Antibiotic resistance emergence in a bacterial population is produced by several factors, of which rate of mutation to resistance and fitness cost of mutants have been considered critical . A consistent fitness cost may appear when chromosomal mutations cause structural modifications in the cellular target of a drug and when efflux pumps are overexpressed [2,3]. In the case of fluoroquinolones, resistance is mainly acquired by increasing efflux pump and DNA gyrase and topoisomerase mutations, . This resistance has been widely shown to slow the growth and the virulence of Escherichia coli and Salmonella enterica [5,6]. Contradictory data are observed in E. coli resistant to ciprofloxacin due to parC and gyrA mutations, which did not decrease the growth rate in vitro and in vivo . In the same line, the relative growth rate of a ciprofloxacin-resistant S. pneumoniae strain due to parC mutation was identical to its susceptible isogenic parental strain .
Acinetobacter baumannii has become one of the leading nosocomial pathogens worldwide. Resistance to fluoroquinolones has increased in A. baumannii . In the UK, 49.1% of 226 A. baumannii isolates were resistant to ciprofloxacin and 68.5% of these resistant isolates had an MIC of ≥32 mg/L and mutations in gyrA and parC . Fitness costs in A. baumannii remain largely unknown. A. baumannii and Acinetobacter sp. resistant to colistin and rifampin impaired fitness and increased duplication times compared with the parental strain [11,12]. Besides these works, there are no reports on the fitness cost for A. baumannii resistant to fluoroquinolones.
Expanded details of all materials and methods are given in the Supplemental Data.
Bacterial strains: 77wt is a clinical isolate susceptible to ciprofloxacin (CipS) . 77R is a ciprofloxacin-resistant strain (CipR) obtained by exposing CipS to repeated in vitro subinhibitory concentrations of ciprofloxacin.
In vitro susceptibility testing: Ciprofloxacin MICs for CipS and CipR were determined by microdilution in the presence or absence of Phe-Arg-β-naphthylamide  and reserpine .
Biofilm: Biofilm formation assay was performed as described previously .
Cellular viability: A549 cells were infected with CipS and CipR. A. baumannii cytotoxicity was assessed using the MTT assay , the LIVE/DEAD and the cytotoxicity detection kits.
Sepsis peritoneal infection model: A peritoneal sepsis model was used . Mice were inoculated with different bacterial incoula of CipS and CipR.
Competition experiments: The in vitro and in vivo competition index (CI) was determined from growth cultures of CipS, CipR and a mixed CipS and CipR.
Ciprofloxacin MICs against CipS and CipR were 1 and 32 mg/L, respectively. CipR incubation with efflux pump inhibitors PAβN and reserpine showed a decreased ciprofloxacin MIC four and six-fold, respectively (Table S1). No additional mutations in gyrA and parC between the CipS and the CipR were observed (data not shown). These results imply that ciprofloxacin resistance in CipR is due at least to the overexpression of an active efflux pump. Furthermore, we showed that the CipR produced less biofilm than the CipS (Fig. S1).
Cell survival assessment by MTT assay showed that incubation of A549 cells with CipS reduced cell viability to 66.23 ± 4.01%. In contrast, CipR induced a little reduction in cell viability to 86.53 ± 2.44%. As a positive control, incubation of A549 cells with staurosporine reduced cell viability to 33.12 ± 10.06% (Fig. 1a). A. baumannii cytotoxicity was also assessed using a two-colour fluorescence cell viability assay. We showed that CipS induced more cell death, as indicated by red fluorescence, than CipR (Fig. 1b). Additionaly, lower levels of LDH liberation, a necrotic cell sign, were detected in infected A549 cells with CipR (5.98 ± 1.81%) than with CipS (48.38 ± 3.5%) (Fig. 1c). These results showed that acquisition of ciprofloxacin resistance by an active efflux pump impaired cytotoxicity induced by A. baumannii.
