The High Prevalence of Plasmid-Mediated Quinolone Resistance Among Very Low Birth-Weight Infants in Poland

Abstract

The aim of this study was to investigate the prevalence of plasmid-mediated quinolone resistance (PMQR) determinants in Escherichia coli from neonatal intensive care units (NICUs) in Poland. The study was conducted on 80 E. coli isolates from different types of infections collected between 2009 and 2012. Six (5%) isolates were not susceptible to ciprofloxacin, 16% to ofloxacin, and 6.2% to levofloxacin. Among the 80 isolates, 27.5% carried at least one PMQR determinant (n=22). qnrB was found in 8.8% of isolates (7/80), and qnrS was found in 2.5% (2/80). No isolates carrying qnrA, qnrD, qnrC, qepA, or oqxAB were found. The prevalence of aac(6′)-Ib-cr was 27.5%, and 54.5% of the isolates contained aac(6′)-Ib-cr and bla_CTX-M. qnr-positive strains were more likely to produce extended-spectrum β-lactamases (ESBLs) than qnr-negative strains, which may suggest an association between qnr and ESBLs (88.9% vs. 17%, p=0.0003). PMQR-positive isolates had significantly higher ciprofloxacin MIC₅₀ values (28.8-fold higher when comparing the MIC₅₀ values) than the PMQR-negative strains (0.23 vs. 0.008 mg/L), regardless of the presence of quinolone resistance-determining region mutations. Five of the nine isolates, for which conjugation was performed successfully, harbored plasmids that contained CTX-M together with qnr and aac(6′)-Ib-cr. Our data suggest that the number of qnr genes detected in E. coli from newborns may be related to the selection of qnr through antimicrobial exposure. Even if fluoroquinolones are not commonly used in the NICU, E. coli isolates may carry PMQR. The high prevalence of PMQR is of serious concern, as it may be horizontally transferred to other pathogenic bacteria.

Introduction

Fluoroquinolones are currently one of the three most prescribed antimicrobial families in the world. As a consequence of their use, quinolone resistance rates have increased considerably in recent years.^13,16

Multidrug resistance in Escherichia coli is an increasing public health problem challenging the clinical use of antibiotics. The most frequent resistance determinant leading to multidrug resistance in E. coli is the production of extended-spectrum β-lactamases (ESBL).^39,40

One of the major concerns is the association between ESBL production and fluoroquinolone resistance (FQ-R) or aminoglycoside resistance.²⁴ The latter leads to the coselection of FQ-R by β-lactams or aminoglycosides, and vice versa, with β-lactam or aminoglycoside resistance caused by fluoroquinolones. The mobile elements (such as plasmids) carrying the resistance genes can easily spread to other microorganisms and individuals, which may increase the number of circulating antimicrobial-resistant organisms.²

Plasmid-mediated quinolone resistance (PMQR) was more recently described worldwide. Four major PMQR determinants were described: Qnr proteins, which protect topoisomerases against quinolone activity, Aac(6′)-Ib-cr, which acetylates ciprofloxacin and norfloxacin, and QepA and OqxAB, which are efflux pumps.⁴⁵ Although PMQR confers a low level of quinolone resistance, increasing PMQR prevalence has been described, and these determinants significantly reduce the activity of fluoroquinolones.^1,3,35 Consequently, detection of these genes among clinical bacterial isolates is worthwhile.

Most of the studies describing PQMR prevalence in Enterobacteriaceae were conducted on strains isolated from adults. Fluoroquinolones are associated with a wide range of side effects in children and neonates, and for this reason, they are not recommended for pediatric use. Ciprofloxacin is not recommended in the current guidelines for neonatal sepsis but is more commonly used to treat sepsis caused by multiple-resistant organisms sensitive to ciprofloxacin.^15,36 A systematic review of the use of ciprofloxacin in neonates was conducted by Kaguelidou et al., and no serious adverse events were observed (the evaluation was predominantly clinical, and follow-up was limited to a few months after the end of treatment).²⁷

However, to our knowledge, fluoroquinolones are used in Poland in pediatric patients together with other antibiotics when no other option is available. A survey conducted in four Polish neonatal intensive care units (NICUs) in 2007 on a group of newborns revealed that the most frequently used antibiotics were β-lactams (48.71–75.67% of all antibiotics used in those hospitals), but fluoroquinolones were used in two hospitals (∼2% of all antibiotics used).⁴⁴ In another survey that was conducted in 273 neonatal wards in Poland, the most intensively used antibiotics (as first resort antibiotics) were lactams (in ∼40% of wards) in combination with aminoglycosides and penicillins.²²

To our knowledge, little data describing PMQR-producing Enterobacteriaceae isolated from pediatric patients are available. Studies conducted in 2005–2006 in China on 292 isolates showed that 20 (6.8%) were positive for PMQR determinants, and 24 (8.2%) were positive for aac(6′)-Ib, of which 10 had the -cr variant.²⁰ The qepA gene was not detected among those isolates.²⁰ A high prevalence of qnrB (54.0%) and qnrS (14.0%) was discovered in pools of commensal enterobacterial isolates from 310 healthy children (age: 6–72 months) living in Peru and Bolivia using a metagenomic approach.³⁴ Given studies reporting PMQR in children, we decided to design this study because there were no data available for Poland.

