Effects of a Mutation in the gyrA Gene on the Virulence of Uropathogenic Escherichia coli

Abstract

Fluoroquinolones are among the drugs most extensively used for the treatment of bacterial infections in human and veterinary medicine. Resistance to quinolones can be chromosome or plasmid mediated. The chromosomal mechanism of resistance is associated with mutations in the DNA gyrase- and topoisomerase IV-encoding genes and mutations in regulatory genes affecting different efflux systems, among others. We studied the role of the acquisition of a mutation in the gyrA gene in the virulence and protein expression of uropathogenic Escherichia coli (UPEC). The HC14366M strain carrying a mutation in the gyrA gene (S83L) was found to lose the capacity to cause cystitis and pyelonephritis mainly due to a decrease in the expression of the fimA, papA, papB, and ompA genes. The levels of expression of the fimA, papB, and ompA genes were recovered on complementing the strain with a plasmid containing the gyrA wild-type gene. However, only a slight recovery was observed in the colonization of the bladder in the GyrA complement strain compared to the mutant strain in a murine model of ascending urinary tract infection. In conclusion, a mutation in the gyrA gene of uropathogenic E. coli reduced the virulence of the bacteria, likely in association with the effect of DNA supercoiling on the expression of several virulence factors and proteins, thereby decreasing their capacity to cause cystitis and pyelonephritis.

INTRODUCTION

Fluoroquinolones are among the drugs most extensively used for the treatment of bacterial infections in human and veterinary medicine. They act by inhibiting the DNA gyrase and topoisomerase IV, which are tetrameric enzymes constituted by two A subunits and two B subunits. These subunits are encoded by the gyrA and gyrB genes, respectively, in the case of the DNA-gyrase and by the parC and parE genes, respectively, in the case of topoisomerase IV (1). The quinolones bind the DNA and the topoisomerase forming a quinolone-DNA-topoisomerase complex, avoiding the transcription or replication of DNA (1). The main mechanism of quinolone resistance is the accumulation of mutations in these two enzymes (2). Quinolone resistance can also be caused by the acquisition of qnr, a plasmid-mediated horizontally transferable gene (3). Two additional plasmid-mediated mechanisms of resistance to quinolones have also been identified, the AAC(6′)-Ib-cr protein, a variant aminoglycoside acetyltransferase capable of reducing ciprofloxacin (CIP) activity (4), and the efflux pump QepA (5).

The primary cellular target of fluoroquinolones in Escherichia coli is a type II topoisomerase (DNA gyrase) enzyme, which is unique in catalyzing negative supercoiling of covalently closed circular double-stranded DNA in an ATP-consuming reaction and is, therefore, essential for maintenance of DNA topology. Topoisomerase IV has been shown to be a secondary quinolone target in E. coli and decatenates the chromosome before cell division (6). Changes in DNA supercoiling in response to environmental factors contribute to the control of bacterial virulence (7).

Quinolone- and fluoroquinolone-resistant uropathogenic E. coli (UPEC) strains display reduced virulence in the invasion of immunocompromised patients. In contrast, susceptible E. coli strains are more virulent and affect immunocompetent hosts, showing a larger number of virulence factors contained in pathogenicity islands (PAIs) (8, 9). It has been demonstrated that a resistant E. coli strain becomes less virulent following the acquisition of a gyrA mutation (10) and that the loss of virulence by acquisition of quinolone resistance may take place before the acquisition of mutations and/or quinolone resistance levels (11).

The biological cost of quinolone resistance differs among different bacteria and depends on the level of resistance and the number of resistance mutations (12).

Compared to commensal strains, UPEC has several virulence factors that allow it to colonize host mucosal surfaces, injure and invade host tissues, overcome host defense mechanisms, and incite a host inflammatory response.

Among these virulence factors, type 1 fimbriae, P-fimbriae, and outer membrane proteins play an important role in several steps of urinary tract infection (UTI). Thus, type 1 pili promote adherence of UPEC isolates to superficial bladder epithelial cells, initiating a cascade of events that directly influence the pathogenesis of UTIs (13). In addition, type 1 fimbriae have been associated with invasion of bladder epithelial cells and the ability of bacteria to replicate intracellularly, forming “internal biofilms” (14).

P-fimbria (a mannose-resistant adhesin of UPEC) has been shown to be associated with acute pyelonephritis (at least 90% of acute pyelonephritis) (15).

