Investigation of Smith's quinolone killing mechanisms during the PAE of ciprofloxacin on Escherichia coli.

Quinolone antibacterials interact with the DNA-DNA gyrase complex, but subsequent events that lead to cell death are unresolved. Three distinct mechanisms of quinolone lethality have been identified by Smith and co-workers: Mechanism A, which requires RNA and protein synthesis and cell division for expression; Mechanism B, which remains active when these functions are precluded; and Mechanism C, which is active on non-dividing cells. Exposure to 4x MIC ciprofloxacin (Cip) in nutrient broth (NB) for 3 h reduced the viability of Escherichia coli AB1157 to 0.25%. Addition of rifampicin (Rif) or chloramphenicol (Cm), to inhibit RNA or protein synthesis, respectively, increased survival 70-fold. Treatment of cells with Cip in phosphate-buffered saline (PBS), to inhibit cell division, increased survival 20-fold. No further cell death occurred once the various drug combinations or PBS had been washed out and cells resuspended in drug-free nutrient broth. These latter conditions allow expression of the post-antibiotic effect (PAE). PAE was lengthened in cells exposed to Cip in the presence of Rif or Cm, probably as a result of delay in the initiation of inducible DNA repair.

Flow cytometric investigation of filamentation, membrane patency, and membrane potential in Escherichia coli following ciprofloxacin exposure.

Ninety-eight percent of the cells in a population of Escherichia coli in log-phase growth lost colony-forming ability after being exposed for 3 h to the quinolone antibiotic ciprofloxacin at four times the MIC in nutrient broth, a concentration easily reached in vivo. Flow cytometric analysis, however, demonstrated that only 68% of this bacterial population had lost membrane potential, as judged by the membrane potential-sensitive dye bis-(1,3-dibutylbarbituric acid) trimethine oxonol [DiBAC(4)(3)], and only 30% could no longer exclude the nucleic acid-binding dye propidium iodide (PI), reflecting lost membrane integrity, efflux mechanisms, or both. Subsequent removal of ciprofloxacin and resuspension in nutrient broth resulted in renewed cell division after 2 h, with a calculated postantibiotic effect (PAE) time of 57 min. The proportion of DiBAC- and PI-fluorescent cells in this recovering population remained stable for more than 4 h after antibiotic removal. Eighty percent of cells present at drug removal were filamentous. Their number subsequently decreased with time, and the increase in particle count seen at the end of the PAE resulted from the division of short cells. Exposure to ciprofloxacin in the presence of the protein synthesis inhibitor chloramphenicol increased colony-forming ability to 60% of starting population numbers. In contrast to ciprofloxacin alone, this antibiotic combination resulted in insignificant filamentation and no dye uptake. Subsequent drug removal and resuspension in nutrient broth resulted in the appearance of filaments within 1 h, with 69% of the population forming filaments at 3 h. Dye uptake was also seen, with 20% of the population fluorescing with either dye after 4 h. We were unable to relate dye uptake to the viable count. Cell division resumed 240 min after removal of both drugs, yielding a PAE calculated at 186 min. Inhibition of protein synthesis with chloramphenicol prevented ciprofloxacin-induced changes in bacterial morphology, cell membrane potential, and ability to exclude nucleic acid-binding dye. These changes persisted beyond the end of the classically defined PAE and were not a definite indicator of cell death as defined by loss of colony formation, which related at least in part to filamentation.

Susceptibilities of Actinobacillus actinomycetemcomitans biofilms to oral antiseptics.

The susceptibilities of Actinobacillus actinomycetemcomitans cultures, grown as 1- or 3-day-old biofilms or as planktonic suspensions, to concentrations of chlorhexidine digluconate, cetylpyridinium chloride or triclosan used in commercial mouthwashes were compared. Three-day biofilms were the most resistant form of the organism and chlorhexidine was the most active antiseptic. Comparison of solutions of the pure antibacterial agent with commercial products containing the same concentration of antiseptic showed little difference in in-vitro activities. The results emphasise that the testing of antimicrobial mouthwashes should be performed on bacteria grown as biofilms.

Post-UV survival of Escherichia coli strains carrying more than one UV-sensitising plasmid.

