Characterization of genetic structures of the QepA3 gene in clinical isolates of Enterobacteriaceae

QepA is one of the genes that confer quinolone resistance in bacteria. The aim of this study was to analyze the genetic structures of plasmids that carry a qepA3, a recently discovered allele of qepA in Enterobacteriaceae clinical isolates. 656 non-redundant Enterobacteriaceae clinical isolates were screened for the qepA3 gene and five isolates were identified to carry the gene. Plasmids were isolated from these isolates and were found to increase antibiotic resistance once the plasmids were transferred to Escherichia coli. These plasmids were subcloned and sequenced to analyze the genetic structures surrounding the qepA3 gene. The results showed that the five plasmids had different genetic structures; two of the qepA3-containning isolates had either the blaCTX-M-14 or blaTEM-12 gene instead of the blaTEM-1 gene. The structures of both pKP3764 and pECL3786 have not been previously described. In comparison with pHPA, there were a number of changes in DNA sequences up- and down-stream of the qepA3 gene. These findings provide better understanding of the genetic variations in qepA3 and would be useful for diagnosis and control of quinolone resistance in clinical settings.

QepA is one of the genes that confer quinolone resistance in bacteria. The aim of this study was to analyze the genetic structures of plasmids that carry a qepA3, a recently discovered allele of qepA in Enterobacteriaceae clinical isolates. 656 non-redundant Enterobacteriaceae clinical isolates were screened for the qepA3 gene and five isolates were identified to carry the gene. Plasmids were isolated from these isolates and were found to increase antibiotic resistance once the plasmids were transferred to Escherichia coli. These plasmids were subcloned and sequenced to analyze the genetic structures surrounding the qepA3 gene. The results showed that the five plasmids had different genetic structures; two of the qepA3-containning isolates had either the bla CTX−M−14 or bla TEM 12 gene instead of the bla TEM 1 gene. The structures of both pKP3764 and
In this study, we investigated the qepA gene in 656 Enterobacteriaceae isolates from hospitalized patients and only 13 isolates were found positive. This result confirm that qepA has low prevalence (1.98%) (0.91% for qepA1, 0.31% for qepA2 and 0.76% for qepA3 ) in patients of our area. However, five isolates were determined to harbor novel qepA3 structures. These structure were characterized to provide better understanding of the gene for potential management of plasmid-mediated antibiotic resistance.

Bacterial Strains and Luria-Bertani (LB) Agar Plates
Enterobacteriaceae isolates EC3157, EC3587, CD4359, KP3764, and ECL3786 were identified using the Vitek 2 system (bioMérieux, France) according to the manufacturer's instructions. Escherichia coli isolates EC3157 and EC3587 were isolated from the blood and sputum samples of ICU inpatients, respectively. CD4359, an isolate of Citrobacter koseri, was obtained from the sputum of an inpatient in Infection Unit, while KP3764, an isolate of Klebsiella pneumoniae, was from a blood sample of an ICU inpatient. ECL3786, an isolate of Enterobacter cloacae, was from the chest wound secretions of a cardiothoracic surgical inpatient. E. coli JM109 was used as the host for cloning and an azide-resistant E. coli (strain J53) was used as the recipient strain for conjugation experiments. The LB1 agar medium for plasmid transformation contained 100 μL of isopropyl β-D-1-thiogalactopyranoside (IPTG, 24 mg/mL), 200 μL of 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside (X-Gal, 20 mg/mL) and 100 μL of ampicillin (Amp, 100 mg/mL) in 100 mL medium. LB2 agar plates were used for conjugation experiments and were supplemented with sodium azide (150 μg/mL) and ciprofloxacin (0.25 μg/mL).

Susceptibility Testing
The minimum inhibitory concentrations (MICs) of eight antimicrobial agents (nalidixic acid, ofloxacin, ciprofloxacin, cefotaxime, ceftazidime, amikacin, and gentamicin) were determined using the MicroScan microdilution panel (Scott, USA) broth dilution method. E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as the controls. The results were interpreted according to the CLSI guidelines (Clinical and Laboratory Standards Institute, 2014).