Then, we focused on E. coli, as it is an important pathogen in newborns. According to EARS-Net (www.ecdc.europa.eu) in 2010 in Poland, 26% of E. coli clinical isolates that were isolated from invasive infections were found to be FQ-R. Because fluoroquinolones are mainly given to adults, there is lack of data about FQ-R mechanisms in E. coli isolated from Polish infants. The prevalence of PMQR determinants among E. coli in Poland has not been well studied. To the best of our knowledge, only one study has reported PMQR prevalence in Poland. This later study focused only on qnrA, qnrB, and qnrS genes in a 215 Enterobacteriaceae strain collection with reduced susceptibility to fluoroquinolones.³⁸ The isolates were collected in 2010 in a regular hospital in Warsaw, and 8.3% of the isolates carried qnr genes with qnrB as the most common qnr type. Most of the qnr-positive strains (88.9%) were ESBL producers with predominantly CTX-M and TEM types.³⁸

The aim of this study was to investigate the prevalence of qnr, aac(6′)-Ib-cr, qepA, and oqxAB genes in Enterobacteriaceae isolates collected from NICUs in Poland. We have shown that qnr genes are carried by mobile elements harboring multiple antimicrobial resistance genes. These genes may be coselected by the use of not only quinolone antimicrobials but the use of other antimicrobials also plays a critical role in the selection of qnr genes.

Materials and Methods

The surveillance conducted between January 1st 2009 and December 31st 2012 included 1,768 newborns in five Polish NICUs that participated in the activities of the Polish Neonatal Surveillance Network. The surveillance concerned the infants whose birth weight was below 1,500 g (very low birth weight) and who were hospitalized at the cooperating units from birth to discharge or until a weight of 1,800 g was reached or until death. The group comprised 14.6% of extremely low birth-weight infants with a birth weight up to 750 g, 27.3% of infants with a birth weight ranging from 751 to 999 g, and 58.2% of infants with a birth weight ranging from 1,000 to 1,500 g. In cases of infection, local microbiology laboratories carried out bacterial identification. Various diagnostic materials, including blood, tracheal/bronchial aspirates, swabs, and urine, were collected for culture and assessment of the microbial etiology of infections. Among hospitalized infants, 264 early-onset infections and 714 episodes of late-onset infections were diagnosed. Among all the infections, 96 were of E. coli etiology. Isolates from other infections were other gram-negative bacteria, gram-positive bacteria, atypical bacteria, and yeast.

Antibiotic consumption

Data about antibiotic treatment were entered into the database by the ward personnel based on the physicians' orders. Precise information about the type of drug, daily dose, and length of treatment for each antibiotic was collected and used for the calculation of two indicators: duration of treatment and defined daily dose (DDD) in reference to each case of infection. The aggregate sum of the number of days during which at least one dose of antibiotic was received for each antibiotic used (days of treatment) was expressed in days and the DDD, according to the ATC/DDD system of the World Health Organization (Anatomical Therapeutic Chemical, group “J01”). Only antibiotics for systemic use were taken into account.

Bacterial isolates

Altogether, 96 E. coli strains were isolated in five collaborating hospitals (Hospitals I–V) and 6 were not stored. All strains were isolated from cases of infection. Clinical samples came from the respiratory tract (n=50), blood (n=24), urine (n=8), and other areas (n=8). All isolates were identified by API ID 32E (bioMérieux, Marcy l'Etoile, France). Then, E. coli isolates were analyzed using the standardized pulsed-field gel electrophoresis (PFGE) protocol developed at the Centers for Disease Control and Prevention by the PulseNet program. GelCompar software (Applied Maths, Kortrijk, Belgium) was used to establish a DNA similarity matrix. A dendrogram was constructed using the unweighted pair group method of the arithmetic average clustering method with the Dice coefficient. PFGE results were interpreted according to international recommendations.⁴⁶ Isolates that were 90% similar or more were grouped together.

E. coli isolates showed very different pulsotypes, dominant epidemic clones were not detected, and all strains from separate patients were different. Cluster analysis based on PFGE of the 90 isolates showed 71 unique types, some of which were less than 70% similar, suggesting a genotypically variable population.⁷ Isolates with identical pulsotypes derived from the same patient were excluded from further analysis. We were unable to grow two strains. After all of the above exclusions were made, 80 nonredundant E. coli isolates were included for further analysis.⁷

Minimal inhibitory concentrations (MICs) of nalidixic acid (NA), ciprofloxacin (CIP), levofloxacin (LVX), and ofloxacin (OFX) were determined using the E-test method (bioMérieux). The results were interpreted using the clinical breakpoints as defined by the current guidelines of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) (www.eucast.org/clinical_breakpoints/). Antimicrobial susceptibility testing of other antibiotics, as well as ESBL detection, was performed previously.⁷

PCR-based plasmid replicon typing, the determination of the sequence-type (ST) 131 clone, and phylogenetic classification were conducted previously.⁷

Detection of PMQR determinants

The qnr, aac(6′)-Ib-cr, qepA, and oqxAB genes were systematically detected using real-time PCR and pyrosequencing as described elsewhere.^18,19

Briefly, after DNA extraction, qnr gene detection by multiplex real-time PCR was carried out using a LightCycler^® 480 (LC480) with HRM Supermix (Roche Molecular Diagnostics, Mannheim, Germany). The melt-curve analysis showed characteristic curves for each qnr gene family. Alleles of the qnr genes detected were characterized using high-resolution melting analysis or by conventional sequencing.¹⁷

qepA gene detection by simplex real-time PCR was performed using a LC480 with LC480 SYBR Green I Master (Roche Molecular Diagnostics). oqxAB gene detection was performed using conventional PCR as described elsewhere.³⁰

aac(6′)-Ib-cr was detected by pyrosequencing one amplification product that targeted the 77 bp portion with T304C/A and another product that targeted the 102 bp portion with G535T. T304C/A and G535C were easily detected on pyrograms. Appropriate positive controls were included.

