Coprevalence of Plasmid-Mediated Quinolone Resistance Determinants QepA, Qnr, and AAC(6′)-Ib-cr among 16S rRNA Methylase RmtB-Producing Escherichia coli Isolates from Pigs

Plasmid-mediated quinolone resistance determinants, including Qnr peptides and AAC(6′)-Ib-cr, are increasingly identified worldwide among various clinical isolates of Enterobacteriaceae (7, 9, 10). Very recently, a novel plasmid-mediated fluoroquinolone-resistant determinant, QepA (quinolone efflux pump), which showed a considerable similarity to the major facilitator superfamily-type efflux pumps, was first identified in an Escherichia coli clinical isolate from Japan (13) and later found also in an E. coli isolate in Belgium (6). Interestingly, both of the two qepA-harboring E. coli isolates also contained the rmtB gene encoding a 16S rRNA methyltransferase, an emerging new molecular mechanism responsible for high-level pan-aminoglycoside resistance among gram-negative pathogens (3, 4, 6, 13, 14).

Our previous study showed that rmtB was highly prevalent among E. coli isolates from pigs in China (1). The aim of this study was to investigate the prevalence of plasmid-mediated quinolone resistance determinants among rmtB-producing E. coli isolates from pigs in China and to identify the association of the qepA gene with rmtB.

One hundred fifty-one E. coli isolates were obtained from pig feces sampled at two pig farms. These isolates were collected from 2005 to 2006, and 48 of them were identified as producers of RmtB. (Some of these data were published previously [1].) Screening for qepA, qnrA, qnrB, qnrS, and aac(6′)-Ib-cr genes was carried out by PCR amplification among the 48 rmtB-positive isolates. For qepA, the following primers were used to produce a 218-bp amplicon: qepA-F (5′-GCAGGTCCAGCAGCGGGTAG-3′) and qepA-R (5′-CTTCCTGCCCGAGTATCGTG-3′). Positive results were confirmed by direct sequencing of PCR products. qnrA, qnrB, qnrS, and aac(6′)-Ib-cr genes were detected by PCR using specific primers (the used qnrB primers were able to detect almost all known qnrB alleles except qnrB8), as previously described (5, 8, 11), and were finally confirmed by sequencing of each PCR product.

Overall, qepA, qnrB, qnrS, and aac(6′)-Ib-cr were detected in 28 (58.3%), 1 (2.1%), 9 (18.8%), and 6 (12.5%) of 48 RmtB-producing E. coli isolates, respectively (Table 1). The qnrB genes were identified as qnrB6 alleles by sequencing. The qnrS genes were confirmed as qnrS1 (four isolates) and qnrS2 (five isolates) alleles by sequencing. Four isolates with uniform pulsed-field gel electrophoresis (PFGE) patterns harbored qepA, qnrS2, and aac(6′)-Ib-cr genes concurrently.

TABLE 1.

Characteristics of E. coli isolates and transconjugants harboring rmtB, as well as qnr, qepA, and/or aac(6′)-Ib-cr