The post-UV phenotypes conferred by wild-type plasmids R391 and pYD1, which increase UV-induced mutagenesis but sensitise Escherichia coli AB1157 umuC+ uvrB+ to UV, were compared, alone and in combination with that of plasmid pGW16, which sensitises AB1157 to low, but protects against high UV doses. All three plasmids increased UV resistance when present in Shigella sonnei. No plasmid significantly affected the UV sensitivity of E. coli TK501 umuC uvrB, in which pKM101, the parent of pGW16 increases UV resistance up to 1000-fold. Both pYD1 and R391 reduced the UV protective effect of pKM101, and increased UV-sensitisation conferred by pGW16. UV-sensitisation conferred by pYD1 and R391 was additive when the plasmids were together in strain AB1157, and both pKM101 and pGW16 reduced this additive sensitisation.

Positive R plasmid mutator effect on chromosomal mutation to nalidixic acid resistance in nalidixic acid-exposed cultures of Escherichia coli.

Mutation frequencies to nalidixic acid resistance (15 mg/L in nutrient agar) were determined for derivatives of Escherichia coli AB1157 carrying the mutator plasmids R46, R391 or pYD1, or the non-mutator plasmid RP4. Frequencies of mutation remained constant in cultures of AB1157(R46) growing exponentially in drug-free broth, at a level about 12-fold higher than in the strain without plasmid. Mutation frequencies in cultures of strains AB1157(R391) and AB1157(pYD1) were about three times greater than in the control, whereas plasmid RP4 had no effect on spontaneous mutation frequency to nalidixic acid resistance. Exposure of strain AB1157 to 6 mg/L nalidixic acid in nutrient broth killed 80% of cells after 4 h. This enriched the proportion of nalidixic acid-resistant cells present in the surviving cell population giving enhanced "apparent" mutation frequencies. These were further increased by cell division of resistant mutants in the nalidixic acid-containing medium. "Apparent" resistance mutation frequencies in nalidixic acid-exposed cultures of the R46-, R391- or pYD1-carrying derivatives were, at their peak, 447-, 53- and 38-fold higher than in the control, the strain without plasmid, or the RP4-containing strain, respectively. These data illustrate how mutator plasmids like R391 and pYD1, which mediate only small increases in spontaneous mutation, can contribute to the development of clinically-significant levels of quinolone resistance.

Epidemiology and susceptibilities to mercury preservatives of staphylococci isolated from used eye-drops preserved with thiomersal.

Minimum inhibitory concentrations (MICs) of seven independent isolates of Staphylococcus hominis isolated in the same week from used eye-drops, preserved with thiomersal and collected from wards and clinics in the same hospital, ranged between 1 and 0.03 mg L-1 for thiomersal, 1 and 0.01 mg L-1 for phenyl mercuric nitrate and 10 and 3 mg L-1 for mercuric chloride. Although MIC values determined on solid nutrient medium indicated a 100-fold variation in susceptibility to the bacteriostatic effect of phenyl mercuric nitrate, after 5 h in an aqueous solution containing the bactericidal concentration of 10 mg L-1 phenyl mercuric nitrate, the survival levels of the six S. hominis isolates were similar, with a mean of 13.4% (s.d. 11.0), compared with 100 and 0.8%, respectively, for the most resistant and most sensitive control staphylococcal strains tested. Antibiotic susceptibilities and plasmid profiles of the S. hominis isolates indicated they were the same strain. It is concluded that laboratory indicators of preservative efficacy, such as MIC determination or susceptibility to bactericidal concentrations of preservatives, do not necessarily correlate with the epidemiology of contaminating bacterial strains or their survival in preserved pharmaceuticals.

Teicoplanin or vancomycin in the treatment of gram-positive infections?

The glycopeptide antibiotics vancomycin and teicoplanin have similar mechanisms of action on bacterial cell wall synthesis. Their spectra of activity are limited to Gram-positive bacteria, with the degree of bactericidal activity depending on the species of micro-organism. Staphylococcus aureus, Staphylococcus epidermis, enterococci and Clostridium difficile are generally sensitive, including methicillin-resistant strains of S. aureus and S. epidermidis. Glycopeptide resistance has recently emerged in staphylococci and enterococci. Vancomycin has a shorter half-life than teicoplanin and requires multiple dosing to maintain adequate serum levels. It can only be given by prolonged intravenous infusion over 1 h. In contrast, the pharmacokinetics of teicoplanin allow for once-daily dosing, either by rapid intravenous infusion or by the intramuscular route. The latter offers reliable absorption for patients with limited venous access and is also of benefit for out-patient therapy. Teicoplanin is a safer drug than vancomycin. It is associated with a lower incidence of nephrotoxicity or ototoxicity. Compared to vancomycin, the availability of the intramuscular route and the absence of a requirement for routine serum monitoring, together with the reduced need to treat drug-related side-effects make teicoplanin more cost-effective. It is as effective as vancomycin for most indications, is safe, easy to administer and an important agent for treating Gram-positive infections. Its role in hospitals is likely to increase if the price of drug acquisition is kept low.