Plasmid analysis

The transfer of PMQR genes was studied using the conjugation assay with a previously described method.¹⁴ The azide-resistant and rifampicin-resistant E. coli J.53 strain was used as the recipient strain. Transconjugants were selected on Tryptic soy (TS) agar plates with sodium azide (100 mg/L), rifampicin (250 mg/L), and cefotaxime (1 mg/L) or ciprofloxacin (0.06 mg/L). Phenotypic resistance was analyzed on TS agar plates with cefotaxime (1 mg/L) or ciprofloxacin (0.06 mg/L). ESBLs were identified using the double-disk synergy test prepared on Mueller-Hinton agar plates with rifampicin (250 mg/L). Plasmids (both from isolates and transconjugants) were classified into incompatibility groups using the PCR-based replicon typing scheme described by Carattoli et al.⁵

The positive controls used in the replicon typing were kindly provided by Prof. Alessandra Carattoli (Instituto Superiore di Sanita, Rome, Italy).

Statistics

Analyses were performed using Statsoft Statistica software. The normality of the distribution of continuous variables was tested using the Shapiro–Wilk test. For dichotomous variables, a chi-square test was used for expected frequencies >10, a chi-square test with Yates' correction was used for expected frequencies between 5 and 10, and a chi-square test with confirmation by Fisher's exact test was used for expected frequencies of 5 or lower. Values of p<0.05 were considered statistically significant.

Results

E. coli was isolated from 12.2% of all infections with a 5.4% incidence rate. In our wards, the consumption of β-lactams in a group of 44 newborns with E. coli infections was 83.91 DDD and fluoroquinolones was 2.8 DDD (1 newborn). Of the 25% ESBL-producing E. coli isolates (20/80), 11 harbored a bla_CTX-M-15 gene, 8 bla_CTX-M-3, and 1 isolate carried bla_SHV-5. Ten isolates coharbored bla_TEM-1 together with bla_CTX-M. OXA-1 was found in 13 isolates (16.3%). Among all the isolates, the majority belonged to the B2 group (52/80, 65.0%), and 27 of the isolates were ST131 clones (27/52, 52.0%). Twelve different plasmid Inc types were detected among the isolates. The most prevalent were IncFIB (57/80, 71.3%), IncF (47/80, 58.8%), and IncFIA (23/80, 28.8%).⁷

Among ESBL-positive isolates, 60.0% coharbored PMQR (12/20). Most of these PMQR isolates carried the bla_CTX-M-15 gene (11/12, 91.6%).

The highest level of resistance was observed against trimethoprim/sulfamethoxazole (SXT) (26/80, 32%) and cefuroxime (CXM) (20/80, 25%). According to the E-test results, 7.5% (6/80) of E. coli isolates were not susceptible to CIP, 16% (13/80) to OFX, and 6.2% to LVX (5/80). The coresistance among fluoroquinolones occurred as follows: four isolates were resistant to CIP, OFX, and LVX simultaneously. The MIC₅₀ for NA was 4.0 mg/L, for OFX was 0.094 mg/L, for CIP was 0.008 mg/L, and for LVX was 0.032 mg/L. Among the ESBL-positive strains, 20% (4/20) and 30% (6/20) were CIP and gentamicin (CN) resistant, respectively.

Among the 80 isolates tested, 27.5% carried at least one PMQR determinant (22/80) (Table 1). The prevalence of qnr determinants was 11.3% (9/80): qnrB was found in 8.8% isolates (7/80) and qnrS in 2.5% (2/80). No isolates carrying qnrA, qnrD, or qnrC were found. No isolates were positive for the qepA or oqxAB genes. Screening by PCR for aac(6′)-Ib-cr genes revealed 27.5% positive isolates (22/80).

Table 1.

Prevalence of Plasmid-Mediated Quinolone Resistance Determinants in Escherichia coli Isolates Studied

Strain	aac(6′)-Ib-cr	aac(6′)-Ib-cr+qnrB	aac(6′)-Ib-cr+qnrS	Total
ESBL-Escherichia coli	4	6	2	12 (54.5%)
Non-ESBL-E. Coli	9	1	0	10 (45.5%)
Total	13	7	2	22 (100%)

ESBL, extended-spectrum β-lactamase.

Of the isolates belonging to the B2 group, 21.1% had the ESBL phenotype (11/52). Among isolates of the B2 group, 25.0% (13/52) had at least one PMQR determinant: two isolates had qnrS and aac(6′)-Ib-cr, three had qnrB and aac(6′)-Ib-cr, and eight had aac(6′)-Ib-cr.