Conversely, the OmpA protein is critical for promoting persistent infection within the epithelium and has been associated with cystitis and intracellular survival (16).

The aim of this study was to determine the role of the acquisition of a mutation in the gyrA gene in the virulence and protein expression of UPEC.

MATERIALS AND METHODS

Bacterial strains and selection of resistant mutants.

Three strains of E. coli were used in this study: (i) the HC14366 wild-type (HC14366wt) UPEC clinical isolate with an MIC of ciprofloxacin (CIP) of 0.008 mg/liter, (ii) its CIP-resistant mutant (E. coli HC14366M) with a mutation in the gyrA gene (S83L) and an MIC of CIP of 2 mg/liter, and (iii) the E. coli HC14366M mutant transformed with a plasmid carrying the wild-type gyrA gene, generating a complementation of the gyrA gene (E. coli HC14366MC) with an MIC of CIP of 0.064 mg/liter. Strain HC14366wt was grown at 37°C on MacConkey plates in the presence of ciprofloxacin in a multistep selection process to obtain strain HC14366M, a ciprofloxacin-resistant mutant. Ciprofloxacin (Fluka, Steinheim, Germany) was present only in agar plates during the selection procedures, starting at 0.004 mg/liter (half of the MIC for HC14366wt) and increasing 2-fold each step, until reaching a maximum concentration of 2 mg/liter. Single colonies were selected at each step and named according to the ciprofloxacin concentration of selection (e.g., strain HC14366-0.016 was selected at a CIP concentration of 0.016 mg/liter).

Antimicrobial susceptibility.

Susceptibility to several antimicrobial agents was determined in the presence and absence of 20 mg/liter of the efflux pump inhibitor Phe-Arg-β-naphthylamide using the agar dilution method according to the CLSI (17) guidelines as described elsewhere (18).

Virulence profile.

The virulence profile was analyzed by PCR using gene-specific primers for 17 virulence genes, including genes encoding hemolysin (hly), cytotoxic necrotizing factor (cnf), autotransporter (sat), P-fimbriae (pap genes), type 1C fimbriae (foc), yersiniabactin (fyu), heat-resistant hemagglutinin (hra), S-fimbriae (sfa), invasin (ibeA), adhesin (iha), aerobactin (aer), siderophores (iucC, iutA, iroN), and antigen 43 (Ag43 gene) (19).

Motility and type 1 fimbria expression.

The motility of each isolate was analyzed by growth in mannitol agar. Expression of type 1 fimbriae was determined by agglutination of Saccharomyces cerevisiae by the procedure described in reference 20.

Doubling time analysis.

The strains were grown in LB medium at 37°C with shaking. The optical density at 600 nm (OD₆₀₀) of each culture was measured in a Cecil CE 2302 spectrum. Aliquots were taken every 30 min for 6 h (21).

Animal model.

The virulence of each strain was tested in a murine model of an ascending UTI protocol approved by the Danish Ministry of Justice Animal Ethics Committee (approval no. 2004/561-835) and described in reference 22. In short, mouse bladders were emptied by gently pressing the abdomen, and 50 μl (5 × 10⁶ CFU) of each bacterial suspension was slowly inoculated transurethrally into 4 to 6 outbred female albino CFW1 mice (26 to 30 g; Harlan Netherlands, Horst, Netherlands) with the use of plastic catheters. The mice were housed 4 to 6 to a cage and were given free access to food and 5% glucose-containing water. Seventy-two hours after inoculation, urine was collected from each mouse. The mice were then euthanized by cervical dislocation, and the bladder and kidneys were removed and stored in Eppendorf tubes. The urine samples were processed the same day by spotting (20 μl) of a series of 10-fold dilutions (10⁰ to 10⁻⁶) in duplicate on bromothymol blue agar plates (SSI Diagnostika, Hillerød, Denmark). The bladder and kidneys were stored in 0.9% saline solution and were then incubated at room temperature for 1 h and subsequently homogenized using a TissueLyser (Qiagen, Ballerup, Denmark). Plates for bacterial counting were processed as described above. The detection limit was 25 CFU/sample. The experiment was repeated twice. The three strains were tested in parallel on the same day using the same batch of mice.

Reverse transcriptase PCR.

