Abstract

Objectives: To determine the mutant prevention concentrations (MPCs) of ciprofloxacin and enrofloxacin against four strains of Salmonella enterica serovar Enteritidis and four strains of S. Typhimurium including one fully susceptible, one multiply resistant (MAR), one GyrA mutant and one GyrA/MAR mutant. Further, to examine mutants arising after exposure to sub-MPC concentrations of the antibiotics for susceptibility to ciprofloxacin and enrofloxacin, and cyclohexane tolerance.

Methods: MICs were determined using the agar dilution method of the BSAC. The MPC was recorded as the lowest concentration of antibiotic to inhibit growth from an inoculum of 1010 cfu.

Results: The MPCs and resulting MPC/MIC ratios of enrofloxacin were generally two- to four-fold higher than for ciprofloxacin. At 24 h for both antibiotics, MPCs were lowest for the fully susceptible strains (0.25–0.5 mg/L), similar for the MAR (1–4 mg/L) and GyrA (2–4 mg/L) mutants and highest for the GyrA/MAR mutants (1–8 mg/L). MPC/MIC ratios at 24 h were 2–16 for all strains except those for the MAR strains without mutation in gyrA where the ratios were 8–64.

Conclusions: The ability to eradicate Salmonellain vivo depends on many factors such as antibiotic susceptibility of the strain, dose and route of administration. It is suggested that these MPC values will be useful when considering dosing strategies. In view of the high MPC/MIC ratio, MAR strains with wild-type gyrA, although susceptible to ciprofloxacin (MICs 0.06–0.13 mg/L), may give rise to treatment failures.

Introduction

Fluoroquinolones are the drug of choice for the treatment of invasive Salmonella enterica infections in humans and are also used for animal treatment.1 Of increasing concern is the rise in numbers of strains resistant to ciprofloxacin; from 20% to 26% for S. Typhi and from 6% to 17% for S. Paratyphi from 1999 to 2001.2 Whilst a high proportion of fluoroquinolone-resistant strains arise from individuals returning to the West from travel in the developing world,2,3 antimicrobial resistance in zoonotic microorganisms has been attributed to the use of antimicrobials in food-producing animals and widespread use in humans worldwide.24 Recent reports state that fluoroquinolones were ineffective at resolving infections in humans when the MICs of fluoroquinolones were 0.06–2 mg/L.1 Furthermore, decreased susceptibility to fluoroquinolones can develop during therapy.1

Mutations in gyrA, the target enzyme of quinolones, is the primary mechanism by which S. enterica become fluoroquinolone-resistant.5 Since spontaneous mutants usually arise at a frequency associated with mutation in a single gene (10−7 to 10−8), it is argued that preventing the selective enrichment of mutant bacterial populations may help restrict the development of antibiotic resistance.6 The mutant prevention concentration (MPC) has been defined as the lowest concentration of antibiotic to inhibit the emergence of mutants from 1010 cfu.6 The aim of this study was to determine MPCs of ciprofloxacin and enrofloxacin in vitro against fully susceptible strains of S. Enteritidis and Typhimurium, strains with mutation in gyrA, multiply antibiotic-resistant (MAR) strains and strains with a mutation in gyrA which were also MAR. Additionally, the MICs of ciprofloxacin and enrofloxacin and susceptibility to cyclohexane, a marker for MAR,8 were determined for any mutants which arose at concentrations between the MIC and MPC.

Materials and methods

Bacterial strains

Isolates were obtained from the Veterinary Laboratories Agency (VLA), Weybridge, UK (Table 1).

Determination of MICs, MPCs, mutation rate and cyclohexane resistance

MICs were determined by the agar dilution method as described by the BSAC.7 Cyclohexane resistance was determined as previously described.8 To determine MPCs, strains were first grown overnight in 100 mL LB broth at 37°C, centrifuged and concentrated by resuspending in 10 mL PBS before determining viable counts. Strains in PBS (100 μL of each strain containing ≥109 cfu) were then plated onto each of 10 plates supplemented with dilutions of ciprofloxacin or enrofloxacin (1 × to 128 × MIC). Plates were incubated at 37°C for 24 h, the colonies counted and incubated again for a further 24 h. The MPC was recorded as the lowest concentration of antibiotic to prevent the emergence of any mutants after 24 h and 48 h incubation.6

At antibiotic concentrations between the MIC and MPC, the frequency at which resistant mutants were selected was calculated as the number of mutants growing in the presence of antibiotic per mL of inoculum divided by the cfu/mL of inoculum. At each antibiotic concentration, five mutants were selected at random and stored on Dorset egg slopes until required for further analysis. The MPC/MIC ratio was determined by dividing the MPC values by the MIC values.