Of the 20 ESBL-positive isolates, 20.0% (4/20) harbored aac(6′)-Ib-cr, 30.0% (6/20) coharbored aac(6′)-Ib-cr and qnrB genes, and 10.0% (2/20) coharbored aac(6′)-Ib-cr and qnrS genes. Among the 22 aac(6′)-Ib-cr-positive isolates, 54.5% (12/22) were positive for both aac(6′)-Ib-cr and bla_CTX-M (CTX-M-15 or −3). Among non-ESBLs, aac(6′)-Ib-cr occurred in 15% (9/60) of isolates; aac(6′)-Ib-cr together with qnrB was found in 1 isolate (Table 1).

Among the 9 qnr-positive isolates, 88.9% (8/9) produced ESBLs, whereas among the 71 qnr-negative isolates, 17.0% (12/71) produced ESBLs. qnr-positive strains were more likely to produce ESBLs, which may suggest an association between qnr and ESBLs (p=0.0003).

All PMQR-positive isolates had significantly higher ciprofloxacin MIC₅₀ values (28.8-fold higher when comparing the MIC₅₀ values) than the PMQR-negative strains (0.23 vs. 0.008 mg/L), regardless of the presence of quinolone resistance-determining region (QRDR) mutations. A similar situation was observed for all tested fluoroquinolones: the MIC₅₀s for ofloxacin and levofloxacin in PMQR-positive strains were 0.75 and 0.38 mg/L and in PMQR-negative strains were 0.094 and 0.032 mg/L, respectively. The rate of nonsusceptibility to ciprofloxacin was significantly higher in the PMQR-positive isolates (6/22, 27.3%) than in the PMQR-negative isolates (0, 0% [p<0.001]). A similar situation was observed for ofloxacin (p<0.0001).

Conjugation experiments involving the qnr- and/or aac(6′)-Ib-cr-positive isolates revealed that these genes were successfully transferred from 9 of 22 isolates (Table 2). Five of the nine isolates, for which conjugation was performed successfully, harbored plasmids that contained CTX-M together with qnr and aac(6′)-Ib-cr.

Table 2.

Characteristic of Isolates with qnr and aac Determinants and Their Transconjugants

Isolate no.	Qnr gene	aac type	NA [mg/L]	OFX [mg/L]	LVX [mg/L]	CIP [mg/L]	PFGE pattern	Plasmid-mediated β-lactamases	OXA-1-like	Replicon type
1	—	aac-(6′)-Ib-cr	3	0.064	0.032	0.004	u	TEM-1	—	—
Tc1	—	—	2	0.032	0.032	0.004		—	—	—
7	qnrB	aac-(6′)-Ib-cr	2	0.064	0.032	0.38	u	CTX-M-15, TEM-1	OXA-1-like	FIB
Tc 7 (1)	—	—	2	0.016	0.023	0.006		—	OXA-1-like	FIB
Tc 7 (2)	qnrB	aac-(6′)-Ib-cr	2	0.023	0.008	0.016		CTX-M-15, TEM-1	—	—
14	qnrB	aac-(6′)-Ib-cr	4	0.75	0.25	0.5	u	CTX-M-15+TEM-1	OXA-1-like	FIB, FIIS
Tc 14	qnrB	aac-(6′)-Ib-cr	4	0.75	0.25	0.25		TEM-1	OXA-1-like	FIB
114	qnrB	aac-(6′)-Ib-cr	8	1	0.38	0.75	u	CTX-M-3+TEM-1	OXA-1-like	HI2
Tc114	qnrB	aac-(6′)-Ib-cr	4	0.19	0.19	0.25		CTX-M-3+TEM-1	OXA-1-like	HI2
160	qnrS	aac-(6′)-Ib-cr	256	32	12	>32	D	CTX-M-15	OXA-1-like	FIA, F
Tc160	qnrS	—	6	0.38	0.25	0.38		CTX-M-15	OXA-1-like	FIA
166	qnrS	aac-(6′)-Ib-cr	8	0.75	0.38	0.25	D	CTX-M-15, TEM-1	OXA-1-like	FIA, F
Tc166	qnrS	aac-(6′)-Ib-cr	6	0.023	0.047	0.047		—	OXA-1-like	F
189	—	aac-(6′)-Ib-cr	12	1	0.38	0.38	u	TEM-1	—	FIA, F
Tc189	—	aac-(6′)-Ib-cr	12	0.5	0.38	0.19		—	—	FIA
204	—	aac-(6′)-Ib-cr	3	0.094	0.016	0.064	u	TEM-1	—	F, B/O
Tc204	—	aac-(6′)-Ib-cr	3	0.094	0.016	0.064		—	—	F
205	qnrB	aac-(6′)-Ib-cr	24	0.5	0.75	0.38	u	CTX-M-15+TEM-1	OXA-1-like	FIB, F
Tc 205	qnrB	aac-(6′)-Ib-cr	24	0.19	0.75	0.38		CTX-M-15+TEM-1	OXA-1-like	FIB, F

NA, nalidixic acid; OFX, ofloxacin; LVX, levofloxacin; CIP, ciprofloxacin; PFGE, pulsed-field gel electrophoresis; Tc, transconjugant; u, unique; D, clone D.

Discussion

PMQR genes lead to low-level resistance to fluoroquinolones.³³ As the prevalence of PMQR increases^3,35 and PMQR determinants reduce the activity of fluoroquinolones,¹ detection of these genes among clinical bacterial isolates is worthwhile.