Isolate(s)a	PFGE type	Resistance gene detected	MIC (μg/ml) of enrofloxacin	Fold increase in quinolone MIC for transconjugant vs recipientb
				NOR	ENR	CIP	NAL	LEV
GZ3	A1	qepA	64	16	4	8	8	2
GZ4	A2	qepA	32	4	4	8	2	2
GZ5, GZ6	B	qepA	16	16,8	8, 2	8, 16	1	2, 4
GZ8	C	qepA	64	8	16	4	4	1
GZ9	D	qepA	32	16	2	8	4	2
GZ11	E	qepA	128	4	4	4	2	2
GZ12, GZ13, GZ14	F	qepA	16, 8, 16	16, 8, 16	4, 2, 8	8, 16, 4	1, 4, 1	8, 2, 4
GZ15	G	qepA	8	4	8	4	1	2
GZ16	H	qepA	>128	8	2	8	2	4
CQ15	I1	qepA	2	8	16	16	1	1
CQ18, CQ2, CQ5	I2	qepA	2, 2, 4	8, 16, 16	1, 16, 2	1, 3, 2, 8	1	1, 4, 1
CQ4c	J1	qepA	0.03	2	1	1	2	1
CQ20	J1	qepA	0.03	4	4	2	1	2
CQ26	K	qepA	0.5	4	2	4	1	1
CQ10	L	qepA	0.25	32	2	8	1	2
CQ14	M	qepA	16	32	2	4	1	2
GZ7	N	qepA, qnrS1	32	16	1	2	16	1
GZ1	O	qnrS1	4	16	32	16	4	4
GZ2c	O	qnrS1	2	2	2	4	4	2
CQ22c	P	qnrS1	4	2	2	1	1	1
CQ13	K	qnrS2	0.5	4	8	4	4	4
CQ6, CQ7, CQ12, CQ16	Q	qepA, qnrS2, aac(6′)-Ib-cr	2	16	2, 2, 4, 2	16, 4, 8, 8	16, 1, 1, 2	4, 1, 2, 2
GZ10	R	qepA, aac(6′)-Ib-cr	16	16	4	16	16	4
CQ19	S	qepA, aac(6′)-Ib-cr	2	16	4	4	1	1
CQ1c	U	qnrB6	0.25	2	1	1	2	1

^aIsolates with the same letters were isolated from the same farm.

^bThe quinolone MICs of the recipient strains were 4 μg/ml for nalidixic acid (NAL); 0.015 μg/ml for ciprofloxacin (CIP); and 0.03 μg/ml for norfloxacin (NOR), enrofloxacin (ENR), and levofloxacin (LEV).

^cRmtB-positive transconjugants not containing any plasmid-mediated quinolone resistance determinants.

To investigate the association of rmtB and qepA, rmtB-positive E. coli transconjugants described previously (1) were subjected to PCR amplification of qepA, and all transconjugants that originated from the 28 qepA-positive isolates selected with aminoglycoside resistance were positive for the qepA gene except one, suggesting an strong linkage of qepA with rmtB. Two rmtB-positive transconjugants also harbored qnrS1 or qnrS2.

MICs of ciprofloxacin, enrofloxacin, levofloxacin, nalidixic acid, and norfloxacin for the 27 qepA-positive and 2 qnrS-positive transconjugants were determined by the agar dilution method according to CLSI guidelines (2). The increase (fold) in quinolone MICs for transconjugants compared with those of recipients is shown in Table 1. The MICs for transconjugants strongly indicated that qepA as well as qnrS conferred quinolone resistance, with a 4- to 32-fold increase in norfloxacin MICs and 1- to 32-fold increase in enrofloxacin and ciprofloxacin MICs. However, variations in the quinolone MICs for different transconjugants suggested that the QepA may be expressed at variable levels. Xu et al. (12) recently reported that different promoter strengths may cause the differences in qnrA expression levels and in ciprofloxacin MICs of different transconjugants. Further studies are needed to find out whether the wide range of MICs of quinolones for different qepA-harboring transconjugants depends on the diversities in qepA expression levels due to different promoter strengths. MICs of enrofloxacin for all isolates were also determined by the agar dilution method according to CLSI guidelines. As indicated in Table 1, most isolates were resistant to enrofloxacin (MIC, ≥2 μg/ml), but six isolates were susceptible to enrofloxacin.

This study shows the high prevalence of plasmid-mediated quinolone resistance determinants among E. coli isolates recovered from food-producing animals. A total of 58.3% (28/48) of rmtB-positive E. coli isolates harbored qepA gene, indicating a close relationship between qepA and rmtB, which has been reported in the previous studies (6, 13). This is also the first time three different plasmid-mediated quinolone resistance determinants (QepA, Qnr, and AAC(6′)-Ib-cr) were identified in an E. coli strain. Coproduction of QepA, Qnr, AAC(6′)-Ib-cr, and RmtB may well facilitate the survival of bacteria under selective pressure of antimicrobial agents in both veterinary and human clinical environments, and the resistance determinants in food-producing animals could be transmitted to humans via the food chain. Further spread of these resistance determinants among pathogenic microbes may occur in the near future. Thus, it is necessary to monitor and minimize the spread of such resistance determinants among hazardous bacteria in both humans and animals.