla_SHV, and bla_TEM.

The resistance patterns of E. coli are different in mcr-1-positive and mcr-1-negative isolates

The mcr-1-positive E. coli isolates displayed more resistance to other antimicrobials than those of mcr-1-negative isolates (Table 4). For mcr-1-positive E. coli, all isolates, 100% (197/197), possessed not less than 3 antibiotics resistance pattern, while about 96.15% (75/78) of COL-negative isolates did (P=0.006), with 3 COL-sensitive isolates which displayed resistance to not more than 2 antibiotics. In addition, about two-thirds of mcr-1-positive E. coli isolates, 68.52% (135/197), showed resistance to at least 9 drugs, but only 3.84% (3/78) of mcr-1-negative isolates did (P=0.000). Interestingly, 68.02% (134/197) of mcr-1-positive strains presented resistance profiles to 10 drugs, but no COL-negative E. coli did (P=0.000).

Table 4 The resistance patterns of E. coli with or without mcr-1

Resistance profiles	mcr-1 positive (n=197)	mcr-1 negative (n=78)	Chi-square (P-value)
≥11 drugs	0 (0)	0.0 (0)	-
≥10 drugs	68.02 (134)	0.0 (0)	0.000
≥9 drugs	68.52 (135)	3.84 (3)	0.000
≥8 drugs	81.21 (160)	58.97 (46)	0.000
≥7 drugs	89.84 (177)	76.92 (60)	0.005
≥6 drugs	93.90 (185)	88.46 (69)	0.125
≥5 drugs	99.49 (196)	91.02 (71)	0.000
≥4 drugs	100.0 (197)	93.58 (73)	0.000
≥3 drugs	100.0 (197)	96.15 (75)	0.006
≥2 drugs	100.0 (197)	98.71 (77)	0.111

Coexistence of mcr-1 and ESBL genes screened among the strains

Based on the results of this study, among 197 mcr-1 positive isolates, 39.59% (78/197) were identified as carrying both mcr-1 and ESBL genes. Distribution of all mcr-1 positive E. coli isolates (n=78) harboring ESBL genes are analyzed and presented in Figure 1 and Table 5. Our findings indicate that the combination of mcr-1 with bla_CTX-M was the most prevalent with the rate of 70.51% (55/78) followed by the combination of mcr-1 and bla_SHV (14.10%, 11/78). Finally, 7.69% (6/78) isolates were identified carrying both mcr-1 and bla_TEM. Furthermore, combination of mcr-1 with two or more than two ESBL genes was also identified. Results showed that a total of 5.12% (4/78) of mcr-1-positive isolates also carried bla_CTX-M and bla_SHV, while 1.28% (1/78) of mcr-1-positive isolates carried bla_CTX-M and bla_TEM. Interestingly, a single isolate was carrying (mcr-1+ bla_CTX-M + bla_SHV+ bla_TEM). In this study, mcr-1 and bla_CTX-M were identified as the dominant genes (Table 5).

Table 5 Distribution of various resistance genes among 275 MDR Escherichia coli isolates from pigs

Colistin resistant gene	Extended-spectrum beta-lactamase genes			No. of isolates
mcr-1	blaCTX	blaSHV	blaTEM
+				119
				1
	+			34
		+		13
			+	15
+	+			55
+		+		11
+			+	6
+	+	+	+	1
+	+	+		4
+	+		+	1
	+	+	+	1
	+	+		2
	+		+	2
		+	+	1
N.D.	N.D.	N.D.	N.D.	9
197	100	33	27	275

Abbreviations: +, positive; N.D., not determined.

Figure 1 Distribution of various resistance genes in combination.

As mcr-1 in combination with bla_CTX-M were identified as the most prevalent (70.51%, 55/78), we further sequenced the bla_CTX-M genes to explore the subtypes. Sequencing analysis of these 55 bla_CTX-M isolates showed that all these mcr-1-positive isolates were harboring bla_CTX-M-1 group. The most prevalent variants identified in these 55 isolates belonged to this group were bla_CTX-M-15 in 38/55 (69%) isolates, followed by bla_CTX-M-55 in 16/55 (29%) isolates and bla_CTX-M-1 in one isolate (1.8%).

mcr-1 and ESBL genes could be conjugative transfer by plasmids with different replicon type

Conjugation experiments were performed on random 15 mcr-1 positive isolates. Of the 15 mcr-1 resistant isolates, 10 isolates were carrying additional bla-_CTX-M, while a single isolate was harboring bla_TEM. Of these 15 isolates, 12 were successfully transferred to E. coli EC-600. The resistance profiles of the 12 transconjugants were identical to those of the mcr-1 and bla_CTX-M carrying E. coli donor isolates, indicating the transfer of antibiotic resistance. In addition, resistant to several non-β-lactam antibiotics, such as aminoglycosides, fluoroquinolones, TET, macrolides and sulfonamides, were also co-transferred along with COL and β-lactam resistance. MICs of COL of these transconjugants revealed 4- to 8-fold increase as compared with the recipient EC-600 (0.125 µg/mL).