E. coli remains one of the leading causes of infections in neonates. ESBLs associated with resistance to fluoroquinolones and aminoglycosides have emerged. CTX-M is currently the most reported type of ESBL in Poland, with the main types being CTX-M-15, due to the global spread of the B2-ST131 clone, and CTX-M-3.^10,32 The majority of ESBLs reported in this study were CTX-M. As the E. coli population has a clonal organization, it has been shown that extraintestinal pathogenic E. coli belongs mainly to phylogenetic group B2 and to a lesser extent to group D, but intestinal commensal isolates belong to groups A and B1.^23,37,39 Recently, the multilocus sequence typing method has shown that ST131 is prevalent in various parts of the world and includes highly virulent isolates of E. coli. ST131 isolates very often belong to group B2 and are frequently resistant to fluoroquinolones and carry the CTX-M-15 gene.³⁷

Approximately 65% of the tested strains belonged to phylogenetic group B2 (52/80) and 27 of these strains were ST131 clones (27/52, 52.0%). In this study, we detected that a huge proportion of the isolates belonging to the B2-ST131 clone contained ESBL (37.0%, 10/27) or PMQR (33.3%, 9/27). Such multiresistant strains may be a major public health concern. The dissemination of the E. coli B2-ST131 clone associated with CTX-M-15 β-lactamase is global.⁸

Previous surveys conducted in Europe described less PMQR in E. coli isolates than we found in our study (14.0% in Canada and 12.7% in Spain vs. 27.5% in our study), but these studies were conducted on strains isolated from adults.^3,39 The prevalence of the qnr genes among all strains (11.3%, 9/80) was quite high and was not consistent with other studies conducted in France, Norway, Sweden, and Canada, where the prevalence was ∼1–1.6%.^28,39,41 A high prevalence has also been detected in China (5.3%) and the United States (5%).^26,42 As previously described in Tunisia and China, qnrB was found to be the most common qnr gene in all our isolates (7/80, 8.8%).^9,11 However, we did not detect the qnrA gene, which was the most prevalent qnr gene reported in a Spanish survey.³¹ Similar proportions of qnr genes were also detected in other studies in France and China.^26,41,42 This comparison should be made cautiously, as the selection of the isolates was quite different among all these studies, and the isolates came mainly from adults.

A few studies of PMQR have been conducted on isolates from children. Among 292 ESBL-E. coli clinical isolates collected from five children's hospitals in China, 20 (6.8%) were positive for PMQR determinants and 24 (8.2%) were positive for aac(6′)-Ib (of which 10 had the -cr variant). Similar to our studies, qepA was not detected in the Chinese isolates.²⁰ An analysis performed on commensal enterobacterial pools from 310 healthy children (aged 6–72 months) living in four urban areas of Latin America showed that 54% of isolates carried qnrB and 14% carried qnrS, but qnrA was not detected.³⁴

To our knowledge, no data are available on OqxAB prevalence in neonates, and unfortunately, we detected no OqxAB in this study.

Among the 22 aac(6′)-Ib-cr-positive isolates, 12 (54.5%) were positive for both aac(6′)-Ib-cr and bla_CTX-M (CTX-M-15 or -3). Several studies have reported that aac(6′)-Ib-cr is linked to bla_CTX-M in Enterobacteriaceae.^12,29 As reported in other studies, our data note that the majority of isolates carrying PMQR determinants are able to produce ESBLs as well.^9,11,41 This situation makes treatment more expensive and limits the available treatment options.

Interestingly in our study, several qnr-positive isolates were detected in nalidixic acid-susceptible isolates (MIC <16 mg/L), similar to other studies.⁴ These findings could be explained by the fact that fluoroquinolones are not used for pediatric infections to avoid FQ-R enterobacterial isolate selection with multiple chromosomal mutations in QRDR.

Our data suggest that the number of qnr genes detected in E. coli from newborns may be related to the selection of qnr through antimicrobial exposure. The coexistence of qnrB with other resistance genes, such as bla_CTX-M-14 or bla_CTX-M-15, on the same plasmid is a well-known phenomenon; PMQR plasmids frequently contain other resistance genes, particularly determinants encoding resistance to β-lactams.^29,45

A number of newborns enrolled in this study were treated with β-lactams (consumption of β-lactams was 83.91 DDD), which substantiates the coselection hypothesis. A survey conducted in four Polish hospitals in 2007 revealed that β-lactams were the most widely used antibiotic family in Polish NICUs, and fluoroquinolones were used in two hospitals (∼2% of all antibiotics used).⁴⁴ These data highlight that the high prevalence of PMQR in very low birth-weight infants may indicate that there is a coselection of fluoroquinolones by β-lactams. PMQR can help microorganisms to develop resistance mutations when exposed to therapeutic concentrations of quinolone.⁴³ We have shown that isolates with qnr genes had significantly higher MIC values for ciprofloxacin, which was also demonstrated by other authors among pediatric patients.²⁹

In addition, isolates that were positive for multiple PMQR genes were found in this study. Seven isolates contained qnrB and aac(6′)-Ib-cr, and one isolate contained aac(6′)-Ib-cr and qnrS. Success in transferring the qnrS and qnrB genes by conjugation has frequently been reported, but both of these genes have been detected in nonconjugative plasmids.^6,21,31 The cotransfer of aac(6′)-Ib-cr together with qnr was detected among 9 of the 22 isolates. In many situations, qnr and CTX-M were located on the same plasmid. In one case (isolate 14), CTX-M-15 was not transferred simultaneously with qnrB, which was unexpected because qnrB and CTX-M genes were regularly reported to be on the same plasmids.^6,42 In contrast, the qnrS gene was in one case identified on a different plasmid than the CTX-M gene, which was previously reported.²⁵ The cotransfer of aac(6′)-Ib-cr may suggest that this gene can be located on the same plasmid with qnr and the β-lactamase gene. Unfortunately, in some cases, we were not able to perform conjugation, and this failure may suggest that the genes involved are located on nontransferrable plasmids. This issue needs further investigation.