Plasmid Isolation and Sequence Analysis
Bacterial plasmid DNA was extracted using a plasmid extraction kit (TaKaRa, Japan) according to the manufacturer's instructions. PCR amplifications were preformed using primers based on the pHPA ( Table 1, Yamane et al., 2007). PCR was run for 3 min at 94 • C followed by 30 cycles of 1 min of denaturing at 94 • C and annealing at 56.9 • C, with a final elongation of 10 min at 72 • C on Life Veriti R PCR machine (Invitrogen, USA). The total reaction volume was 20 μl containing 4 μL 5X PCR buffer, 0.4 μL of 10 mM dNTPs, 1 μL each of 10 μM primers and 0.2 μL Polymerase, with nuclease-free water filled up to 20 μL, and bacterial plasmids harboring the qepA gene as template. The amplicons were digested with DraI and BamHI (TaKaRa, Japan), ligated to linearized pMD19-T (TaKaRa, Japan), and transformed into E. coli JM109 competent cells. Plasmids DNA from Amp resistant colonies recovered on LB1 plate were sequenced.

Conjugation
The five isolates were conjugated with E. coli J53 as described previously (Wang et al., 2003). In the conjugation experiments, the isolates were used as donors and azide-resistant E. coli J53 as the recipient strain by filter mating. Transconjugants were selected on the LB2 agar plates. MICs for the donors, transconjugants, and recipients were also measured as described above.  The qep A-containing sequences in the five strains were amplified, digested with DraI and/or BamHI and subcloned in pMD19-T to generate recombinant plasmids pECD1, pECD2, pECD3, pECD4, pCDD2, pKPD1, and pECLD1 ( Figure 1B). The inserts in these plasmids were sequenced and analyzed for qep A and its flanking structures (GenBank accession numbers KR259130, KR259131, KR259132, KR259133, and KR259134, Supplementary Material) (Figure 2). The results showed that in addition to qepA3, which was present in all plasmids, the inserts from pEC3157 and pCD4359 contained the bla CTX−M−14 and rmtB genes; the inserts from pEC3587 had the bla TEM−12 gene; inserts from all plasmids except pCD4359 had the truncated dfr2 gene; the inserts from pCD4539 contained the bla TEM−1 and rmtB genes; the insert from pKPD1 and pECLD1 contained truncated dfr2 gene. No other resistant genes were found in these sequences.

Antibiotic Susceptibility
Through conjugation, the plasmids from the five isolates (donors) were transferred to E. coli J53 (recipients). The MIC values for cefotaxime and ceftazidime in the transconjugants with the isolates EC3157, EC3587, and CD4359 were much higher than those of the recipient E. coli J53 (Table 2). Similarly, the MIC values for nalidixic acid, norfloxacin, ofloxacin, and ciprofloxacin in the transconjugants with the five isolates were generally higher than those of E. coli J53 ( Table 2). Contrastingly, the MIC values nalidixic acid, ofloxacin, ciprofloxacin, amikacin, and gentamicin in the transconjugant involving the EC3157, EC3587, CD4357, and ECL3786 were generally higher than those of E. coli J53 (recipient) ( Table 2). These results demonstrated that the antibiotic resistant genes are located in the plasmids.