The strains were grown to an OD₆₀₀ of 0.5 in Luria-Bertani medium. One milliliter was centrifuged, and RNA from the pellet was extracted with TRI-Reagent solution (Ambion, Spain) following the manufacturer's instructions and treated with 1 μl of DNA-free DNase (Ambion, Spain). Reverse transcriptase PCR (RT-PCR) was performed using the AccessQuick RT-PCR system (Promega, Spain). Five hundred nanograms of RNA was used as the template. Specific primers were used for the housekeeping gap gene (used as an expression control) (5′-GTATCAACGGTTTTGGCCG-3′/5′-AGCTTTAGCAGCACCGGTA-3′) generating an amplicon of about 550 bp, the fimA gene (GGACAGGTTCGTACCGCATC/ACGTTGGTATGACCCGCATC) generating an amplicon of about 250 bp, the marA gene (CATTCATAGCTTTTGGACTGGAT/GTGTAAAAAGCGCGATTCGCC) generating an amplicon of about 150 bp, the papA gene (GGGGCAGGGTAAAGTAACTT/CAGGGTATTAGCATCACCT), and the papI gene (CGATGAGTGAATATATGAA/CACGAATTCTTATTAAGTTGTGGAAGA). The PCR was performed under the following conditions: one cycle of 45 min at 45°C and 3 min at 94°C followed by 26 to 28 cycles (fimA, marA, papA, and papI genes) or 16 cycles (gap gene) of 1 min at 94°C, 1 min at 56°C, and 1 min at 72°C. The PCR products were run in commercial acrylamide gels (GeneGel Excel; GE Healthcare, Spain) and stained with the PlusOne DNA silver staining kit (GE Healthcare, Spain). All experiments were carried out in triplicate.

Protein analysis.

Purification of whole proteins was performed using a sonication-based method (23). Two-dimensional gel electrophoresis was run for the protein extracts of these three strains, which were stained using a silver staining protocol to compare their patterns. The spots in the HC14366 wild-type E. coli showing a variation in the level of abundance compared to the mutant strain (E. coli HC14366M) and restored in the transformed E. coli (E. coli HC14366MC) were sliced and characterized by mass spectrometry analysis (matrix-assisted laser desorption ionization–tandem time of flight mass spectrometry [MALDI TOF/TOF]).

Real-time experiments.

RNA was extracted from exponential cultures and isolated using RNAprotect Bacteria Reagent (Qiagen, Hilden, Germany) and the RNeasy minikit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. All samples were treated with the DNA-free DNase kit (Ambion, Austin, TX) to remove contamination by genomic DNA, and a PCR was performed to confirm the loss of DNA. In this step, quantification of the RNA was carried out by Epoch (Biotek). Three independent RNA extractions of each sample were performed. Using a retro-transcription kit (TaKaRa catalog no. RR037Q), 500 ng of each RNA sample was used to perform reverse transcription. The cDNA template was diluted 1/5 for the RT-PCR. The ompA, ompF (each encoding an outer membrane protein related to virulence), and papB (one of the transcription regulators of papA) genes were selected, and the 16S rRNA gene was used as an endogenous control. Primer Express software was used to design the primers to amplify these genes. After several assays with different primer concentrations, a concentration of 3 μM was found to be optimal. Amplification was performed using a StepOne real-time PCR system (Applied Biosystems) using the Sybr Premix Ex Taq (Tli RNaseH Plus) kit (TaKaRa) and the following universal thermal cycling conditions: 2 min at 50°C (uracil N-glycosylase [UNG] activation) and 10 min at 95°C (enzyme activation) followed by 40 cycles of 95°C for 15 s (denaturation) and 60°C (anneal/extension) for 1 min. Data were analyzed with the StepOne software v2.0, and the relative level of expression of each sample (2^−ΔΔCT) was obtained.

Statistical analysis.

Data from the animal model experiments were analyzed using one-way analysis of variance (ANOVA) with SPSS software version 20. P values of <0.05 were considered to be significant.

RESULTS

The HC14366 UPEC strain was submitted to a multistep selection process in the presence of CIP, starting at 0.004 mg/liter (half of the MIC for the wild-type strain) and increasing 2-fold each step until reaching a maximum concentration of 5.12 mg/liter. The intermediate mutant HC14366M was chosen because it has a mutation in the quinolone resistance-determining region (QRDR) of the gyrA gene but not in the gyrB, parC, or parE genes. This mutation is found in codon 83 from Ser to Leu.