In conclusion, we report that even if fluoroquinolones are not commonly used in the NICU, E. coli isolates may carry PMQR. Our data showed that qnr-positive strains do not have a common antimicrobial phenotype and their prevalence may be underestimated when only MICs are used as markers. MIC values do not give enough information about resistance mechanisms. Molecular screening seems to be mandatory. The prevalence of PMQR is of serious concern, as it may be horizontally transferred to other pathogenic bacteria and may lead to therapeutic failure as a consequence of high levels of antimicrobial resistance.

Our study has some limitations because the group of isolates was not abundant, but our results suggest the wide expression of PMQR genes in E. coli isolates (mainly ESBL positive) and an association between aac(6′)-Ib-cr and bla_CTX-M.

Footnotes

Acknowledgments

The authors wish to thank the staff of the NICUs for their help and interest in the study. Partial results from this publication were presented at a conference in the form of a poster (Summer Conference, Society for Applied Microbiology, Edinburgh, 2012) and at the Symbiosis student conference in May 2014 in Warsaw, Poland. This work was supported by a grant from the Ministry of Science and Higher Education (DEC-2011/01/D/N27/00104) and a partial grant from the University of Reims Champagne-Ardenne. The sponsors provided the funding for the project only.

Disclosure Statement

No competing financial interests exist.

References

Allou

, Cambau

, Massias

, Chau

, and Fantin

2009. Impact of low-level resistance to fluoroquinolones due to qnrA1 and qnrS1 genes or a gyrA mutation on ciprofloxacin bactericidal activity in a murine model of Escherichia coli urinary tract infection. Antimicrob. Agents Chemother., 53:4292–4297.

Bailey

J.K.

, Pinyon

J.L.

, Anantham

, and Hall

R.M.

2010. Commensal Escherichia coli of healthy humans: a reservoir for antibiotic-resistance determinants. J. Med. Microbiol., 59:1331–1339.

Briales

, Rodriguez-Martinez

J.M.

, Velasco

, de Alba

P.D.

, Rodriguez-Bano

, Martinez-Martinez

, and Pascual

2012. Prevalence of plasmid-mediated quinolone resistance determinants qnr and aac(6')-Ib-cr in Escherichia coli and Klebsiella pneumoniae producing extended-spectrum beta-lactamases in Spain. Int. J. Antimicrob. Agents., 39:431–434.

Cano

M.E.

, Rodriguez-Martinez

J.M.

, Aguero

, Pascual

, Calvo

, Garcia-Lobo

J.M.

, Velasco

, Francia

M.V.

, and Martinez-Martinez

2009. Detection of plasmid-mediated quinolone resistance genes in clinical isolates of Enterobacter spp. in Spain. J. Clin. Microbiol., 47:2033–2039.

Carattoli

, Bertini

, Villa

, Falbo

, Hopkins

K.L.

, and Threlfall

E.J.

2005. Identification of plasmids by PCR-based replicon typing. J. Microbiol. Methods, 63:219–228.

Cattoir

, Poirel

, Rotimi

, Soussy

C.J.

, and Nordmann

2007. Multiplex PCR for detection of plasmid-mediated quinolone resistance qnr genes in ESBL-producing enterobacterial isolates. J. Antimicrob. Chemother., 60:394–397.

Chmielarczyk

, Pobiega

, Wojkowska-Mach

, Romaniszyn

, Adamski

, Heczko

P.B.

, and Bulanda

2013. Molecular epidemiology, plasmid analysis, virulence, and resistance of Escherichia coli isolated from neonatal intensive care units in Poland. Diagn. Microbiol. Infect. Dis., 76:542–545.

Coque

T.M.

, Novais

, Carattoli

, Poirel

, Pitout

, Peixe

, Baquero

, Canton

, and Nordmann

2008. Dissemination of clonally related Escherichia coli strains expressing extended-spectrum beta-lactamase CTX-M-15. Emerg. Infect. Dis., 14:195–200.

Dahmen

, Poirel

, Mansour

, Bouallegue

, and Nordmann

2010. Prevalence of plasmid-mediated quinolone resistance determinants in Enterobacteriaceae from Tunisia. Clin. Microbiol. Infect., 16:1019–1023.

10.

Empel

, Baraniak

, Literacka

, Mrowka

, Fiett

, Sadowy

, Hryniewicz

, Gniadkowski

, and Beta-PL Study Group. 2008. Molecular survey of beta-lactamases conferring resistance to newer beta-lactams in Enterobacteriaceae isolates from Polish hospitals. Antimicrob. Agents Chemother., 52:2449–2454.

11.

Fang

, Huang

, Shi

, Hedin

, Nord

C.E.

, and Ullberg

2009. Prevalence of qnr determinants among extended-spectrum beta-lactamase-positive Enterobacteriaceae clinical isolates in southern Stockholm, Sweden. Int. J. Antimicrob. Agents., 34:268–270.