Statistics

The non-parametric Kruskal–Wallis test was used to determine whether mutants isolated with ciprofloxacin were more resistant to ciprofloxacin and enrofloxacin than strains isolated with enrofloxacin.

Results

MICs, MPCs, MPC/MIC ratios and mutant selection frequencies

Although the MICs of ciprofloxacin and enrofloxacin were similar, the MPCs and resulting MPC/MIC ratios tended to be two-to four-fold higher for enrofloxacin compared with ciprofloxacin. MAR strains without mutations in gyrA had MPC/MIC ratios at 24 h of 8–64, whereas all other strains had MPC/MIC ratios at 24 h in the range 2–16 (Table 1). In particular, the MPC/MIC ratios (24 h MPC result) for the S. Typhimurium MAR mutant 3992/96 (no mutation in gyrA) were 32 and 64 for ciprofloxacin and enrofloxacin, respectively (Table 1). The frequency of selection of resistance was similar for Salmonella exposed to either ciprofloxacin or enrofloxacin and ranged from ∼10−7 to ∼10−11 at 24 h (data not shown). The frequency at which mutants were selected decreased as the antibiotic concentration approached the MPC. In most cases, more mutant colonies were seen when the plates were incubated for a further 24 h and for S. Enteritidis 8497/99 and 248/94 and S. Typhimurium 3992/96 and 3992/96 MAR, this had the effect of increasing the MPC by two- to four-fold at 48 h compared to the 24 h value (data not shown). As such, the frequency at which mutants were selected at 48 h was higher (∼1.5- to 10-fold) than that observed at 24 h (data not shown).

Phenotype of selected mutants

Mutants selected with ciprofloxacin were approximately two-fold more resistant to both ciprofloxacin and enrofloxacin compared with mutants isolated with enrofloxacin; this difference was statistically significant (P=0.011) for ciprofloxacin but not for enrofloxacin (P=0.227) (Table 2). None of the mutants derived from the fully susceptible strains became cyclohexane tolerant, but most of the mutants (10/15) derived from the cyclohexane susceptible strains with mutations in gyrA became cyclohexane tolerant (Table 2).

Discussion

In recent years, the concept of the mutant prevention concentration (MPC), particularly for concentration-dependent bactericidal antibiotics such as the fluoroquinolones, has been considered important in preventing the emergence of antibiotic resistance.6 If an antibiotic concentration is maintained above the MPC, resistant bacteria should not be selected.6 MPCs of various quinolones have been determined for a number of bacteria and MPC/MIC ratios for ciprofloxacin have been reported as being 24 for Mycobacterium smegmatis, 53 for M. tuberculosis and 13 for Staphylococcus aureus.6 In Escherichia coli, a norfloxacin MPC/MIC ratio of 36 has been reported,6 whereas other data report MPC/MIC ratios of 4 for ciprofloxacin and norfloxacin for E. coli.9 We are aware of no data for Salmonella with which to compare our data. Although the MICs of ciprofloxacin and enrofloxacin were identical for all but one strain, for which there was a two-fold difference, the MPCs and therefore the MPC/MIC ratios tended to be two- to four-fold higher for enrofloxacin compared with ciprofloxacin. Of interest, ciprofloxacin appeared to select for mutants that were more resistant to ciprofloxacin and enrofloxacin than those selected with enrofloxacin.