PCR-based replicon typing (PBRT) showed that in the E. coli isolates carrying mcr-1 and bla_CTX-M, the plasmids with different replicons, including IncHI2 (n=7), IncFIB (n=7), IncFIC (n=4), IncP (n=4), IncFrepB (n=4), IncN (n=3), IncX4 (n=2) IncY (n=2) and IncI1 (n=1), were detected in the donor strains (Figure 2 and Table 6). However, PBRT of the transconjugants confirmed only five replicons, IncHI2, IncFIB, IncFIC, IncN and IncX4, which were present in both donors and transconjugants and were associated with the transfer of the mcr-1 and ESBL genes.

Table 6 Conjugation experiments and plasmid replicon type detection for 15 mcr-1-positive E. coli

Strain	CTX MIC (mg/L)	COL MIC (mg/L)	Plasmid (Inc) types	Resistance genes	Resistance profile for non-beta-lactam antibiotics GEN ENR SUL CIP KEN CHL TET TRM STR POL-B
EC-9	128	16	HI2,I1,Y, FIC	MCR-1, CTX-M-15	GEN, ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, STR,POL-B
EC-9-T	128	4	HI2, FIC	MCR-1, CTX-M-15	GEN, ENR, NAL, SUL, KEN, CHL, TET, TRM, POL-B
EC-37	256	8	N, P	MCR-1, CTX-M-55	NAL, SUL, CFX, CEF, KEN, CHL, TET, TRM, STR, POL-B
EC-37-T	32	4	N	MCR-1, CTX-M-55	NAL, SUL, CFX, CEF, KEN, CHL, TET, TRM, STR, POL-B
EC-48	256	4	FIB, P, FrepB, N	MCR-1, CTX-M-55	GEN, ENR, NAL, SUL, KEN, CHL, TET, TRM, STR,POL-B
EC-48-T	32	2	FIB, N	MCR-1, CTX-M-55	GEN, ENR, NAL, SUL, KEN, CHL, TET, TRM, STR,POL-B
EC-52	128	4	FIB,N, P	MCR-1, CTX-M-55	ENR, NAL, SUL, CEF, KEN, CHL, TET, TRM, STR,POL-B
EC-52-T	256	2	FIB,N	MCR-1, CTX-M-55	ENR, NAL, SUL, CEF, KEN, CHL, TET, TRM, STR
EC-29	256	8	HI2, FIB	MCR-1, CTX-M-15	GEN, NAL, SUL, CFX, CEF, KEN, CHL, TET, TRM, STR,POL-B
EC-29-T	0.5	2	HI2, FIB	MCR-1, CTX-M-15	GEN, NAL, SUL, KEN, CHL, TET, TRM, POL-B
EC-34	256	8	FIB, HI2, FrepB	MCR-1, CTX-M-15	GEN, NAL, SUL, CEF, KEN, CHL, TET, TRM, STR,POL-B
EC-34-T	>256	8	HI2,FIB	MCR-1, CTX-M-15	GEN, NAL, SUL, CEF, KEN, CHL, TET, TRM, POL-B
EC-40	128	4	HI2, FIB, FrepB	MCR-1, CTX-M-15	ENR, NAL, SUL, CFX, CEF, CIP, KEN, CHL, TET, TRM, POL-B
EC-40-T	64	2	HI2, FIB	MCR-1, CTX-M-15	ENR, NAL, SUL, CFX, CEF, KEN, CHL, TET, TRM, POL-B
EC-1	256	4	X4, FIC	MCR-1, TEM	GEN, ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, STR, POL-B
EC-1-T	256	2	X4, FIC	MCR-1, TEM	GEN, NAL, SUL, CEF, CIP, KEN, CHL, TET, TRM, STR, POL-B
EC-2SF	256	2	HI2,FIB, P	MCR-1, CTX-M-15	GEN, ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, STR,POL-B
EC-2SF-T	>256	4	HI2, FIB	MCR-1, CTX-M-15	ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, STR,POL-B
EC-55	64	8	HI2,FIB, FIC,Y	MCR-1, CTX-M-55	ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, STR,POL-B
EC-55-T	64	4	HI2,FIC	MCR-1, CTX-M-55	ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, POL-B, STR
EC-20	2	4	X4, FIC, FrepB	MCR-1	GEN, ENR, NAL, SUL, CIP, KEN, CHL, TET, TRM, POL-B
EC-20-T	0.25	4	X4, FIC	MCR-1	GEN, NAL, SUL, KEN, CHL, TET, TRM, STR,POL-B
EC-19	128	32	HI2	MCR-1	GEN, NAL, SUL, CIP, KEN, CHL, TET, TRM, POL-B
EC-19-T	256	4	HI2	MCR-1	GEN, NAL, SUL, KEN, CHL, TET, TRM, POL-B
EC-600	0.25	0.125	ND	ND