12.

Fihman

, Lartigue

M.F.

, Jacquier

, Meunier

, Schnepf

, Raskine

, Riahi

, Sanson-le Pors

M.J.

, and Bercot

2008. Appearance of aac(6')-Ib-cr gene among extended-spectrum beta-lactamase-producing Enterobacteriaceae in a French hospital. J. Infect., 56:454–459.

13.

Garau

, Xercavins

, Rodriguez-Carballeira

, Gomez-Vera

J.R.

, Coll

, Vidal

, Llovet

, and Ruiz-Bremon

1999. Emergence and dissemination of quinolone-resistant Escherichia coli in the community. Antimicrob. Agents Chemother., 43:2736–2741.

14.

Garcia

, Navarro

, Miro

, Villa

, Mirelis

, Coll

, and Carattoli

2007. Acquisition and diffusion of bla CTX-M-9 gene by R478-IncHI2 derivative plasmids. FEMS Microbiol. Lett., 271:71–77.

15.

Gendrel

, Chalumeau

, Moulin

, and Raymond

2003. Fluoroquinolones in paediatrics: a risk for the patient or for the community?. Lancet Infect. Dis., 3:537–546.

16.

Goossens

2009. Antibiotic consumption and link to resistance. Clin. Microbiol. Infect. 15 Suppl 3:12–15.

17.

Guillard

, de Champs

, Moret

, Bertrand

, Scheftel

J.M.

, and Cambau

2012. High-resolution melting analysis for rapid characterization of qnr alleles in clinical isolates and detection of two novel alleles, qnrB25 and qnrB42. J. Antimicrob. Chemother., 67:2635–2639.

18.

Guillard

, Duval

, Moret

, Brasme

, Vernet-Garnier

, and de Champs

2010. Rapid detection of aac(6')-Ib-cr quinolone resistance gene by pyrosequencing. J. Clin. Microbiol., 48:286–289.

19.

Guillard

, Moret

, Brasme

, Carlier

, Vernet-Garnier

, Cambau

, and de Champs

2011. Rapid detection of qnr and qepA plasmid-mediated quinolone resistance genes using real-time PCR. Diagn. Microbiol. Infect. Dis., 70:253–259.

20.

Han

, Yang

, Wang

, Lu

, Xu

, Wang

, Liu

, Deng

, and Shen

2010. The prevalence of plasmid-mediated quinolone resistance determinants among clinical isolates of ESBL or AmpC-producing Escherichia coli from Chinese pediatric patients. Microbiol. Immunol., 54:123–128.

21.

Hata

, Suzuki

, Matsumoto

, Takahashi

, Sato

, Ibe

, and Sakae

2005. Cloning of a novel gene for quinolone resistance from a transferable plasmid in Shigella flexneri 2b. Antimicrob. Agents Chemother., 49:801–803.

22.

Helwich

, Zejda

J.E.

, Brozek

, and Swietlinski

2008. Analysis of infections in neonatal units in Poland. Med. Wieku Rozwoj., 12:826–836.

23.

Herzer

P.J.

, Inouye

, and Whittam

T.S.

1990. Phylogenetic distribution of branched RNA-linked multicopy single-stranded DNA among natural isolates of Escherichia coli. J. Bacteriol., 172:6175–6181.

24.

Hooper

D.C.

2000. Mechanisms of action and resistance of older and newer fluoroquinolones. Clin. Infect. Dis., 31 Suppl 2:S24–S28.

25.

Hopkins

K.L.

, Wootton

, Day

M.R.

, and Threlfall

E.J.

2007. Plasmid-mediated quinolone resistance determinant qnrS1 found in Salmonella enterica strains isolated in the UK. J. Antimicrob. Chemother., 59:1071–1075.

26.

Jiang

, Zhou

, Qian

, Wei

, Yu

, Hu

, and Li

2008. Plasmid-mediated quinolone resistance determinants qnr and aac(6')-Ib-cr in extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae in China. J. Antimicrob. Chemother., 61:1003–1006.

27.

Kaguelidou

, Turner

M.A.

, Choonara

, and Jacqz-Aigrain

2011. Ciprofloxacin use in neonates: a systematic review of the literature. Pediatr. Infect. Dis. J., 30:e29–e37.

28.

Karah

, Poirel

, Bengtsson

, Sundqvist

, Kahlmeter

, Nordmann

, Sundsfjord

, Samuelsen

, and Norwegian Study Group on PMQR. 2010. Plasmid-mediated quinolone resistance determinants qnr and aac(6')-Ib-cr in Escherichia coli and Klebsiella spp. from Norway and Sweden. Diagn. Microbiol. Infect. Dis., 66:425–431.

29.

Kim

E.S.

, Jeong

J.Y.

, Jun

J.B.

, Choi

S.H.

, Lee

S.O.

, Kim

M.N.

, Woo

J.H.

, and Kim

Y.S.

2009. Prevalence of aac(6')-Ib-cr encoding a ciprofloxacin-modifying enzyme among Enterobacteriaceae blood isolates in Korea. Antimicrob. Agents Chemother., 53:2643–2645.

30.

Kim

H.B.

, Wang

, Park

C.H.

, Kim

E.C.

, Jacoby

G.A.

, and Hooper

D.C.