The HC14366M mutant was transformed with a plasmid carrying the wild-type gyrA gene. The resulting strain (HC14366MC) showed an MIC of CIP of 0.064 mg/liter. The MICs of different antimicrobial agents in the presence/absence of the efflux pump inhibitor Phe-Arg-β-naphthylamide were also determined (Table 1). The complemented strain HC14366MC was found to be less resistant to CIP, nalidixic acid, norfloxacin, and chloramphenicol than the mutant strain HC14366M.

TABLE 1.

MIC of the strains studied

Straina	MIC (mg/liter) forb:
	CIP	NAL	NAL-Inh	NX	NX-Inh	C	C-Inh
HC14366wt	0.008	3	0.19	0.047	0.125	6	2
HC14366M	2.56	>256	>256	6	16	24	4
HC14366MC	0.064	6	0.38	0.5	0.5	16	3

^aHC14366wt, wild-type strain; HC14366M, gyrA mutant strain; HC14366MC, complemented strain.

^bCIP, ciprofloxacin; NAL, nalidixic acid; Inh, efflux pump inhibitor Phe-Arg-β-naphthylamide; NX, norfloxacin; C, chloramphenicol.

The HC14366 wild-type strain and its mutants showed the following virulence factors: hemolysin (hly), cytotoxic necrotizing factor (cnf1), autotransporter (sat), yersiniabactin (fyuA), type 1 fimbriae (fimA), P-fimbriae (pap genes), hemagglutinin (hra gene), S-fimbriae (sfaS), and siderophore (iroN). The HC14366M and HC14366MC strains showed a decrease in the motility through mannitol and in the expression of type 1 fimbriae in comparison with the wild-type strain. Therefore, expression of type 1 fimbriae and motility are not affected by a mutation in the gyrA gene.

The doubling time of the three strains was studied, showing that a mutation in the gyrA gene affects bacterial growth, and the complemented strain showed a higher doubling time value than the mutant strain but did not fully recover the wild-type levels (data not shown).

These three strains were inoculated into six mice of an animal model of ascending UTI, and urine, bladder, and kidney samples were collected. It is noteworthy that the HC14366M strain lost the capacity to cause cystitis and pyelonephritis, with an average of 10⁵ CFU/ml, 10² CFU, and 10⁰ CFU found in urine, the bladder, and the kidneys, respectively, compared with the values observed in the wild-type strain: 10⁸ CFU/ml urine (P = 0.032), 10⁷ CFU/bladder (P = 0.002), and 10⁴ CFU/two kidneys (P = 0.042). The HC14366MC strain had an increased capacity to cause cystitis, with around 10⁴ CFU (P = 0.011) in the bladder, but did not have the capacity to cause pyelonephritis (P = 0.043) (Fig. 1).

FIG 1.

Results of an animal model of ascending urinary tract infection. (A) HC14366 wild-type strain; (B) HC14366M strain; (C) HC14366MC strain. Vertical axes, CFU per ml of urine or per gram of bladder or pair of kidneys.

In order to determine the cause of the decrease of colonization in the mutant strain, RT-PCR was carried out using specific primers for the fimA and papA genes, involved in cystitis and pyelonephritis, respectively. The expression of the two genes was found to be decreased in the HC14366M strain, and only fimA expression was recovered in the complemented strain. On the other hand, marA was overexpressed in the mutant and complemented strains compared with the wild-type strain (Fig. 2).

FIG 2.

RT-PCR of the strains studied. MC, HC14366MC strain; M, HC14366M strain; wt, HC14366 wild-type strain.

In order to study the cause of the decrease in the expression of the papA gene in the mutant and complemented strain, we studied the regulators Lrp, PapI, and PapB. A total inhibition of papB and papI gene expression was found in the HC14366M strain, with papB expression recovered in the HC14366MC strain (Fig. 3).

FIG 3.

Real-time PCR of the genes selected. HC14366, wild-type strain; HC14366M, mutant strain; HC14366MC, complemented strain; RQ, relative quantity.

Protein analysis revealed changes in protein expression in the three strains (Table 2; Fig. 4). These changes included proteins implicated in cellular permeability, metabolic functions, and DNA replication. Among the proteins with decreased expression in the HC14366M strain but with the recovery of wild-type levels in the HC14366MC strain, we found the outer membrane protein A precursor, aspartate ammonia-lyase, the maltose-binding periplasmic protein, tryptophanyl-tRNA synthetase, the d-ribose periplasmic binding protein, the pyruvate kinase I protein, and a phosphate acetyltransferase. Conversely, the DNA-directed RNA polymerase, two dehydrogenases, and the heat shock protein HtpG were overexpressed in the HC14366M but not in the HC14366wt or HC14366MC strain. In addition, the expression of the outer membrane protein F (porin) decreased in HC14366M and its complemented strain (Table 2).