Abbreviations: COL, colistin; CTX, cefotaxime; GEN, gentamycin; ENR, enrofloxacin; NAL, nalidix-acid; SUL, sulfamethaxozole; CIP, ciprofloxacin; KEN, kenamycin; CHL, chloramphenicol; TET, tetracycline; TRM, trimethropim; STR, streptomycin; POL-B, polymyxin-B; ND, not determined.

Figure 2 Detection of plasmid replicon types in multidrug-resistant E. coli using PCR assay.

S1-PFGE analysis demonstrated that these six strains carried multiple plasmids varying in sizes ranging from ~97 kb to 242 kb (Figure 3A). Southern hybridization assay confirmed that the mcr-1 gene recovered from these six strains was positioned on the following five different types of plasmids: with the size of approximately 97, 130, 160, 227 and 242 kb, respectively (Figure 3B).

Figure 3 (A) S1-nuclease pulsed-field gel electrophoresis profiles of six E. coli and (B) Southern Blot hybridization of six E. coli carrying the mcr-1 plasmid.

Notes: M – Salmonella H9812 (15.0–242.5 kb); 1: E-55; 2: E-01; 3: E-09; 4: E-02; 5: E-44; and 6: E-37. The arrows in the figure represent the position of the mcr-1 plasmid.

Discussion

China alone produces and consumes roughly half the planet’s pigs, about 500 million annually, and has been the leading consumption of antibiotics in the world.³³ The increased usage of antibiotics may trigger the emergence of AMR. Reports on emergence of AMR particularly resistance of β-lactam and COL are increasing all over the world.^34–37 The prevalence of ESBL in animal origin has been rising since 2003, with slight variances amongst terrestrial regions and different animal species.^4,14,38 We report on the high incidence of mcr-1-carrying ESBL-producing E. coli recovered from pigs in Jiangsu, China. Our results indicated that all the mcr-1 and ESBL-producing E. coli isolates showed MDR. The majority of these isolates (77–86%) showed resistance to cephalosporin (Table 2). In addition, high resistance was also observed to common β-lactam and non-β-lactam antimicrobials such as AMP, fluoroquinolones, aminoglycosides, amphenicol, quinolones, sulfonamides and TET which are commonly using in human as well in veterinary practice. Many recent studies have reported MDR ESBL-producing E. coli isolated from poultry,³ pigs,³⁹ cattle³⁶ and humans.⁴⁰

Recently, plasmid-mediated COL-resistant genes mcr-1 to mcr-8 have been widely discovered around the world.^34,41–47 Herein, we screened 275 MDR E. coli isolated from 14 pig farms from Jiangsu province for the presence of mcr-1 to mcr-5 genes. Only mcr-1 gene was detected in isolates from every farm, and the carriage rate was extremely high in 71.6% (197/275). The high prevalence rate of the mcr-1 found in this study from pigs in Jiangsu is consistent with very recent reports from China.^48,49 These recent studies in pigs reported similarly high mcr-1-positive carriage (79.2% and 76.2%), ranging from 45% to 100% in different provinces, while the mcr-1 rate in Jiangsu province reported by⁴⁸ was 71.9% which is very similar to our findings 71.6%. The present study and, together with the previous studies, confirmed a surprisingly high rate of mcr-1 in swine farms and is likely associated with the prolonged and extensive practice of COL as a growth promoter in pigs.

The coexistence of mcr-1 with other resistance genes in an E. coli was reported in China.⁵⁰ One recent study also assumed that a historic bridge existed between mcr-1 and ESBL.²³ However, there is scarcity in the incidence of the coexistence of mcr-1 and ESBL in the pig origin. Herein, we screened 275 MDR E. coli–resistant strains from pig farms in Jiangsu during 2016–2017 and found a high occurrence of mcr-1-positive strains with ESBL in the swine 39.5% (78/197), which was very high from the previous report.⁵¹ A very recent longitudinal study from China investigated the co-rising of mcr-1 and ESBL in chicken isolates.¹⁴ Among ESBL-positive strains, we found that bla_CTX-M is the most predominant.

