Macrophage-Produced Peroxynitrite Induces Antibiotic Tolerance and Supersedes Intrinsic Mechanisms of Persister Formation

ABSTRACT

Staphylococcus aureus is a leading human pathogen that frequently causes chronic and relapsing infections. Antibiotic-tolerant persister cells contribute to frequent antibiotic failure in patients. Macrophages represent an important niche during S. aureus bacteremia, and recent work has identified a role for oxidative burst in the formation of antibiotic-tolerant S. aureus. We find that host-derived peroxynitrite, the reaction product of superoxide and nitric oxide, is the main mediator of antibiotic tolerance in macrophages. Using a collection of S. aureus clinical isolates, we find that, despite significant variation in persister formation in pure culture, all strains were similarly enriched for antibiotic tolerance following internalization by activated macrophages. Our findings suggest that host interaction strongly induces antibiotic tolerance and may negate bacterial mechanisms of persister formation established in pure culture. These findings emphasize the importance of studying antibiotic tolerance in the context of bacterial interaction with the host and suggest that modulation of the host response may represent a viable therapeutic strategy to sensitize S. aureus to antibiotics.

INTRODUCTION

Staphylococcus aureus is a Gram-positive bacterial pathogen that frequently causes chronic and relapsing infections, ranging from relatively minor skin and soft tissue infections to more serious infections like necrotizing pneumonia and bacterial sepsis (1). Despite the availability of antibiotics to treat S. aureus infections, treatment failure is common (2, 3). Antibiotic tolerance is the capacity of bacterial cells to survive for extended periods in the presence of a bactericidal antibiotic, and subpopulations of antibiotic-tolerant cells, called persisters, are frequently implicated in antibiotic treatment failure (4–6). Despite the apparent clinical importance of antibiotic tolerance and persister cell formation, the mechanism(s) underlying their formation during infection remain unclear (7–9).

Antibiotic tolerance and persister cell formation have long been studied using pure bacterial cultures under in vitro conditions. Numerous mechanisms have been identified using this approach, including toxin-antitoxin (TA) modules, induction of the stringent response, reduced respiration, ATP depletion, and metabolic collapse (8, 10–17). However, the relative contributions of these bacterial mechanisms to antibiotic survival during host interaction remain unclear. Recent work from our laboratory and others has identified a role for the host response in inducing antibiotic tolerance (7, 8, 53). Upon bacteremia, S. aureus cells are rapidly phagocytosed by macrophages and reside in the mature phagolysosome (18, 19). Although macrophages are effective at killing S. aureus, subpopulations of bacterial cells survive and these survivors are highly tolerant to antibiotics (7, 8, 19). We previously showed that macrophage-derived reactive oxygen species (ROS) collapse tricarboxylic acid (TCA) cycle activity and deplete ATP, leading to a metabolic state that is incompatible with antibiotic killing (7).

During oxidative burst, a myriad of reactive oxygen and nitrogen species (ROS/RNS) are produced (Fig. 1a). The NADPH oxidase (NOX) complex located on the phagosomal membrane generates superoxide anions from the reduction of molecular oxygen. These superoxide anions are generated inside the phagosomal lumen and penetrate the bacterial membrane poorly (20). Superoxide can be dismutated into oxygen and hydrogen peroxide. The latter is further detoxified into oxygen and water. This occurs spontaneously, but S. aureus also produces a catalase enzyme that facilitates this detoxification. In addition to the production of superoxide, the inducible nitric oxide synthase (iNOS) protein located in the macrophage cytoplasm converts l-arginine to nitric oxide. Nitric oxide diffuses freely into the phagosomal lumen, as well as through the bacterial cell membrane (21). Nitric oxide can be detoxified by S. aureus through the Hmp and Nar proteins (22, 23). Additionally, superoxide and nitric oxide react at a diffusion-controlled rate to form the potent reactive species peroxynitrite (24). Once formed, peroxynitrite can oxidize and nitrate various cellular components, including nucleic acids, proteins, and lipids (24–26). Importantly, the rate of formation of peroxynitrite from nitric oxide and superoxide is an order of magnitude faster than that of the superoxide dismutase-catalyzed conversion of superoxide into hydrogen peroxide, suggesting that peroxynitrite is likely highly abundant during oxidative burst, when superoxide and nitric oxide are both produced (24).