Post-antibiotic effects of cefdinir on Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus and Streptococcus pyrogenes.

The post-antibiotic effects (PAEs) of a new cephalosporin, cefdinir, were determined against a range of organisms using a viable counting technique. Cefdinir exerted considerable PAEs against Staphylococcus aureus and Streptococcus pyogenes, but no overall post-antibiotic inhibition of growth was detected against Escherichia coli or Klebsiella pneumoniae. Exposure to cefdinir made the gram-negative organisms susceptible to the washing procedure used for drug removal, but this was followed by rapid recovery of viability in drug-free broth.

Antagonism between bactericidal activities of 4-quinolones and coumarins gives insight into 4-quinolone killing mechanisms.

At concentrations exceeding their MICs, novobiocin and coumermycin antagonised the bactericidal activities of nalidixic acid, ciprofloxacin, ofloxacin and norfloxacin against Escherichia coli KL16. The sensitivities to killing by ciprofloxacin of four mutant derivatives of KL16 carrying gyrA, nalB, nal24 or nal31 alleles were also antagonised by novobiocin. The activities of drug combinations were tested in nutrient broth, which allowed expression of 4-quinolone killing mechanisms A, B and C. They were also tested in nutrient broth plus rifampicin to inhibit mechanisms A and C of the 4-quinolones, and in phosphate-buffered saline, which inhibited mechanism A. Results showed that novobiocin antagonised mechanism C, but not B, of both ciprofloxacin and ofloxacin, but did not antagonise mechanism C of norfloxacin. A review of these and other data indicates that mechanism B may result from the activities of SOS error-prone DNA repair on an irreversibly-bound drug-gyrase-DNA complex, and that mechanism C is mediated via drug interaction with the B subunit of DNA gyrase.

4-Quinolone bactericidal mechanisms.

The bactericidal activity of nalidixic acid against Escherichia coli strain KL16 in nutrient broth was abolished by the addition of rifampicin. Cells suspended in phosphate-buffered normal saline (PBS) were also not killed by nalidixic acid. Experiments with modern 4-quinolones showed their activities varied according to the conditions under which they were tested. Rifampicin did not affect the concentration at which ofloxacin became bactericidal in nutrient broth, but did limit the extent of ofloxacin-induced death. However, rifampicin produced a 10-fold increase in the concentration at which ciprofloxacin became bactericidal in nutrient broth, and completely abolished the bactericidal activity of norfloxacin. Unlike nalidixic acid all of the modern 4-quinolones killed cells suspended in PBS. Based on these results it was possible to differentiate 3 processes by which 4-quinolones induced death. Mechanism A was only active against dividing bacteria and required RNA and protein synthesis; it was therefore not active against bacteria suspended in PBS and was inhibited in nutrient broth by the addition of rifampicin. Mechanism B required neither RNA nor protein synthesis and was also active against non-dividing bacteria; it was therefore not inhibited by rifampicin nor by suspending bacteria in PBS. Mechanism C killed non-dividing bacteria, but required protein and RNA synthesis: it therefore functioned in PBS, but was inhibited by rifampicin.(ABSTRACT TRUNCATED AT 250 WORDS)

Post-antibiotic effects of ofloxacin on Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus, and Streptococcus pyogenes.

A viable counting technique was used to determine the post-antibiotic effect (PAE) of ofloxacin against four bacterial species, treated with either once of four times the minimum inhibitory concentration for 1 or 3 h. Similar to the results obtained previously with ciprofloxacin, ofloxacin gave PAE values with Escherichia coli, Staphylococcus aureus, and Streptococcus pyogenes. Cell division of Klebsiella pneumoniae was inhibited on removal of ofloxacin, but no clear PAE was demonstrated with this species because once replication recommenced, the mean generation times of drug-treated cultures were much shorter than those of untreated controls. Therefore, although the results obtained with ciprofloxacin and ofloxacin imply a consistency of PAE for 4-quinolones within a species, the response to DNA damage induced by 4-quinolones is multifaceted and species dependent. 4-quinolones inhibit both DNA replication and cell division, whilst at the same time stimulating DNA repair pathways. Thus, in some cases PAEs result from an increased post-treatment lag phase which may be followed by nearly normal multiplication, whereas in other cases a long lag may be followed by abnormally rapid cell division, with the generation times of treated cultures being shorter than those of corresponding drug-free controls. The PAE of a drug-induced lag may thus be masked by rapid cell division once growth resumes.