In this study, 39.5% mcr-1-positive E. coli strains were detected which coexist in different ESBL genes. Among them, bla_CTX-M was the predominant one which was found in 55 mcr-1-positive E. coli 70.5% (Figure 1). On sequence-based analysis of these 55 bla_CTX-M isolates, interestingly the bla_CTX-M-15 gene was found to be the most prevalent bla_CTX-M gene (69%) followed by bla_CTX-M-55 (29%). The spread of the bla_CTX-M-15 gene is a common bla_CTX-M enzyme and detected widely in Enterobacteriaceae of human origin.^19,52 Very few studies have reported the co-occurrence of mcr-1 and bla_CTX-M-15 of human origin.^22,53 From China, one recent study in dairy cows also found bla_CTX-M-15 as the second prevalent ESBL gene 21.4% (62/275), but there was no mcr-1 gene detected in those bla_CTX-M-15 isolates.⁵⁴ Another study also reported the coexistence of mcr-1 and bla_CTX-M-15 in Turkey hen meat.⁵⁵ Herein, this is the first investigation that reveals the coexistence of mcr-1 and bla_CTX-M-15 in E. coli strains of pig origin in a very high proportion.

A significant increase in the bla_CTX-M-55 was found in E. coli by over a period of ten years.⁴ Many previous studies reported that bla_CTX-M-55 in human strains in China has become the second dominant bla_CTX-M type and even the occurrence of bla_CTX-M-55 was higher than bla_CTX-M-15.⁵¹ A high rate of mcr-1 and bla_CTX-M-55 was recently detected from the chicken origin in China,¹⁴ which was consistent with our results. Thus, the concurrent dissemination of the mcr-1 harboring bla_CTX-M-15 and bla_CTX-M-55 mediated by a single bacterial clone is existing which suggests that mcr-1 is found in the diverse reservoirs.

In addition, bla_CTX-M-15 and bla_CTX-M-55 were previously reported on conjugative plasmids, ie, FIB, IncI1, IncHI2, IncK, IncP and IncN.^35,56 Therefore, we also detected these incompatibility types by PCR typing. While mcr-1 gene was often found on conjugative plasmids like IncF, IncI2, IncHI2, IncN, IncP and X4, which exhibit an unexpected diversity.⁵⁷ In conjugation experiment, IncHI2, IncFIB, IncFIC, IncN and IncX4, were found in both donors and transconjugants and were associated with the transfer of the mcr-1 and ESBL encoding genes.

Genetic representation of mcr-1-carrying plasmids demonstrated that this gene is located on different conjugative elements of ~97 kb and 242 kb in size. The fact that mcr-1-carrying E. coli isolates display divergent PFGE profiles suggests that these elements may play a vital role in mcr-1 transmission. The incidence of closely related plasmids that carry mcr-1 and ESBL resistance genes among genotypically varied E. coli strains from various origins is a threat for alarm as it indicates that plasmids can easily disseminate from animals to humans and the spread of these plasmids may be remarkably challenging to control.

Considering that bla_CTX-M has become the most prevalent ESBL type of animal origin in the last few years, this situation may suggest that mcr-1 and bla_CTX-M emerged and arose due to the extensive use of antimicrobial practice in animal farming in the last decade. Our results also suggested that bla_CTX-M-15 and bla_CTX-M-55 and other β-lactamase genes coharboring with mcr-1 positive isolates is a potential threat to public health as the pig carrying these genes may enter the food chain. It is recommended that we should pay high consideration in monitoring the incidence of ESBL-producing & COL-resistant E. coli in both clinical and food-producing animals.

Conclusion

Our study reported a high incidence of the mcr-1-carrying ESBL-producing E. coli recovered from pigs in Jiangsu, China. The coexistence of the mcr-1- and bla_CTX-M-15-carrying isolates displaying MDR, recovered from pig origin is a major concern for both humans and veterinary medicine. The presence of these genes on the conjugative plasmids with the ability to transfer between similar strains which contain other drug resistance genes emphasizes on urgent intervention.

Acknowledgments

This work was supported by the National Key Research and Development Program of China (2018YFD0500300), the National Natural Science Foundation of China (31702292, 31872517), the Natural Science Foundation of Jiangsu Province (BK20170710), the China Postdoctoral Science Foundation (2017M611841, 2018T110515) and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

Disclosure

The authors report no conflicts of interest in this work.

Supplementary materials

Figure S1 Map of sampling sites in Jiangsu province

Figure S2 Screening of mcr-1 and ESBL encoding genes in E. coli.