Whilst it is not always possible to extrapolate from an in vitro to an in vivo situation, MICs determined in vitro are extrapolated to the in vivo situation. Extrapolating from the in vitro MPC data indicate that it should be possible to eradicate fully susceptible strains of Salmonellain vivo, bearing in mind the MPC values of 0.25 and 0.5 mg/L for ciprofloxacin and enrofloxacin, respectively, although it is possible that mutants could be selected when antibiotic levels in vivo fall below the MPC. However, eradication of strains in vivo that already have some resistance, such as MAR or gyrA, will be more difficult, in view of their higher MPC values and the difficulty that may be associated with attaining higher levels of fluoroquinolones in vivo. One of the MAR mutants (S. Typhimurium 3992/96) had very high MPC/MIC ratios (32 and 64 for ciprofloxacin and enrofloxacin, respectively at 24 h) and as such, the low-level resistance of this organism (ciprofloxacin and enrofloxacin MICs of 0.06 mg/L) may be of greater clinical significance than previously considered. Future work will investigate if other MAR mutants have high MPC/MIC ratios. It is possible that the MAR phenotype affords some initial protection against supra-MIC concentrations of quinolones, allowing for selection of strains with higher level resistance conferred by mutations in gyrA, and this may result in the high MPC/MIC ratios seen for one of the MAR mutants that originally did not have mutations in gyrA. Further work would be needed to validate this. However, the efficacy of in vivo treatment depends on many factors such as dose, route of administration, target species (human or animal) and the location of infection. A recent study in pigs showed that increasing the enrofloxacin dose from 2.5 mg/kg to 7.5 and 15 mg/kg body weight prevented colonization with an experimentally introduced nalidixic acid-resistant strain of S. Typhimurium with a ciprofloxacin MIC of 0.5 mg/L.10 However, this regimen did not prevent colonization with a ciprofloxacin-resistant E. coli.10 Current thinking is that shorter courses of fluoroquinolone treatment at higher doses may be one way to reduce the development of resistance in bacteria. Studies in humans have shown that 3 days of ofloxacin at 15 mg/kg is as effective as 5 days treatment at 10 mg/kg against multidrug-resistant typhoid caused by strains such as S. Typhi, Paratyphi and Choleraesuis.11 However, further work is needed to ascertain how effective such treatment may be in humans and various animal species infected with different resistant phenotypes and genotypes of Salmonella and the effect higher treatment levels may have on resident gut flora.

Studying the MPCs of different antibiotics for different organisms in vitro will give information that will be useful in predicting the development of resistance or clearance of organisms in clinical situations. However, it is accepted that only a small number of strains were examined in this preliminary study and future work is proposed.

Table 1.

MICs and MPCsa of ciprofloxacin and enrofloxacin for S. enterica

    MICs (mg/L)
 
 MPCs (mg/L)
 
 MPC/MIC ratio
 
 
Serotype Strain reference number Mutations in gyrAb cyxc CIP ENR CIP ENR CIP ENR 
Enteritidis LA5 WT 0.03 0.03 0.25 0.5 16 
 LA5 MARd WT 0.13 0.13 
 8497/99 Ser83-Phe 0.5 0.5 
 248/94 Asp87-Asn 0.5 
Typhimurium 3992/96 WT 0.03 0.03 0.25 0.5 16 
 3992/96 MARd WT 0.06 0.06 32 64 
 5120/98 Ser83-Tyr 0.5 0.5 
 6785/98 Ser83-Phe 
    MICs (mg/L)
 
 MPCs (mg/L)
 
 MPC/MIC ratio
 
 
Serotype Strain reference number Mutations in gyrAb cyxc CIP ENR CIP ENR CIP ENR 
Enteritidis LA5 WT 0.03 0.03 0.25 0.5 16 
 LA5 MARd WT 0.13 0.13 
 8497/99 Ser83-Phe 0.5 0.5 
 248/94 Asp87-Asn 0.5 
Typhimurium 3992/96 WT 0.03 0.03 0.25 0.5 16 
 3992/96 MARd WT 0.06 0.06 32 64 
 5120/98 Ser83-Tyr 0.5 0.5 
 6785/98 Ser83-Phe 

CIP, ciprofloxacin; ENR, enrofloxacin.

a

The mutant prevention concentration (MPC) is the concentration of antibiotic to prevent any mutants from 1010 cfu and results here are at 24 h.

b

As previously described,12 WT, wild-type.

c

cyx, cyclohexane; S, susceptible; T, tolerant as previously described.8

d

Laboratory multiple antibiotic-resistant mutant (MAR) derived from plates with 5 mg/L tetracycline.

Table 2.

MICs (mg/L) of ciprofloxacin and enrofloxacin against mutants generated

 Mutants (no. tested) selected from growth with
 
    
 ciprofloxacin
 
 enrofloxacin
 
  
Resistance type of parent strainsa CIP MIC100 ENR MIC100 CIP MIC100 ENR MIC100 No. of mutants cyxT 
gyrA– and cyxS 0.5 (4) 0.5 (4) 0.5 (8) 0.5 (8) 0/12 
gyrA–and cyxT 4 (12) 4 (12) 2 (14) 4 (14) NA 
gyrA+and cyxS 8 (8) 4 (8) 4 (7) 4 (7) 10/15 
gyrA+and cyxT >8 (8) 8 (8) 2 (5) 8 (5) NA 
GMMb 2.39 (32) 2.16 (32) 1.14 (34) 1.47 (34)  
 Mutants (no. tested) selected from growth with
 