2009. oqxAB encoding a multidrug efflux pump in human clinical isolates of Enterobacteriaceae. Antimicrob. Agents Chemother., 53:3582–3584.

31.

Lavilla

, Gonzalez-Lopez

J.J.

, Sabate

, Garcia-Fernandez

, Larrosa

M.N.

, Bartolome

R.M.

, Carattoli

, and Prats

2008. Prevalence of qnr genes among extended-spectrum beta-lactamase-producing enterobacterial isolates in Barcelona, Spain. J. Antimicrob. Chemother., 61:291–295.

32.

Livermore

D.M.

, Canton

, Gniadkowski

, Nordmann

, Rossolini

G.M.

, Arlet

, Ayala

, Coque

T.M.

, Kern-Zdanowicz

, Luzzaro

, Poirel

, and Woodford

2007. CTX-M: changing the face of ESBLs in Europe. J. Antimicrob. Chemother., 59:165–174.

33.

Martinez-Martinez

, Pascual

, and Jacoby

G.A.

1998. Quinolone resistance from a transferable plasmid. Lancet, 351:797–799.

34.

Pallecchi

, Riccobono

, Mantella

, Bartalesi

, Sennati

, Gamboa

, Gotuzzo

, Bartoloni

, and Rossolini

G.M.

2009. High prevalence of qnr genes in commensal enterobacteria from healthy children in Peru and Bolivia. Antimicrob. Agents Chemother., 53:2632–2635.

35.

Paltansing

, Kraakman

M.E.

, Ras

J.M.

, Wessels

, and Bernards

A.T.

2013. Characterization of fluoroquinolone and cephalosporin resistance mechanisms in Enterobacteriaceae isolated in a Dutch teaching hospital reveals the presence of an Escherichia coli ST131 clone with a specific mutation in parE. J. Antimicrob. Chemother., 68:40–45.

36.

Pandolfini

, Kaguelidou

, Sequi

, Jacqz-Aigrain

, Choonara

, Turner

M.A.

, Manzoni

, and Bonati

2013. Wide intra- and inter-country variability in drug use and dosage in very-low-birth-weight newborns with severe infections. Eur. J. Clin. Pharmacol., 69:1031–1036.

37.

Peirano

, and Pitout

J.D.

2014. Fluoroquinolone-resistant Escherichia coli sequence type 131 isolates causing bloodstream infections in a canadian region with a centralized laboratory system: rapid emergence of the H30-Rx sublineage. Antimicrob. Agents Chemother., 58:2699–2703.

38.

Piekarska

, Rzeczkowska

, Zacharczuk

, Chrost

, Januszkiewicz

, Bareja

, Olak

, and Gierczynski

2012. Prevalence of qnr genes in clinical Enterobacteriaceae non-susceptible to fluoroquinolone in Poland. Med. Dosw. Mikrobiol., 64:211–219.

39.

Pitout

J.D.

, Wei

, Church

D.L.

, and Gregson

D.B.

2008. Surveillance for plasmid-mediated quinolone resistance determinants in Enterobacteriaceae within the Calgary Health Region, Canada: the emergence of aac(6')-Ib-cr. J. Antimicrob. Chemother., 61:999–1002.

40.

Poirel

, Bonnin

R.A.

, and Nordmann

2012. Genetic support and diversity of acquired extended-spectrum beta-lactamases in Gram-negative rods. Infect. Genet. Evol., 12:883–893.

41.

Poirel

, Leviandier

, and Nordmann

2006. Prevalence and genetic analysis of plasmid-mediated quinolone resistance determinants QnrA and QnrS in Enterobacteriaceae isolates from a French university hospital. Antimicrob. Agents Chemother., 50:3992–3997.

42.

Robicsek

, Jacoby

G.A.

, and Hooper

D.C.

2006. The worldwide emergence of plasmid-mediated quinolone resistance. Lancet Infect. Dis., 6:629–640.

43.

Rodriguez-Martinez

J.M.

, Velasco

, Garcia

, Cano

M.E.

, Martinez-Martinez

, and Pascual

2007. Mutant prevention concentrations of fluoroquinolones for Enterobacteriaceae expressing the plasmid-carried quinolone resistance determinant qnrA1. Antimicrob. Agents Chemother., 51:2236–2239.

44.

Rozanska

, Wojkowska-Mach

, Borszewska-Kornacka

, Cmiel

, Gadzinowski

, Gulczynska

, Helwich

, Kordek

, Pawlik

, Szczapa

, and Heczko

P.B.

2012. Antibiotic consumption and its costs of purchase in Polish neonatology networks units. Przegl. Epidemiol., 66:513–519.

45.

Strahilevitz

, Jacoby

G.A.

, Hooper

D.C.

, and Robicsek

2009. Plasmid-mediated quinolone resistance: a multifaceted threat. Clin. Microbiol. Rev., 22:664–689.

46.

van Belkum

, Tassios

P.T.

, Dijkshoorn

, Haeggman

, Cookson

, Fry

N.K.

, Fussing

, Green

, Feil

, Gerner-Smidt

, Brisse

, Struelens

, and European Society of Clinical Microbiology and Infectious Diseases (ESCMID) Study Group on Epidemiological Markers (ESGEM). 2007. Guidelines for the validation and application of typing methods for use in bacterial epidemiology. Clin. Microbiol. Infect., 13 Suppl 3:1–46.