The Role of Reactive Oxygen Species in Antibiotic-Induced Cell Death in Burkholderia cepacia Complex Bacteria

Abstract

It was recently proposed that bactericidal antibiotics, besides through specific drug-target interactions, kill bacteria by a common mechanism involving the production of reactive oxygen species (ROS). However, this mechanism involving the production of hydroxyl radicals has become the subject of a lot of debate. Since the contribution of ROS to antibiotic mediated killing most likely depends on the conditions, differences in experimental procedures are expected to be at the basis of the conflicting results. In the present study different methods (ROS specific stainings, gene-expression analyses, electron paramagnetic resonance, genetic and phenotypic experiments, detection of protein carbonylation and DNA oxidation) to measure the production of ROS upon antibiotic treatment in Burkholderia cepacia complex (Bcc) bacteria were compared. Different classes of antibiotics (tobramycin, ciprofloxacin, meropenem) were included, and both planktonic and biofilm cultures were studied. Our results indicate that some of the methods investigated were not sensitive enough to measure antibiotic induced production of ROS, including the spectrophotometric detection of protein carbonylation. Secondly, other methods were found to be useful only in specific conditions. For example, an increase in the expression of OxyR was measured in Burkholderia cenocepacia K56-2 after treatment with ciprofloxacin or meropenem (both in biofilms and planktonic cultures) but not after treatment with tobramycin. In addition results vary with the experimental conditions and the species tested. Nevertheless our data strongly suggest that ROS contribute to antibiotic mediated killing in Bcc species and that enhancing ROS production or interfering with the protection against ROS may form a novel strategy to improve antibiotic treatment.

Fig 1. Influence of pH on fluorescence of H2DCFDA and HPF.

Error bars represent SEM. Solutions for which fluorescence was significantly different from fluorescence at pH 7 are indicated with an asterisk, p < 0.05, n = 3.

Fig 2. Fluorescence generated over time in treated (4 x MIC, up to 13 h) (red circles) and untreated (blue squares) biofilms (left) and planktonic cultures (right).

(A) Cultures incubated with (closed circles and squares) or without (open circles and squares) H2DCFDA. (B) Cultures incubated with (closed circles and squares) or without (open circles and squares) HPF. Data are shown of a single representative experiment. Error bars represent SEM (calculated on 3 technical replicates).

Fig 3. Luminescence generated over time after treating a B. cenocepacia K56-2 oxyR:lux promoter fusion mutant biofilm with 0.03% H₂O₂, Cip (4 x MIC), Mer (4 x MIC) or Tob (2 x MIC) (up to 6 h) compared to luminescence in biofilms exposed to LB alone (blue).

Data are shown of a single representative experiment.

Fig 4. EPR spectrum of the spin probe CMH and superoxide radicals formed in planktonic cultures of B. cenocepacia K56-2 treated with Tob (4 x MIC) for 30 min.

Data are shown of a single representative experiment.

Fig 5. EPR determination of radicals formed in treated and untreated (blue) planktonic cultures of B. cenocepacia K56-2.

Cultures were treated with Tob (4 x MIC, 30 min, red), Cip (4 x MIC, 30 min and 180 min, orange) or Mer (4 x MIC, 30 min and 180 min, green). The probe was added 30 min before measurement. Error bars represent SEM. Statistically significant differences are indicated with an asterisk, p < 0.05, n = 4.

Fig 6. The fraction of surviving cells after treating biofilms (top) or planktonic cultures (bottom) with Tob or Cip (4 x MIC, 24 h) in combination with an antioxidant, compared to the fraction surviving cells after treatment with Tob or Cip alone.

Error bars represent SEM. Statistically significant differences are indicated with an asterisk, p < 0.05, n = 3.

Fig 7. Correlation between fold change in survival and fold change in fluorescence after treatment with Tob or Cip (4 x MIC, 24 h) in combination with an antioxidant compared to treatment with Tob or Cip alone.

Fig 8. Summary of import aspects of ROS detection methods.

Fig 9. Fluorescence generated over time after treatment with Tob (4 x MIC, up to 20 h) (red), Cip (16 x MIC, up to 20 h) (orange) or Mer (16 x MIC, up to 20 h) (green) in biofilms and planktonic B. cenocepacia K56-2 cultures compared to fluorescence in pH-matched controls (blue).

Cultures were pre-incubated with H2DCFDA and treated with antibiotics or control solutions with the same pH. Data are shown of a single representative experiment. Error bars represent SEM (calculated on 3 technical replicates).

Fig 10. Fluorescence generated over time in biofilms and planktonic B. cenocepacia K56-2 cultures after treatment with Tob (red) in different concentrations (4xMIC, MIC, MIC/4, up to 8 h) compared to fluorescence in pH-matched control (blue).

Cultures were pre-incubated with H2DCFDA and treated with antibiotics or control solutions with the same pH. Data are shown of a single representative experiment. Error bars represent SEM (calculated on 3 technical replicates).

Fig 11. H2DCFDA fluorescence in treated (Tob 4 x MIC, Cip 16 x MIC, Mer 16 x MIC, 24 h) versus untreated planktonic cultures of B. multivorans LMG 13010, B. vietnamiensis LMG 10929, B. cepacia LMG 1222, B. metallica LMG 24068 and B. cenocepacia K56-2.

Error bars represent SEM. Statistically significant differences are indicated with an asterisk, p < 0.05, n ≥ 3.