TABLE 2.

Proteins characterized by two-dimensional SDS-PAGE

IDa	Protein	Spot intensityb
		HC14366wt	HC14366M	HC14366MC
0J	Aspartate ammonia-lyase	+++	+	+++
3J	Glycerol kinase	+	−	−
6J	Outer membrane protein (OmpF)	++	+	−
7J	Maltose-binding periplasmic protein precursor	+++	+	++
8J	Aminomethyltransferase	++	+	+
9J	Outer membrane protein A (OmpA)	++	−	++
11J	Phosphotransferase enzyme IIAB, mannose specific	++	−	+
12J	d-Ribose periplasmic binding protein	+++	+	++
13J	DNA-directed RNA polymerase	+	++	+
14J	Pyruvate kinase I	++	+	++
15J	6-Phosphogluconate dehydrogenase	+	++	+
16J	Succinyl-coenzyme A synthetase	+	++	++
18J	Dihydrolipoamide dehydrogenase	++	+++	++
20J	Tryptophanyl-tRNA synthetase	+	−	+
21J	Phosphate acetyltransferase	+	−	+
22J	HtpG, heat shock protein	−	+	−
24J	Adenylosuccinate synthetase	++	−	+
25J	Phosphoglycerate kinase	+++	+	++
26J	Tartronate semialdehyde reductase	+	−	+
28J	Isocitrate dehydrogenase	−	+	−
29J	Cell division inhibitor	+	−	++

^aID, identification number from Fig. 4.

^bHC14366wt, wild-type strain; HC14366M, gyrA mutant strain; HC14366MC, complemented strain. +++, high protein expression; ++, moderate protein expression; +, low protein expression; −, no protein expression.

FIG 4.

Two-dimensional SDS-PAGE protein gels. (A) HC14366 wild-type strain; (B) HC14366M strain; (C) HC14366MC strain.

RNA expression of the genes encoding some proteins possibly related to virulence (MalE, OmpA, OmpF, and PapB) was analyzed, confirming the data obtained in the protein experiments (Fig. 3).

DISCUSSION

Since their introduction into clinical use in 1983, fluoroquinolones have played an essential role in the treatment of infectious diseases caused by enteric bacteria such as E. coli. However, a progressive increase in the emergence of fluoroquinolone-resistant strains has been observed in the last few decades (24). Two types of mutants are predominantly found among clinical isolates: low-level resistant isolates (CIP MIC, <2 mg/liter) frequently carrying a single gyrA mutation, which generates a substitution of serine 83 to leucine (S83L), and high-level resistant isolates (CIP MIC, >4 mg/liter) carrying two gyrA mutations in addition to mutations affecting serine 80 (S80) and glutamic acid 84 (Glu84) in parC (21).

The in vitro mutant obtained in our laboratory contained the single gyrA mutation most frequently found in clinical isolates (S83L).

The level of global supercoiling in E. coli isolates is mainly regulated by the DNA-gyrase (25). The accumulation of mutations in genes that encode the essential enzymes involved in the control of DNA topology can affect the regulation of the degree of supercoiling. Thus, the expression of supercoiling-regulated genes in laboratory mutants is commonly associated with a fitness cost (probably due to the overexpression of an unknown efflux system), observed as a reduced growth rate and/or virulence in the absence of antibiotic (21). In accordance with the results obtained in our study, Bagel et al. (21) observed that a mutant with a single S83L mutation in the gyrA gene showed an increase in the doubling time and, therefore, a decrease in the growth rate compared with the wild-type strain. Moreover, in the present study, an increase in the doubling time was observed when a gyrA wild-type gene was introduced into the mutant strain, albeit not to wild-type levels. These results indicate that gyrA is involved in the growth rate of E. coli.

Changes in DNA supercoiling affect antimicrobial resistance levels. Thus, the introduction of the gyrA wild-type gene in the HC14366M strain caused a reduction in the MICs of CIP and nalidixic acid (from 2.56 to 0.064 mg/liter and from >256 to 6 mg/liter, respectively), indicating that this mutation contributes to the expression of quinolone resistance as described previously (21).