FIG 1.

Peroxynitrite is the main driver of tolerance of rifampicin in macrophages. (a) Schematic representing the production of reactive oxygen and nitrogen species during oxidative burst in macrophages. (b) S. aureus strain HG003 was grown to mid-exponential phase in vitro and exposed to 0.3 U/liter xanthine oxidase (O₂·⁻), 1 mM DEA NONOate (NO·), 2.5 mM H₂O₂, or 2.5 mM peroxynitrite (ONOO⁻) for 20 min prior to the addition of rifampicin (rif) at time zero. At indicated times, an aliquot was removed, washed in 1% NaCl, and plated to enumerate survivors. (c) S. aureus strain HG003 was grown to mid-exponential phase in vitro and exposed to 0.1 mM, 1 mM, or 2.5 mM peroxynitrite (blue bars) or H₂O₂ (orange bars) for 20 min prior to the addition of rifampicin (10 μg/ml) at time zero. At indicated times, an aliquot was removed, washed in 1% NaCl, and plated to enumerate survivors. Data representing T24 are shown. (d) Percent survival of S. aureus cells after internalization by J774 macrophages that were stimulated overnight with LPS + IFN-γ and then treated with 10 μM VAS, 2 mM L-NAME, or 200 μM FeTPPS for 1 h prior to infection. rifampicin (10 μg/ml) was added at time zero, after 30 min of incubation with S. aureus to allow for phagocytosis. After 4 h of rifampicin treatment, S. aureus survival was compared to the survival in the untreated control. (b and d) Data are representative of n = 3 biological samples, and all experiments were repeated 3 times on separate days. Error bars represent the standard deviations (SD). Statistical significance was determined using one-way ANOVA with Sidak’s multiple-comparison test.

In the current study, we aimed to identify which reactive species was predominantly responsible for the induction of antibiotic tolerance in phagocytosed S. aureus. We find that peroxynitrite generated by activated macrophages indiscriminately induced antibiotic tolerance in a variety of S. aureus strains, including clinical bacteremia isolates and mutants that displayed decreased persister formation in vitro. These findings suggest that targeting macrophage-produced peroxynitrite may improve antibiotic efficacy against recalcitrant S. aureus populations in the host and that antibiotic tolerance needs to be investigated in the context of host interaction if physiologically relevant mechanisms of tolerance are to be overcome.

RESULTS

Macrophage-derived peroxynitrite induces tolerance to rifampicin.

We previously reported that ROS produced by stimulated macrophages induced antibiotic tolerance in internalized S. aureus cells (7). To determine which reactive species is the main driver of intracellular antibiotic tolerance, we tested the capacity of superoxide, nitric oxide, peroxynitrite, and hydrogen peroxide to induce rifampicin tolerance in vitro. S. aureus cultures were grown to exponential phase and exposed to various sources of reactive species for 20 min prior to rifampicin challenge. Redox cycling agents like menadione and paraquat have been used extensively to induce superoxide production (7). However, these agents generate superoxide within the bacterial cytoplasm, which does not accurately recapitulate the production of superoxide in the phagosomal lumen. Superoxide produced in the phagosomal lumen poorly penetrates the bacterial membrane (20). To better replicate extracellular superoxide production in vitro, S. aureus was exposed to xanthine oxidase (XO) with acetaldehyde, a well-characterized enzyme that produces superoxide in the growth medium (27), prior to antibiotic challenge. Neither superoxide nor nitric oxide, produced via the nitric oxide donor diethylamine (DEA) NONOate, induced tolerance to rifampicin (Fig. 1b). In contrast, pretreatment with peroxynitrite induced complete tolerance to rifampicin. Hydrogen peroxide was previously shown to induce tolerance when added at high levels (120 mM) (7). However, at a concentration equivalent to that of peroxynitrite, hydrogen peroxide had no impact on rifampicin tolerance (Fig. 1b). The addition of these oxidizing agents in the absence of rifampicin did not cause bacterial cell death (Fig. S1a in the supplemental material). We then examined whether the induction of antibiotic tolerance by peroxynitrite was concentration dependent by challenging cultures with increasing concentrations prior to rifampicin treatment. At sublethal peroxynitrite concentrations (Fig. S1b), tolerance of rifampicin was induced in a dose-dependent manner (Fig. 1c). Equivalent concentrations of hydrogen peroxide had no impact on antibiotic susceptibility (Fig. 1c; Fig. S1b). These results suggest that although superoxide, hydrogen peroxide, or nitric oxide may have the capacity to induce antibiotic tolerance, peroxynitrite does so with far greater potency.

