PMQR determinants among clinical isolates of ESBL and Amp C producing Serratia marcescens in Mansoura University Hospitals: A 6-year study

Background: Plasmid mediated quinolone resistant determinants (PMQRs) are found to be of clinical relevance because it causes decreased susceptibility to quinolones which are valuable antibiotics against intrinsically resistant Serratia marcescens. This study aimed to determine the occurrence of PMQR determinants (qnr, aac(6′)-Ibcr and qep) and their relation with extended-spectrum β-lactamase (ESBL) and / or Amp Cproducing S. marcescens. Material and Method: One hundred and eleven S. marcescens were tested for resistance to five quinolones by disc diffusion method. Testing for qnr, aac(6′)-Ib-cr and qep was carried out by multiplex PCR, while ESBL and Amp C production were evaluated by double disc synergy test and cefoxitin Hodge test, respectively. Results: A total of 51/111 (45.9%) S. marcescens were resistant to one or more of the tested quinolones. Among S. marcescens isolates, 13 (11.7%) were ESBL producers and 5 (4.5%) were Amp C producers. Multiplex PCR confirmed the presence of PMQR determinant in 14 (12.6%) of S. marcescens isolates and these were positive for qnrA and qnr B in 3 and 10 of the isolates, respectively, whereas one isolate coharbored aac(6’)-Ib-cr and qnrA. Among 13 ESBL producers, 3 and 5 isolates were positive for qnrA and qnr B, respectively (8/13; 61.5%),whereas one isolate which has co-expressed aac(6’)-Ib-cr and qnrA was associated with Amp C production (1/5;20%). THE INTERNATIONAL ARABIC JOURNAL OF ANTIMICROBIAL AGENTS


Introduction
Serratia marcescens was considered in the past to be a harmless and non-pathogenic organism in humans, it is now becoming an important cause of hospital-associated infections including urinary tract infections, respiratory tract infections, septicemia, wound infections and meningitis [1].
Quinolones are powerful antibacterial agents against a wide range of bacteria species causing clinical infections [2]. They are particularly valuable against infections caused by Gram-negative bacteria species such as S. marcescens which is intractable to other antibiotics because of its intrinsic resistance [1].
Plasmid-mediates quinolone resistance (PMQR) involves three families, qnrA, qnrB, and qnrS [3]. The presence of these enzymes is of great importance because they are not merely able to confer resistance against most available quinolones but as well are often related to extended-spectrum β-lactamases (ESBLs) and/or plasmid-mediated Amp C β-lactamases (pAmp Cs) [4]. Six determinants of QnrA has been recorded worldwide from different enterobacterial species. This gene codes for a 218 amino acid protein of the pentapeptide family that prevent binding of quinolone to gyrase and topoisomerase IV protecting DNA [5]. Two other PMQR genes, called, qnrB and qnrS, have been recognized that code for QnrB and QnrS and share 41% and 60% amino acid homology with QnrA, respectively [6].
Another mechanism of conjugal quinolone resistance via an enzymatic inactivation of some quinolones by means of aminoglycoside acetyltransferase that codes for cr variant of aac(6′)-Ib was detected. It acts by N-acetylation of their piperazinyl amine causing decreased susceptibility to quinolone. They have the same mechanism of action like QnrA [7].
Lastly, the extra PMQR determinants is the quinolone efflux pump gene, qepA. It shared a substantial resemblance to the major facilitator superfamily-type efflux pumps. This protein cause reduced susceptibility to many quinolones [8].
This study was conducted to find out the rate of PMQR determinants (qnr, aac(6′)-Ib-cr and qep) in S. marcescens isolated from hospitalized patients at the Mansoura University Hospitals and their association with extended-spectrum β-lactamase (ESBL) and / or Amp C-producing S. marcescens.

Study design
This prospective study was carried out over a period of 6 years from January 2008 through January 2014, where S. marcescens strains were isolated from various clinical samples obtained from patients at the Mansoura University Hospitals (MUHs), Mansoura, Egypt. The samples processing were carried out in diagnostics and infection control unit (MDICU).
All samples were cultured on Blood and MacConkey's agar. Identification of S. marcescens isolates to species level was done primarily by Gram stain, colonial morphology, biochemical reactions including indole, methyl red, voges proskauer and citrate test (IMVC), motility on semisolid agar, DNAase, gelatin liquefaction tests and confirmed later with API 20E (bioMérieux) [9].
ESBL detection was detected by the use of a double-disk synergy test, and a test was considered positive if any distortion or zone increase towards the disc of amoxicillin-clavulanate was detected as reported by Paterson and Bonomo [11].
The cefoxitin Hodge test was used to screen strains producing Amp C β-lactamase and the test was interpreted positive in case of the presence of clear growth of the control strain (ATCC E. coli 25922) along with the test isolate within the 22 mm inhibition zone after overnight Incubation at 37°C [12].