Figure 1. Cell death induced by Acinetobacter baumannii. A549 cells were infected with 108 CFU/mL of A. baumannii CipS and CipR for 24 h. A. baumannii cytotoxicity was assessed by monitoring the mitochondrial reduction activity using the MTT assay (a), illustrated with representative microscopic fields of the two-colour fluorescence cell viability assay (LIVE/DEAD Kit) (b), and determined by following LDH release using the cytotoxicity detection assay (c). The positive control was 10 mg/L staurosporine (Str). Representative results of three independent experiments are shown and data are the means ± SD. p < 0.05: *between A549 cells and inoculated groups, † between inoculated groups.
Download figure to PowerPoint
We detect a significant difference in mortality caused by CipS and CipR, independently of the size of inoculums (Table S2). Kaplan–Meier analysis revealed differences between animal groups receiving the same inoculum (p = 0.002, p = 0.07 and p = 0.04 for approximately 8.3, 7.3 and 6.3 log CFU/mL, respectively) (Fig. S2). These results indicate that acquisition of ciprofloxacin resistance by an active efflux pump interferes with infective capacity and mortality caused by A. baumannii in vivo.
An in vitro competition experiment showed statistical differences between CipS and CipR (CI = 0.02; p = 0.04). Growth rates of CipS and CipR cultures were similar, with the exception of the experiments in which both strains were grown together, showing a lower growth of 1.7 log of the CipR (Fig. 2a). These data were confirmed in vivo. CFUs in spleens from animals infected with CipS or CipR differed by 2.02 log at 24 h, whereas the difference was 2.82 log in the mixed infection group (CI = 0.03; p < 0.001) (Fig. 2b). These data demonstrate that acquisition of ciprofloxacin resistance by an active efflux pump affects the fitness of A. baumannii in vitro and in vivo.
Figure 2. In vitro and in vivo competition growth of A. baumannii. CipS, CipR and mixed inoculum were grown in vitro (a) and in vivo after i.p. inoculation with 8.3 Log CFU/mL in mice (b) for 24 h.
Download figure to PowerPoint
The data presented in our study revealed overexpression of efflux pumps in the CipR strain with ciprofloxacin MIC of 32 mg/L and absence of mutations in gyrA and parC. In another study, two clinical A. baumannii isolates with ciprofloxacin MICs of 8 and >32 mg/L presented a mutation in parC but not in gyrA and an overexpression of an efflux pump adeB . Decreased growth in fluoroquinolone resistance isolates could result from inopportune efflux of nutrients and metabolic intermediates by overexpressed MDR efflux pumps . However, evidence to support this hypothesis is currently lacking.
Our results have also demonstrated that the CipR strain was less virulent in vitro and in vivo. Previously, we reported that CipR A. baumannii induced less cell death than CipS strains . Overexpression of efflux pumps by P. aeruginosa was associated with reduction of the gene expression that encodes the type III secretion system and was proposed to cause a concomitant lack of fitness . Gene expression associated with Acinetobacter sp. virulence was autoinduced by quorum-sensing signal molecules . As efflux pumps have been shown to expel quorum-sensing signals in P. aeruginosa , we suggest that altered quorum-sensing signal homeostasis, resulting from increased efflux pump activities, may have contributed to the decreased virulence of the CipR strain observed in our study. More studies are needed to confirm the relationship between efflux pumps and specific genes involved in A. baumannii virulence.
- Top of page
- Transparency Declaration
- Supporting Information
Figure S1. One-day biofilm formation. Biofilm formation of Acinetobacter baumannii CipS and CipR is shown. Representative results of three independent experiments are shown and data are the means ± S.D. p <0.05: * between MHB and inoculated groups, † between inoculated groups.
Figure S2. Mortality curves. Mice were infected with three different inocula (8.3, 7.3, and 6.3 CFU/mL) of Acinetobacter baumannii 77wt strain (black lines) and its ciprofloxacin resistant mutant 77R strain (gray lines). Mortality curves are analysed by Kaplan-Meier survival analysis.
Table S1. In vitro activity of ciprofloxacin against Acinetobacter baumannii strains in the presence and absence of PAβN and reserpine.
Table S2. Seven-day mortality monitoring in a murine model of peritoneal sepsis. Mice were infected with three different inocula of Acinetobacter baumannii CipS and CipR.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.