Next, to assess the relative contribution of each species during macrophage infection, J774 macrophages were treated with inhibitors of oxidative and nitrosative burst prior to infection with S. aureus and subsequent exposure to rifampicin. Macrophages were treated with inhibitors of burst for 1 h, followed by infection with S. aureus for 30 min before the addition of rifampicin, to allow for bacterial uptake in the absence of antibiotic. The addition of VAS2870 (VAS), an inhibitor of the NOX complex (which produces superoxide) (28), to the J774 macrophages restored the rifampicin susceptibility of intracellular S. aureus (Fig. 1d; Fig. S1c). Similarly, the addition of L-NAME (l-NG-nitroarginine methyl ester), an inhibitor of iNOS (which produces nitric oxide) (29), to the J774 macrophages restored the rifampicin susceptibility of intracellular S. aureus (Fig. 1d; Fig. S1c). Since superoxide and nitric oxide spontaneously react to produce peroxynitrite, we hypothesized that peroxynitrite was driving the antibiotic tolerance of intracellular S. aureus. In support of this, the addition of FeTPPS [Fe(III)5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrinato chloride], a peroxynitrite decomposition catalyst (30), also sensitized S. aureus to rifampicin (Fig. 1d; Fig. S1c). Taken together, these data suggest that peroxynitrite is primarily responsible for the induction of antibiotic tolerance in phagocytosed S. aureus cells.

Because the duration of an oxidative burst is limited and high levels of ROS are only transiently produced (31, 32), we examined whether the tolerant state was maintained after 2 h of internalization (compared to 30 min in previous experiments). Indeed, we found that tolerance was stable for at least that duration, further suggesting the physiological relevance of the phenotype (Fig. S1d).

Exposure to peroxynitrite reduces TCA cycle activity through posttranslational modification of aconitase.

We have previously shown that exposure to ROS through treatment with menadione reduces TCA cycle activity and depletes ATP levels in S. aureus cells, thus conferring protection from antibiotic killing (7). Additionally, stochastic variation in expression of TCA cycle enzymes was also shown to be involved in the formation of antibiotic-tolerant persister cells (54). Since peroxynitrite appears to be the main mediator of antibiotic tolerance of S. aureus cells internalized by macrophages, we aimed to determine if peroxynitrite does so through TCA cycle inactivation. We found that only S. aureus cells exposed to peroxynitrite displayed reduced levels of intracellular ATP (Fig. 2a), while S. aureus cells exposed to extracellular superoxide, nitric oxide, or hydrogen peroxide had levels of ATP similar to the levels in control untreated cells (Fig. 2a). ATP levels were also significantly reduced in S. aureus cells recovered from stimulated macrophages compared to the levels in unstimulated macrophages (Fig. 2b).

FIG 2.