DISCUSSION
Previous studies have shown that quinolone and fluoroquinolone resistance genes in E. coli and Klebsiella isolates are located on plasmids, which often carry other antimicrobial resistant genes, and can be transferred to other strains by conjugation or transformation (Ruiz et al., 2012). Since the discovery of FIGURE 2 | Genetic structures surrounding the qepA3 gene in pEC3157, pEC3587, pCD4359, pKP3764, and pECL3786. The numbers in parentheses are GenBank accession numbers; "ggatcc" and "tttaaa" indicate restriction sites; (delta) denotes truncated gene; intI1-groEL denotes the truncated intI1 and groEL fusion gene. The figures in the parentheses following the names of the plasmids denote the plasmid size. qepA1 and qepA2 in 2007 and 2008, the genetic environment and location of the qepA genes in the plasmid pHPA have both been well-established (Périchon et al., 2007;Yamane et al., 2007;Cattoir et al., 2008). In this study, five Enterobacteriaceae isolates carrying the qepA3 and surrounding genes were investigated, and are found to have different genetic structures (Figure 2) surrounding qepA3. The plasmid pCD4359 from isolate CD4359 showed a similar structure to pHPA, except for the qepA allele and a truncated dfr2, a gene closely related to vertebrate FGF-receptor. The plasmids pEC3587 and pEC3157 were different from pHPA, where bla TEM−1 in pHPA was replaced by bla TEM−12 or bla CTX−M−14 , respectively (Figure 2). bla CTX−M−14 ,bla TEM−1, and bla TEM−12 code β-lactamase and are important determinants of drug resistance. In addition, in pCD4359, there is an additional intI1 downstream of the rmtB which confers high-level resistance to all aminoglycoside and missing dfr2. The sequence between IS26 and tnpR was also truncated. Furthermore, rmtB or ISCR3C are not linked with qepA3 directly in pKP3764 as reported previously (Périchon et al., 2007;Yamane et al., 2007;Cattoir et al., 2008;Kim et al., 2009;Rocha-Gracia et al., 2010;Cao et al., 2014).
In pEC3587, the bla TEM−12 gene took the place of bla TEM−1 and had a truncated dfr2 gene as compared with pHPA. These results are also different from previous reports (Périchon et al., 2007;Yamane et al., 2007;Cattoir et al., 2008;Kim et al., 2009;Rocha-Gracia et al., 2010;Cao et al., 2014). Surprisingly, in pKP3764 there are three genes qepA3, truncated dfr2, and truncated intI1-groEL fusion gene between the two IS26 insertions, while in pKP3764 and pECL3786, therefore are four genes: rmtB, qepA3, truncated dfr2, and truncated intI1-groEL fusion genes. To date, these genetic structures have not been identified in previous studies Rocha-Gracia et al., 2010;Cao et al., 2014). Moreover, pKP3764 is the only plasmid that does not contain the rmtB gene among the five plasmids studied in this study. According to a previous report, 58.3% (28/48) of rmtB-positive E. coli isolates harbored the qepA gene (Grape et al., 2005). Whether or not these results suggest a strong link between qepA and rmtB remains to be investigated. In these five genetic structures (Figure 2), the qepA3 gene and its downstream genes were nearly all identical except for the truncated dfr2 deletion in pCD4359. After pEC3157 and pEC3587 were digested by the DraI, the size of both products were identical at approximate 6.2 kb. It should be noted that it is also possible that the ISCR3C is not a stable link between ISCR3C and the qepA gene.
QepA gene is a quinolone pump gene that confers resistance to nalidixic acid and norfloxacin. Stains carrying the gene may be resistant or sensitive to ofloxacin and ciprofloxacin with increased MIC. In the conjugation experiments, we found that E. coli cells with pEC3587 and pCD4359 were resistant to penicillins and aminoglycoside antibiotics drugs such as amikacin, gentamicin, gentamicin, and tobramycin. They were tolerant or sensitive to ofloxacin and ciprofloxacin, but the MICs were increased. These results indicate that there might be synergistic effect against antibiotics when the gepA, blaCTX-M 14, and rmtB are present in the same plasmids, leading to multidrug resistance.
Taken together, our works suggest that there are sequence variations surrounding the qepA3 even if a limited number of isolates are analyzed, and it is likely that more variations would exist that may impact the resistance profiles, and subsequently the clinical implications of the bacteria. More studies are needed to link these structure variation to resistance profiles and potential clinical outcomes.

CONCLUSION
The novel genetic structures surrounding the qepA3 gene have been discovered in the isolates obtained from five patients in China although the prevalence for the qepA allele in hospital patients are low. One of the isolates is linked to the non-rmtB-or non-ISCR3C-producing genetic structure in the qepA3 genetic environment. Moreover, the bla CTX−M−14 or bla TEM−12 genes are found to be associated with the qepA3 gene in these structures instead of the bla TEM−1 gene. These results provide new insight into the variation in genetic environments of the qepA3 gene and would be useful for further investigation of the clinical implications in antibiotic resistance management.