Host Cell Oxidative Stress Induces Dormant Staphylococcus aureus Persisters

Abstract

Persisters are transiently nongrowing and antibiotic-tolerant phenotypic variants identified in major human pathogens, including intracellular Staphylococcus aureus. Due to their capacity to regrow once the environmental stress is relieved and to promote resistance, persisters possibly contribute to therapeutic failures. While persistence and its related quiescence have been mostly studied under starvation, little is known within host cell environments. Here, we examined how the level of reactive oxygen species (ROS) in different host cells affects dormancy depth of intracellular S. aureus. Using single-cell approaches, we found that host ROS induce variable dormant states in S. aureus persisters, displaying heterogeneous and increased lag times for resuscitation in liquid medium. Dormant persisters displayed decreased translation and energy metabolism, but remained infectious, exiting from dormancy and resuming growth when reinoculated in low-oxidative-stress cells. In high-oxidative-stress cells, ROS-induced ATP depletion was associated with the formation of visible dark foci similar to those induced by the protein aggregation inducer CCCP (carbonyl cyanide m-chlorophenylhydrazone) and with the recruitment of the DnaK-ClpB chaperone system involved in the clearance of protein aggregates. ATP depletion led to higher fractions of dormant persisters than ROS, due to a counterbalancing effect of ROS-induced translational repression, suggesting a pivotal role of translation in the dormant phenotype. Consistently, protein synthesis inhibition limited dormancy to levels similar to those observed in low-oxidative-stress cells. This study supports the hypothesis that intracellular S. aureus persisters can reach heterogeneous dormancy depths and highlights the link between ROS, ATP depletion, dark focus formation, and subsequent dormancy state. IMPORTANCE By their capacity to survive to antibiotic pressure and to regrow and give rise to a susceptible population once this pressure is relieved, intracellular persisters of S. aureus may contribute to explain therapeutic failures and recurrent infections. Here, we show that the level of dormancy and the subsequent capacity to resuscitate from this resting state are dependent on the level of oxidative stress in the host cells where bacteria survive. This observation nourishes the debate as whether the most appropriate strategy to cope with S. aureus intracellular infections would consist of trying to push persisters to a deep dormancy state from which wakening is improbable or, on the contrary, to prevent ROS-induced dormancy and force bacteria to maintain regular metabolism in order to restore their responsiveness to antibiotics. Importantly also, our data highlight the interest in single-cell analyses with conventional enumeration of CFU to quantify persisters and study host-pathogen interactions.

Keywords: Staphylococcus aureus; dormancy; intracellular infection; macrophages; oxidative stress; persisters.

FIG 1

Host oxidative stress drives transition to dormant states of S. aureus persisters. (A) Bacteria (CFU or total propidium iodide-negative bacteria from flow cytometry profiles) recovered from macrophages (MΦ) exposed to 50× the MIC of oxacillin for 48 h. (B) Time-kill curves against S. aureus infecting macrophages exposed to 50× the MIC of oxacillin, displaying total propidium iodide-negative bacteria from flow cytometry profiles. The kill rate is estimated as a 0.2- or 0.3-log₁₀ decrease in propidium iodide-negative events per hour over the first 3 h of incubation in J774 and human macrophages, respectively, and a 0.02-log₁₀ decrease per hour for longer incubations up to 48 h in both cell types. (C) Flow cytometry profiles of persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h (fill) and their respective postphagocytosis inoculum (line). (D) Cellular ROS production was measured using the oxidation-sensitive fluorescent probe DCF. Cells were incubated with 10 μM DCF for 30 min prior to fluorescence measurement. (E and F) Proportions of dormant persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h (E) or for the indicated periods (F). Dormant bacteria are defined as cells nonproliferating for 24 h on an agar plate in comparison to total propidium iodide-negative bacteria from flow cytometry profiles. (G) Awakening kinetics of persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h and reinoculated in fresh MHB-CA medium, starting from a single FACS event. Lag times were determined by densitometry and calculated in comparison to the awakening time of exponential cultures (n = 30). (H) Flow cytometry profiles of persisters recovered from human macrophages exposed to 50× the MIC of oxacillin for 48 h and then reinoculated in J774 macrophages for an additional 24 h (1 mg L⁻¹ gentamicin to prevent extracellular contamination). Where indicated, J774 macrophages were stimulated with LPS and IFN-γ (J774 MΦ stim), and human macrophages were incubated with BHA (Human MΦ antiox). Data are means ± standard error of the mean (SEM) (A, B, and D to F) or representative results (C, G, and H) from three independent experiments. Statistical analysis was performed with one-way analysis of variance (ANOVA) with Sidak’s posttest. ***, P < 0.001.