Changes in DNA supercoiling can also contribute to the control of bacterial virulence (7). The mutation in the gyrA gene in the strain under study seemed to cause changes in its capacity to develop cystitis and pyelonephritis. First, a reduction in type 1 fimbria expression was exhibited by the mutant strain, preventing it from colonizing the bladder and, therefore, from causing cystitis. The finding that the introduction of the gyrA wild-type gene did not significantly (P = 0.456) modify the capacity of the mutant strain to cause cystitis may be due to the fact that transcription from the fimA promoter was not totally affected by changes in DNA supercoiling, as demonstrated by Dove et al. (26) on introducing a topA::Tn10 mutation or inhibiting the DNA-gyrase with the antibiotic novobiocin.

Another change in virulence as a consequence of the acquisition of a mutation in the gyrA gene is a decrease in P-fimbria expression, leading to a decrease in the capacity of the mutant strain to cause pyelonephritis. Expression of pyelonephritis-associated pili (Pap) in E. coli isolates is under a phase-variation control mechanism in which individual cells alternate between pilus-positive (on) and pilus-negative (off) states through a process involving DNA methylation by deoxyadenosine methylase (Dam) and regulation via Lrp (27).

Control of P-fimbria expression also requires the action of PapI, a positive regulator that increases the affinity of Lrp for the binding sites, and PapB, the second specific regulator of the Pap operon that plays an important role at a transcriptional level primarily by coordinating the expression of papBA and papI promoters (28).

In our strain, Lrp and PapI seem to be functional. However, a decrease in papB and papI expression was found in the mutant strain, with only papB expression recovered in the complemented strain, albeit not at wild-type levels and not reflected in the ability of the complemented strain to colonize the kidney.

Tessier et al. (29) studied F165 adhesin from E. coli. This adhesin belongs to the family of Pap-related fimbriae, the expression of which is mediated by regulatory proteins such as Lrp, Dam-methylase, and FooI and FooB. They found that inactivation of the gyrA gene caused a decrease in supercoiling, producing a decrease in fooB expression and inducing a decrease in P-fimbria expression. FooB is the equivalent of PapB in the P-fimbriae. The decrease of papB expression found in the present study can explain the decrease of P-fimbria expression, thereby making the mutant strain unable to adhere to renal epithelial cells and cause pyelonephritis. Although papB expression was recovered in the complemented strain, the finding that it did not recover the ability to cause pyelonephritis may be due to the fact that other P-fimbria regulators (such as PapI) were not affected by the inclusion of the plasmid containing the functional gyrA gene.

Finally, the introduction of a mutation in the gyrA gene may cause changes in the expression of different proteins.

Treatment with fluoroquinolones can induce heat shock responses (30). For example, levofloxacin produced an overexpression of several heat shock proteins when the strain was incubated with this antibiotic (30), with HtpG as one of these proteins. HtpG is the bacterial homologue of Hsp90 (present in yeast and humans) and is dispensable under nonstress conditions. HtpG comprises a large fraction (0.36%) of all the proteins in E. coli isolates growing at 37°C (31). In the present study, this protein was found to be overexpressed in the mutant strain, and its expression achieved wild-type levels in the strain complemented with a plasmid containing the gyrA gene. Therefore, the transcription of HtpG is mainly regulated by supercoiling.

OmpA is a major, monomeric, integral protein component of the outer bacterial membrane that functions as a critical determinant of intracellular virulence for UPEC, promoting persistent infection within the bladder epithelium (32). The fact that the HC14366M strain has a significantly lower bladder colonization rate than the HC14366 wild-type strain may be in accordance with the decrease in the expression of this gene. The recovery of ompA expression together with that of fimA can explain the increase in bladder colonization from 10² CFU/g to 10⁴ CFU/g.

OmpF is also one of the major outer membrane proteins of E. coli, the expression of which is extremely and specifically sensitive to the level of DNA supercoiling (33). Our results are in accordance with the study by Graeme-Cook et al. (33) which found that a gyrA mutant strain exhibited decreased OmpF expression probably due to overexpression of marA (34).

In conclusion, a mutation in the gyrA gene of UPEC causes a decrease in the virulence of the bacteria due to the effect of DNA supercoiling on the expression of several virulence factors and proteins, thereby decreasing the capacity to cause cystitis and pyelonephritis. This study demonstrates the relationship between virulence and the acquisition of antimicrobial resistance in vivo.