Notes: PCR product was separated on 1% agarose gel. Lane 1 and 6 shows 2000 bp molecular marker (Vazyme, Beijing, China); Lane 2, (mcr-1), Lane 3 (blaCTX-M) isolate; Lane 4, (blaTEM); and Lane 5 shows (blaSHV).

Table S1 List of primers used for plasmid replicons typing in this study

Primers	Sequence (5ʹ to 3ʹ)	Target sites/genes	Annealing temperature	Amplicons size	References
PBRT* primers
HI1-F HI1-R	GGAGCGATGGATTACTTCAGTAC TGCCGTTTCACCTCGTGAGTA	parA-parB	58 °C	471-bp	^Citation1
HI2 -F HI2- R	GGCTCACTACCGTTGTCATCCT CGAAAGCCGGACGGCAGAA	RNAI	58 °C	644-bp	^Citation1
I1-F I1-F	CGAAAGCCGGACGGCAGAA TCGTCGTTCCGCCAAGTTCGT	iterons	58 °C	139-bp	^Citation1
X -F X -R	AACCTTAGAGGCTATTTAAGTTGCTGAT TGAGAGTCAATTTTTATCTCATGTTTTAGC	ori ꭚ	58 °C	376-bp	^Citation1
L/M-F L/M-F	GGATGAAAACTATCAGCATCTGAAG CTGCAGGGGCGATTCTTTAGG	repA, B, C	58 °C	785-bp	^Citation1
N-F N-R	GTCTAACGAGCTTACCGAAG GTTTCAACTCTGCCAAGTTC	repA	58 °C	559-bp	^Citation1
FIA- F FIA- R	CCATGCTGGTTCTAGAGAAGGTG GTATATCCTTACTGGCTTCCGCAG	iterons	58 °C	462-bp	^Citation1
FIB –F FIB -R	GGAGTTCTGACACACGATTTTCTG CTCCCGTCGCTTCAGGGCATT	repA	58 °C	702-bp	^Citation1
W-F W-R	CCTAAGAACAACAAAGCCCCCG GGTGCGCGGCATAGAACCGT	repA	58 °C	242-bp	^Citation1
Y-F Y-R	AATTCAAACAACACTGTGCAGCCTG GCGAGAATGGACGATTACAAAACTTT	repA	58 °C	765-bp	^Citation1
P-F P-R	CTATGGCCCTGCAAACGCGCCAGAAA TCACGCGCCAGGGCGCAGCC	iterons	58 °C	534-bp	^Citation1
FIC –F FIC -R	GTGAACTGGCAGATGAGGAAGG TTCTCCTCGTCGCCAAACTAGAT	repA2	58 °C	262-bp	^Citation1
A/C -F A/C -R	GAGAACCAAAGACAAAGACCTGGA ACGACAAACCTGAATTGCCTCCTT	repA	58 °C	465-bp	^Citation1
T-F T-R	TTGGCCTGTTTGTGCCTAAACCAT CGTTGATTACACTTAGCTTTGGAC	repA	58 °C	750-bp	^Citation1
FIIS –F FIIS -R	CTGTCGTAAGCTGATGGC CTCTGCCACAAACTTCAGC	repA	58 °C	270-bp	^Citation1
FrepB-F FrepB-R	TGATCGTTTAAGGAATTTTG GAAGATCAGTCACACCATCC	RNAI/repA	58 °C	270-bp	^Citation1
K/B -F K/B-R	GCGGTCCGGAAAGCCAGAAAAC TCTTTCACGAGCCCGCCAAA	RNAI	58 °C	160 bp	^Citation1
B/O-F B/O-R	GCGGTCCGGAAAGCCAGAAAAC TCTGCGTTCCGCCAAGTTCGA	RNAI	58 °C	159 bp	^Citation1
X4- F X4-R	AGCAAACAGGGAAAGGAGAAGACT TACCCCAAATCGTAACCTG	-	62 °C	569 bp	^Citation2
IncI2-F IncI2-R	ATTGTTGCGTGGCTTCAT TGGAGAGGATTAAGGAGGAA	-	60 °C	353 bp	This study

References

• Carattoli A, Bertini A, Villa L, Falbo V, Hopkins KL, Threlfall EJ. Identification of plasmids by PCR-based replicon typing. J Microbiol Methods. 2005;63(3):219–228.15935499
   PubMed Web of Science ®Google Scholar
• Johnson TJ, Bielak EM, Fortini D, et al. Expansion of the IncX plasmid family for improved identification and typing of novel plasmids in drug-resistant Enterobacteriaceae. Plasmid. 2012;68(1):43–50.22470007
   PubMed Web of Science ®Google Scholar