    
 ciprofloxacin
 
 enrofloxacin
 
  
Resistance type of parent strainsa CIP MIC100 ENR MIC100 CIP MIC100 ENR MIC100 No. of mutants cyxT 
gyrA– and cyxS 0.5 (4) 0.5 (4) 0.5 (8) 0.5 (8) 0/12 
gyrA–and cyxT 4 (12) 4 (12) 2 (14) 4 (14) NA 
gyrA+and cyxS 8 (8) 4 (8) 4 (7) 4 (7) 10/15 
gyrA+and cyxT >8 (8) 8 (8) 2 (5) 8 (5) NA 
GMMb 2.39 (32) 2.16 (32) 1.14 (34) 1.47 (34)  

CIP, ciprofloxacin; ENR, enrofloxacin; MIC100, concentration of antibiotic to inhibit all strains tested; NA, not applicable.

a

gyrA−/gyrA + , no mutation/mutation in gyrA; cyxS/cyxT, cyclohexane susceptible/tolerant.

b

GMM, geometric mean of MICs of antibiotics.

This work was supported by DEFRA UK, Grant reference number VM2201 to L. P. Randall.

References

1.
Aarestrup, F. M., Wiuff, C., Molbak, K. et al. (
2003
). Is it time to change fluoroquinolone breakpoints for Salmonella spp.?
Antimicrobial Agents and Chemotherapy
 
47
,
827
–9.
2.
Threlfall, E. J., Fisher, I. S., Berghold, C. et al. (
2003
). Trends in antimicrobial drug resistance in Salmonella enterica serotypes Typhi and Paratyphi A isolated in Europe, 1999-2001.
International Journal of Antimicrobial Agents
 
22
,
487
–91.
3.
Hakanen, A., Siitonen, A., Kotilainen, P. et al. (
1999
). Increasing fluoroquinolone resistance in Salmonella serotypes in Finland during 1995-1997.
Journal of Antimicrobial Chemotherapy
 
43
,
145
–8.
4.
Malorny, B., Schroeter, A. & Helmuth, R. (
1999
). Incidence of quinolone resistance over the period 1986 to 1998 in veterinary Salmonella isolates from Germany.
Antimicrobial Agents and Chemotherapy
 
43
,
2278
–82.
5.
Piddock, L. J. V. (
1995
). Mechanisms of resistance to fluoroquinolones: state-of-the-art 1992-1994.
Drugs
 
49
,
29
–35.
6.
Drlica, K. (
2003
). The mutant selection window and antimicrobial resistance.
Journal of Antimicrobial Chemotherapy
 
52
,
11
–17.
7.
Andrews, J. M. (
2001
). Determination of minimum inhibitory concentrations.
Journal of Antimicrobial Chemotherapy
 
48
, Suppl. S1,
5
–16.
8.
Randall, L. P., Cooles, S. W., Sayers, A. R. et al. (
2001
). Cyclohexane resistance in Salmonella of different serovars is associated with increased resistance to multiple antibiotics, disinfectants and dyes.
Journal of Medical Microbiology
 
50
,
1
–6.
9.
Linde, H. & Lehn, N. (
2004
). Mutant prevention concentration of nalidixic acid, ciprofloxacin, clinafloxacin, levofloxacin, norfloxacin, ofloxacin, sparfloxacin or trovafloxacin for Escherichia coli under different growth conditions.
Journal of Antimicrobial Chemotherapy
 
53
,
252
–7.
10.
Wiuff, C., Lykkesfeldt, J., Svendsen, O. et al. (
2003
). The effects of oral and intramuscular administration and dose escalation of enrofloxacin on the selection of quinolone resistance among Salmonella and coliforms in pigs.
Research in Veterinary Science
 
75
,
185
–93.
11.
Hien, T. T., Bethell, D., Hoa, N. T. T. et al. (
1995
). Short courses of ofloxacin for treatment of multidrug-resistant typhoid.
Clinical Infectious Diseases
 
20
,
917
–23.
12.
Liebana, E., Clouting, C., Cassar, C. A. et al. (
2002
). Comparison of gyrA mutations, cyclohexane resistance, and the presence of class I integrons in Salmonella enterica from farm animals in England and Wales.
Journal of Clinical Microbiology
 
40
,
1481
–6.

Author notes

1Veterinary Laboratories Agency (Weybridge), New Haw, Addlestone, Surrey KT15 3NB; 2Antimicrobial Agents Research Group, Division of Immunity and Infection, The Medical School, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK