Pefloxacin, like other fluoroquinolones, accumulates in macrophages and several other types of nucleated cells (but not in erythrocytes). Upon fractionation of macrophage homogenates by isopycnic centrifugation in sucrose gradients, fluoroquinolones are not found associated with any specific cellular structure. We have compared the activities of pefloxacin and roxithromycin against intracellular Staphylococcus aureus in mouse J774 macrophages. Pefloxacin was significantly more active for equivalent intracellular drug concentrations (i.e. expressed by reference to the respective MICs of the drugs as determined in broth), suggesting differences in intracellular availability and/or capacity of the drugs to express their activity in the intracellular environment. The difference was enhanced by incubating the cells in acidic medium. We have also examined the cellular pharmacokinetics and intracellular distribution of pefloxacin in uninfected and Legionella pneumophila infected guinea pig macrophages. In contrast to uninfected cells from which pefloxacin was quickly released, macrophages infected with legionella retained approximately 20-30% of the accumulated pefloxacin after a 60-min wash-out. Cell fractionation studies indicated that the drug remaining in cells was associated with components of high buoyant density. These fractions also contained [3H] if cells had been incubated with [3H] labelled legionella (by in-vitro exposure to [3H]-thymidine, before phagocytosis). These results suggest that part of the intracellular pefloxacin becomes associated with legionella, or with legionella-containing cytoplasmic structures.

Pseudomonas aeruginosa is particularly resistant to most all the antibiotics presently available, essentially because of the very low permeability of its outer membrane. To overcome this, we synthesized four siderophore-based antibiotics formed by two quinolones - norfloxacin and benzonaphthyridone - bound to the pyoverdin of P. aeruginosa ATCC 15692 via two types of spacer arms: one stable and the other readily hydrolyzable. From the comparison of their antibacterial properties with those of the two unbound quinolones, we reached the following conclusions: (a) The adducts inhibit Escherichia coli's gyrase showing that the dissociation of the compounds is not necessary for their activity. However, the presence of the pyoverdin moiety on the molecule decreases the inhibition activity compared to the antibiotic alone. (b) They facilitate the uptake of (55)Fe using the specific pyoverdin-mediated iron-transport system of the bacterium. No uptake was observed either with P. aeruginosa ATCC 27853, which produces a structurally different pyoverdin, or with P. aeruginosa K690, which is a mutant of P. aeruginosa ATCC 15692 lacking FpvA, the outer-membrane pyoverdin receptor. (c) MIC determinations have shown that only strains P. aeruginosa ATCC 15692 and the derived outer-membrane receptor-producing but pyoverdin-deficient P. aeruginosa IA1 mutant present higher susceptibility to the pyoverdin-quinolone adducts, whereas P. aeruginosa ATCC 27853 and K690 are much more resistant. (d) Growth inhibition by these adducts confirmed these results and showed that the adducts with the hydrolyzable spacer arm have better activity than those with the stable one and that the labile spacer arm adducts present much higher activity than the quinolones alone. These results show clearly that the penetration of the antibiotic into the cells is favored when this latter is coupled with pyoverdin: Only the strains possessing the appropriate outer-membrane receptor present higher susceptibility to the adduct. In this case the antibiotic uses the pyoverdin-mediated iron-transport system. Furthermore, better efficiency is obtained when the spacer arm is labile and favors the antibiotic release inside the cell, allowing better inhibition of gyrase.

Three flavonoids which promoted Escherichia coli topoisomerase IV-dependent DNA cleavage were isolated from cottonseed flour and identified as quercetin 3-O-␤-D-glucose-[1,6]-O-␣-L-rhamnose (rutin), quercetin 3-O-␤-D-galactose-[1,6]-O-␣-L-rhamnose, and quercetin 3-O-␤-D-glucose (isoquercitrin). The most active one (rutin) also inhibited topoisomerase IV-dependent decatenation activity (50% inhibitory concentration, 64 g/ml) and induced the SOS response of a permeable E. coli strain. Derivatives of quercetin glycosylated at position C-3 were shown to induce two site-specific DNA cleavages of pBR322 DNA, which were mapped by DNA sequence analysis to the gene encoding resistance to tetracycline. Cleavage at these sites was hardly detectable in cleavage reactions with quercetin or fluoroquinolones. None of the three flavonoids isolated from cottonseeds had any stimulatory activity on E. coli DNA gyrase-dependent or calf thymus topoisomerase IIdependent DNA cleavage, and they were therefore specific to topoisomerase IV. These results show that selective inhibitors of topoisomerase IV can be derived from the flavone structure. This is the first report on a DNA topoisomerase inhibitor specific for topoisomerase IV.