Function of the SOS process in repair of DNA damage induced by modern 4-quinolones.

The recA13 mutant of Escherichia coli strain K-12, which lacks recombination and SOS error-prone DNA repair is hypersensitive to nalidixic acid and to the newer 4-quinolones ciprofloxacin, norfloxacin and ofloxacin. However, whereas recombination-proficient but SOS repair-deficient strains, such as those carrying the lexA3 or recA430 alleles are no more sensitive to nalidixic than the lexA+ recA+ parent, they are more sensitive to the newer quinolones, although not as sensitive as the recA13 derivative. Nalidixic acid possesses only bactericidal mechanism A (which requires RNA and protein synthesis and is only effective on actively dividing cells), whereas the newer 4-quinolones exhibit additional mechanisms B (which does not require RNA and protein synthesis and is effective on bacteria unable to multiply) and C (which requires RNA and protein synthesis but does not depend on cell division). Results obtained with bacteria suspended in phosphate-buffered saline, which inhibits mechanism A, and with bacteria suspended in nutrient broth plus rifampicin, which inhibits mechanisms A and C, showed that the lexA3 mutant was still more sensitive than the lexA+ parent under these conditions. The results suggest that, unlike bactericidal mechanism A, DNA damage that results from bactericidal mechanisms B and C of the newer 4-quinolones is subject to SOS error-prone (mutagenic) repair.

Mutator plasmid in a nalidixic acid-resistant strain of Shigella dysenteriae type 1.

A clinical isolate of Shigella dysenteriae from Kashmir, resistant to seven antibacterial agents including nalidixic acid, carried four plasmids, only one of which was transferable by conjugation. This plasmid, designated pYD1, conferred trimethoprim resistance and increased the frequency of mutation to nalidixic acid resistance in recipient strains. Thus, although nalidixic acid resistance was not carried on a transferable plasmid, the presence of pYD1 increased the frequency at which the strain mutated to nalidixic acid resistance.

Amphotericin B and its delivery by liposomal and lipid formulations.

In recent years, new formulations of the original amphotericin B preparation (Fungizone) have been devised in order to overcome toxicity problems that frequently occur. These preparations represent an improved method of drug delivery, with an increased therapeutic index and a decrease in toxicity to mammalian cell membranes. The new formulations have different physico-chemical characteristics and differ in pharmacokinetic parameters. Their effects must be compared with conventional amphotericin B to ascertain potential roles in future antifungal therapy.

Post-UV survival and mutagenesis in DNA repair-proficient and -deficient strains of Escherichia coli K-12 grown in 5-azacytidine to inhibit DNA cytosine methylation: evidence for mutagenic excision repair.

Inhibition of cytosine methylation by growth in 5-azacytidine (5-azaC), did not affect the sensitivities to DNA damage induced by exposure to ultraviolet light (UV) of Escherichia coli K-12 strains AB1157 dcm+, which is fully DNA repair-proficient, LR68 (a dcm derivative of AB1157), JC3890 dcm+ uvrB, deficient in error-free excision repair, TK702 dcm+ umuC, deficient in error-prone repair, or TK501 dcm+ uvrB umuC, which lacks both excision repair and error-prone repair. However, growth in 5-azaC increased the post-UV survival of strains AB2463 recA(Def), AB2470 recB and AB2494 lexA(Ind-), which are deficient in the induction or expression of recombination repair or error-prone repair of DNA. Spontaneous mutation frequencies were increased in strains LR68, AB2463, AB2470 and AB2494 by growth in 5-azaC, but remained unaltered in strains AB1157, JC3890, TK702 or TK501. Growth in 5-azaC significantly increased UV-induced mutation frequencies in strains AB2463 and AB2470, significantly reduced UV-induced mutation in strain JC3890, but had little effect on UV-induced mutation in the other strains. The results suggest that 5-azaC may induce a normally error-free DNA repair pathway to become error-prone and therefore genotoxic.

Studies on mutational cross-resistance between ciprofloxacin, novobiocin and coumermycin in Escherichia coli and Staphylococcus warneri.