FIG 2

ROS-induced ATP depletion correlates with the formation of dark foci. (A) Intrabacterial ATP concentration of persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h, normalized to the total number of viable bacteria. (B) Bright-field images (confocal microscopy) of persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h, 8-h cultures exposed to CCCP for 24 h, or exponential-phase cultures (scale bar, 2 μm). (C) Violin plots displaying the minimal intensity in the bright-field channel of counted bacteria. Bacteria were recovered from 8-h cultures exposed to 6 μM CCCP for 24 h or from macrophages exposed to 50× the MIC of oxacillin for 48 h. Bacteria from macrophages were classified into two groups based on the level of host cell oxidative stress: low oxidative stress (i.e., J774 MΦ and Human MΦ antiox) and high oxidative stress (i.e., J774 MΦ stim and Human MΦ). For low oxidative stress, n = 63, for high oxidative stress, n = 94, and for CCCP, n = 78. Black lines indicate median values. (D) Proportions of persisters displaying dark foci under confocal microscopy. Persisters were recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h. (E) Quantitative real-time PCR of dnaK and clpB transcripts in persisters recovered from macrophages exposed to 50× the MIC of oxacillin for 48 h. Where indicated, J774 macrophages were stimulated with LPS and IFN-γ (J774 MΦ stim), and human macrophages were incubated with BHA (Human MΦ antiox). Data are means ± SEM (A and C to E) or representative results (B) from three independent experiments. Statistical analysis was performed by one-way ANOVA with Sidak’s posttest. ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns, not significant.

FIG 3

ROS-induced translational and ATP synthesis defects regulate bacterial dormancy depth. (A) Translation rate of intracellular persisters. Macrophages were infected by noninduced bacteria for 48 h with 50× the MIC of oxacillin and then induced for GFP expression for the indicated periods. Where indicated, J774 macrophages were stimulated with LPS and IFN-γ (J774 MΦ stim), and human macrophages were incubated with BHA (Human MΦ antiox). (B) Intrabacterial ATP concentration of exponential-phase cultures and 8-h cultures exposed to 6 μM CCCP or 80 μM menadione for 24 h, normalized to the total number of viable bacteria. (C) Quantitative real-time PCR of dnaK and clpB transcripts in 8-h cultures exposed to 80 μM menadione or 6 μM CCCP. (D) Bright-field images (confocal microscopy) of 8-h cultures grown under the control condition or exposed to 80 μM menadione or 6 μM CCCP ± 0.5× the MIC of gentamicin (scale bar, 2 μm) and proportions of bacteria displaying dark foci (E). (F) Translation rate of 8-h noninduced cultures exposed to 80 μM menadione or 6 μM CCCP ± 0.5× the MIC of gentamicin. Bacteria were then induced for GFP expression for the indicated periods. (G) Proportions of dormant bacteria in 8-h noninduced cultures exposed to 80 μM menadione or 6 μM CCCP ± 0.5× the MIC of gentamicin. Dormant bacteria are defined as cells nonproliferating for 24 h on an agar plate in comparison to total propidium iodide-negative bacteria from flow cytometry profiles. (H) Proportions of dormant bacteria recovered from stimulated J774 macrophages exposed to 50× the MIC of oxacillin or 50× the MIC of gentamicin for 48 h. Dormant bacteria are defined as cells nonproliferating for 24 h on an agar plate in comparison to propidium iodide-negative bacteria from flow cytometry profiles. Data are means ± SEM (A to C and E to H) or representative results (D) from three independent experiments. Statistical analysis was performed with two-tailed Student's t test (H) and one-way ANOVA with Sidak’s posttest for multiple comparison (A to C, F, and G). ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns, not significant.

FIG 4

Host oxidative stress drives heterogeneous dormancy of S. aureus persisters among host cells. (A) Cellular ROS production was measured using the oxidation-sensitive fluorescent probe DCF. Cells were incubated with 10 μM DCF for 30 min prior to fluorescence measurement. (B) Proportions of dormant persisters recovered from different host cell types exposed to 50× the MIC of oxacillin for 48 h. Dormant bacteria are defined as cells nonproliferating for 24 h on agar plates in comparison to total propidium iodide-negative bacteria from flow cytometry profiles. (C) Correlation matrix between ROS production, proportions of bacteria displaying dark foci and proportions of dormant bacteria, and their respective frequency distributions. Pearson correlation coefficients (r) and significance (P) are provided for each correlation. Where indicated, THP-1 monocytes were incubated with 200 ng mL⁻¹ PMA for 48 h (THP-1 activated). Data are means ± SEM from three independent experiments.

FIG 5

Model for dormancy of S. aureus persisters under host oxidative stress. In low-oxidative-stress cells, persisters do not show signs of the dark foci observed during protein aggregation. In an oxidative host environment, persisters face ROS-induced translational and ATP synthesis defects; the resulting balance between them defines the dormancy depth. ATP depletion induces protein aggregates that bacteria clear prior to growth resumption, resulting in increased lag times. On the other hand, translation plays a pivotal role to fine-tune the deepness of dormancy, and translation inhibition decreases protein aggregation, resulting in reduced dormant fractions.