Quinolone resistance locus nfxD of Escherichia coli is a mutant allele of the parE gene encoding a subunit of topoisomerase IV.

The locus nfxD, which contributes to high-level quinolone resistance in Escherichia coli KF111b (gyrAr nfxB nfxD), is only expressed in the presence of a gyrA mutation, and maps to the region of the parC and parE genes, was outcrossed into strain KF130, creating strain DH161 (gyrAr nfxD). DNA sequence analysis of DH161 revealed no changes in the topoisomerase IV parC quinolone resistance-determining region but did identify a single T-to-A mutation in parE at codon 445, leading to a change from Leu to His. Full-length cloned parE+ partially complemented the resistance phenotype in KF111b and DH161, but did not complement the resistance phenotype in strain KF130 (gyrAr). No complementation was seen with cloned, truncated parE+. To confirm these findings, gyrAr was first outcrossed from KF130 into E. coli W3110parE10 [parE temperature sensitive(Ts)] and KL16. The transduced strains KL16 and W3110parE10 were subsequently transformed with plasmids containing cloned parE from DH161 or KL16. Cloned parE from DH161 increased norfloxacin resistance in the parE(Ts) background twofold at 30 degrees C and fourfold at 42 degrees C compared to those for cloned parE from KL16. The same experiment with a non-Ts background revealed a twofold increase in the norfloxacin MIC at both 30 and 42 degrees C. These data identify the nfxD conditional resistance locus as a mutant allele of parE. This report is the first of a quinolone-resistant parE mutant and confirms the role of topoisomerase IV as a secondary target of norfloxacin in E. coli.

The effects of quinolones on the adherence of type-1 fimbriated Escherichia coli to mannosylated agarose beads.

Bacterial adherence is reported to be antagonized by several classes of antibiotics including quinolones, beta-lactams and tetracyclines, based primarily on in-vitro studies in which bacterial cells are exposed to antimicrobials, incubated in the presence of uroepithelial cells (UECs) and assessed for adherence by light microscopy. Some problems associated with the use of this approach, include low sensitivity, high variability and, in the case of adherence of mannose-sensitive Escherichia coli interference by mannose-containing uromucoid. To avoid these problems, mannosylated agarose beads (MABs) were used as a model for UECs. Adherence of E. coli strain AAEC356, which is constitutive for type-1 fimbrial expression, was maximal with 3 x 10(4) beads/mL and 1 x 10(8) bacterial cells/mL co-incubated for 35 min at 37 degrees C. Those bacterial cultures showed 40-60% adherence to MABs but only 4-10% adherence to UECs. This study reports a novel method to detect mannose-sensitive bacterial adherence, using MABs, in order to determine the effects of quinolones, cefdinir and tetracycline on E. coli adherence. Cefdinir and the quinolones ciprofloxacin, enoxacin and PD131628 caused significant reductions in the adherence of AAEC356 to UECs at concentrations equivalent to 1/2 x MIC, while up to 1 x MIC of these antibiotics had no significant effect on adherence to MABs. A direct comparison of UEC to MAB-based techniques showed that PD131628, at concentrations equivalent to 1/16x, 1/4x, 1/2x and 1 x MIC, had no effect on bacterial adherence to MABs, while reductions of 34%, 38%, 87% and 85% respectively were seen in adherence to UECs. The anti-adherent effect mediated by quinolones may not therefore be related to the specific interaction between type-1 fimbriae and mannosylated receptors. While quinolones and cefdinir had no effect on overall bacterial adherence to MABs, there was a decrease in the ability of alpha-methyl-D-mannoside (alpha-MM) to inhibit competitively this adherence. Concurrent exposure of PD131628 or cefdinir with tetracycline prevented this, suggesting that protein synthesis is required for this effect.

Modulation of Escherichia coli type 1 fimbrial expression and adherence to uroepithelial cells following exposure of logarithmic phase cells to quinolones at subinhibitory concentrations.

Quinolone antibiotics at sub-inhibitory concentrations have been shown to antagonize the adherence of Escherichia coli to urinary tract epithelium. This may be due either to reduced expression or to alterations to the structure of the fimbriae which mediate adherence. While E. coli cells in the stationary growth phase have previously been used to investigate quinolone-induced inhibition of adherence, the present study has demonstrated that bacteria in the logarithmic phase also produce type 1 fimbriae and that the adherence of these organisms is reduced following exposure to various quinolones. In all experiments, cells in the logarithmic phase were incubated for 3 h in the presence of ciprofloxacin, enoxacin, CI-960 or PD131628 at a concentration equivalent to 0.5 x MIC. An in-vitro adherence assay which used acid-washed uroepithelial cells and a type 1-fimbriated strain of E. coli showed reductions in adherence of 47%, 72% and 95% after exposure to enoxacin, ciprofloxacin and PD131628, respectively. The effects of ciprofloxacin, enoxacin, CI-960 and PD131628 on two phase variation controlling genes, fimB and fimE, and the main structural gene, fimA, were evaluated by quantifying beta-galactosidase production encoded by chromosomally-located fim::lacZ fusions. All four quinolones tested caused reductions in beta-galactosidase production by a fimA::lacZ fusion strain, but did not significantly affect production of this enzyme by fimB::lacZ and fimE::lacZ fusion strains; these agents also led to decreases in wild-type beta-galactosidase production. Amplification of the invertible element after exposure to enoxacin at 0.25, 0.5 or 1 x MIC revealed no changes in orientation distribution compared with the antibiotic-free control. In addition, a fluorescence assay specific for type 1 fimbriae showed only 23%, 21%, 25% and 11% reductions in the surface expression of the structural subunit after incubation in the presence of ciprofloxacin, enoxacin, CI-960 and PD131628, respectively, at 0.5 x MIC.