Nalidixic acid resistant mutants of Escherichia coli KL16 were tested against ciprofloxacin, coumermycin and novobiocin. The mutants gyrA, nalB and nal-24 were more resistant than KL16 to ciprofloxacin, whereas the nal-31 strain was hypersensitive. Only the nalB mutant was more resistant to novobiocin than KL16, but gyrA, nal-31 and nal-24 mutants were more sensitive to coumermycin than KL16. Newly-isolated novobiocin-resistant mutants of KL16 were not cross-resistant to coumermycin or ciprofloxacin. Some coumermycin-resistant mutants were cross-resistant to novobiocin but not ciprofloxacin, whereas mutants resistant to novobiocin and ciprofloxacin were isolated at higher coumermycin concentrations. Two types of Staphylococcus warneri mutant were isolated on media containing novobiocin or coumermycin. Each was resistant to both coumarins, but one was highly resistant to novobiocin and the other to coumermycin. High level resistance to both coumarins was unstable. E. coli mutants differed in susceptibility to bactericidal concentrations of ciprofloxacin, and S. warneri mutants behaved similarly. These results suggest the modes of action of the coumarins are not identical.

Contributions of post-antibiotic lag and repair-recovery to the post-antibiotic effects of ciprofloxacin on Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus and Streptococcus pyogenes.

A viable counting technique was used to determine the post-antibiotic effect (PAE) of ciprofloxacin against four bacterial species, treated with either one or four times the minimum inhibitory concentration for 1 or 3 h. PAE were demonstrated with Escherichia coli, Staphylococcus aureus or Streptococcus pyogenes after exposure to either concentration for both times. No clear PAE was demonstrated for Klebsiella pneumoniae after any treatment. In some cases, PAE was due to an increased post-treatment lag phase, which was followed by nearly normal multiplication, whereas in other cases a long lag was followed by abnormally rapid cell division, with the generation times of treated cultures being much less than those of the corresponding drug-free controls. This is evidence of recovery of cells that have completed DNA repair. S. aureus, E. coli and K. pneumoniae all exhibited evidence of this type of repair even though K. pneumoniae gave no significant PAE. However, the post-treatment generation times of S. pyogenes, which produced the greatest PAE, gave no evidence of such repair. It is concluded that PAEs may result from a variety of factors.

Investigation of potential genotoxic effects of low frequency electromagnetic fields on Escherichia coli.

Exposure of growing cells of Escherichia coli strain AB1157 to a frequency of 1 Hz with field strengths of 1 or 3 kV m-1 did not affect spontaneous or ultraviolet light (UV)-induced mutation frequencies to rifampicin resistance. Neither did growth in the presence of charge alter the sensitivities of strains AB1157, TK702 umuC or TK501 umuC uvrB to UV. Similarly, although the resistance of strains TK702 umuC and TK501 umuC uvrB to UV was increased by the presence of plasmid pKM101, which carries DNA repair genes, pregrowth of plasmid-containing strains in electric fields did not increase UV resistance. Finally, growth in a low frequency field in the presence of sub-inhibitory concentrations of mitomycin C did not affect mitomycin C-induced mutation frequencies. It is concluded that low frequency electromagnetic fields do not increase spontaneous mutation, induce DNA repair or increase the mutagenic effects of UV or mitomycin C.

Increase in susceptibility to EcoRII restriction of bacteriophage lambda produced by propagation on host cells growing in 5-azacytidine: a new in-vivo method for demonstration of DNA-methylation inhibition.

The efficiency of plating on EcoRII-restricting cells of bacteriophage lambda vir propagated on an Escherichia coli K-12 dcm+ host decreased with increase in concentration of 5-azacytidine (5-azaC) in the propagating medium. This illustrates, in-vivo, the inhibition of DNA-cytosine methylation induced by 5-azaC and provides a simple system for the detection of DNA-methylation inhibitors.

Methicillin-resistant Staphylococcus aureus--an overview.

Methicillin-resistant strains of Staphylococcus aureus (MRSA) cause life-threatening infections, but they are no more pathogenic than methicillin-sensitive strains. Difficulties occur because of incorrect or missed identification of MRSA, and hence inappropriate or ineffective treatment of infections. Therapeutic options are severely limited and the increasing spectrum of resistance in MRSA is worrying. However, new methods of detection and new agents for treatment are being developed in response to the challenge of MRSA. Whilst the organism is a problem and control measures are necessary to contain its spread, the outlook is not bleak. In the medium term, the development of new, effective anti-MRSA agents should prevent the threat of MRSA becoming any greater in the field of hospital infection control.