Multiplex PCR technique
Quick DNA extraction was done by a boiling technique that consists of a heating at 100 o C of one colony in a volume of 100 µL of distilled water for 5 minutes then centrifugation of the cell suspension.
Multiplex PCR amplification of qnrA, qnrB, and qnrS was carried out according to Robicseket al 2006. Primer sets used are listed in table (1) [13] . All primers were added to the template DNA and PCR master mix (Fermentas). The cycling conditions used were 94°C for 45 s (denaturation), 53°C for 45 s (annealing), and 72°C for 60 s (extension), which was repeated for 32 cycles. Another multiplex of PCR was held for amplification of aac(6′)-Ib and qep with sequence of primers [14] mentioned as presented in Table  ( Scientific™) to detect aac(6′)-Ib-cr, that is deficient in the BtsCI restriction site found in the wild-type gene. The wild-type aac(6′)-Ib PCR gave 210bp and 272-bp fragments after restriction [15]. Amplification products were identified by their sizes in 2% ethidium bromide-stained agarose gels using 50bp DNA as control marker.

Statistical Analysis
Chi-square analysis was done to verify correlation between PMQR determinants and ESBl and Amp C producers. All statistical analyses were performed using Statistical Package for Social Science program (SPSS version 15.0 for windows, Chicago, IL). P≤ 0.05 was considered statistically significant.

Discussion
This study has detected quinolone resistance at a rate of 45.9%, whereas a study by Park et al. in USA [16], found a higher rate of resistance to quinolones (66.3%), and the study of Yang et al. in China [17], reported much lower results (10.3%). The overall frequency of PMQR genes among the S. marcescens isolates in our study was 12.6% which is similar to other recent study that carried in China [3], while the study of Park et al. [16], on the contrary has showed much lower frequency (2.4%).
The present study demonstrates that the quinolone resistance was increased over time together with a growing shift to PMQR molecular determinants. This increase may be attributed to performance and enhancement of selection pressure of resistant mutants especially with extensive use of quinolones over the study period. In addition, further increase in the magnitude of the problem could be supported by the hypothesis that PMQR encourage emergence of QRDR mutations in gyrA and parC associated with quinolone resistance [14,17].
This study shows that qnrB is the most dominant (71.4%) among PMQR genes, ant it is similar to many other studies over the world [6,7,[18][19][20][21]. Moreover, this gene was the most common among ESBL producers signifying the relation between this gene and other resistance genes as observed by other researchers [22][23][24][25]. In contrast to the current results, the data reported by Yang et al. [3], found that the aac(6′)-Ib-cr gene was the most prevalent among S. marcescens ESBL producers. Additionally, Robicsek et al. [7], found this gene to be the most common in E. coli and the reason of this finding could not be explained. However, a different geographic distribution could be suggested as a cause for this variation. It is noted that aac(6')-Ib-cr and qnrA were co-expressed in one of our isolates, This finding support other reported results where some plasmids can carry both aac(6′)-Ib-cr and qnrA [14]. This association confers the superior level of resistance to quinolones than do each gene by itself, although other studies found no relation between qnr genes and aac(6′)-Ib-cr gene [7,26].
The frequency of the PMQR determinants in ESBL producing S. marcescens isolated from hospitals in this study was 61.5%., while a study done in Korea has detected only 4.8% as qnr-positive among ESBL producing S. marcescens strains [16]. However, high incidence (71.4%) has also been found in China [3]. A low incidence of qnr genes was detected among ESBL producing isolates of E. coli and Klebsiella spp. in France and Canada(1.6% and 1%, respectively), whereas high frequency of the qnr genes was reported in Egypt (26.6%) among ESBL producing E.coli [26], and in Spain, this prevalence was found among ESBL producing E.coli and Klebsiella at 5% and 8%, respectively [27].
Although our study has not detected any statistically significant results regarding the presence of PMQR determinants in association with pAmpCs (20%), it is still important to report that PMQR genes are frequently linked to ESBLs and/or pAmpCs and they can be transferred by conjugation [3].
The high frequency of PMQR among ESBL and Amp C producers in this study could be attributed to the genetic linkage and horizontal transfer of these resistance genes on plasmids [4].
In conclusion, quinolones are important drugs used in the treatment of infections caused by S. marcescens. Therefore, the increased frequency of PMQRs among this organism over 6-year signifies a real therapeutic problem. Moreover, the close association between PMQR, ESBLs and Amp C genes in this organism would cause failure of treatment. This study recommends continuous surveillance to detect and molecular characterization of resistant bacteria isolates in hospitalized patients.