ATP depletion correlates with antibiotic tolerance. (a) S. aureus strain HG003 was grown to mid-exponential phase and treated with 0.3 U/liter xanthine oxidase (O₂·⁻), 1 mM DEA NONOate (NO·), 2.5 mM H₂O₂, or 2.5 mM peroxynitrite (ONOO⁻) for 30 min. Intracellular ATP was measured using a BacTiter-Glo cell viability assay. (b) Intracellular ATP in S. aureus cells recovered from unstimulated and stimulated macrophages was measured using a BacTiter-Glo cell viability assay. Macrophages were infected with S. aureus strain HG003 at an MOI of 50 for 0.5 h. (a) Data are representative of n = 3 biological samples, and all experiments were repeated 3 times on separate days. (b) Data are from n = 9 biologically independent samples. Error bars represent the SD. Statistical significance was determined using one-way ANOVA with Sidak’s multiple-comparison test (a) or unpaired two-tailed Student’s t test (b).

Aconitase is an iron-sulfur (Fe-S) cluster-containing enzyme known to be extremely sensitive to oxidative stress (33). Peroxynitrite has previously been shown to damage and inactivate mammalian aconitase (33). To confirm that S. aureus aconitase is sensitive to peroxynitrite, we exposed exponential-phase cultures to peroxynitrite and measured aconitase activity. Exposure to peroxynitrite reduced aconitase activity in vitro (Fig. 3a; Fig. S2a). Our prior work using an S. aureus aconitase mutant strain indicates a critical role for aconitase in S. aureus antibiotic tolerance (7). Here, we demonstrate that in vitro, in the absence of ROS, a S. aureus aconitase mutant exhibits increased tolerance to rifampicin relative to the wild type, suggesting that aconitase inactivation is sufficient to induce antibiotic tolerance (Fig. 3b).

FIG 3.

Peroxynitrite-mediated aconitase inactivation is the main driver of antibiotic tolerance. (a) S. aureus strain HG003 was grown to mid-exponential phase and treated with 2.5 mM peroxynitrite for 30 min, and aconitase activity was measured without activation using an aconitase activity kit. The average values from n = 3 biologically independent samples are shown. (b) Survival of S. aureus strain HG003 (black) or HG003 acnA::erm (red). Cells were grown to mid-exponential phase and then treated with 10 μg/ml rifampicin (rif). At indicated times, an aliquot was removed, washed in 1% NaCl, and plated to enumerate survivors. (c) Percent survival of S. aureus strain HG003 (black), HG003/pEPSA5_acnA_FLAG (acnA overexpression) (red), and HG003/pEPSA5 empty vector (green) after internalization by J774 macrophages that were stimulated overnight with LPS + IFN-γ. rifampicin (10 μg/ml) was added at time zero, after 30 min of incubation with S. aureus to allow for phagocytosis. Prior to macrophage infection, all strains were induced with 2% xylose for 1 h. (d) Percent survival of S. aureus strain HG003 (black), HG003 acnA::ermB/pEPSA5_acnA_FLAG (acnA overexpression) (red), and HG003 acnA::ermB/pEPSA5 empty vector (green) after internalization by J774 macrophages that were stimulated overnight with LPS + IFN-γ. rifampicin (10 μg/ml) was added at time zero, after 30 min of incubation with S. aureus to allow for phagocytosis. Prior to macrophage infection, all strains were induced with 0.2% xylose for 1 h. % antibiotic tolerance was calculated relative to no rifampicin treatment at 4 h postinfection (c, d). (e) Western blot for 3-nitrotyrosine (YNO₂). S. aureus strain HG003 acnA::ermB/pEPSA5_acnA_FLAG was induced with 0.2% xylose for 1 h. Unstimulated (unstim) and stimulated (stim) macrophages were infected for 4 h. S. aureus cells were collected, and S. aureus AcnA was immunoprecipitated using anti-FLAG antibody-coupled magnetic beads and blotted for YNO₂. Nonspecific band was used as a loading control. (a to d) Data are representative of n = 3 biological samples, and all experiments were repeated 3 times on separate days. Error bars represent the SD. Statistical significance was determined using one-way ANOVA with Sidak’s multiple-comparison test (b to d) or unpaired two-tailed Student’s t test (a). (e) Blot is representative of n = 3 IP-WB analyses performed on separate days. Bands were quantified using densitometry with FIJI.

If aconitase inactivation is the main driver of antibiotic tolerance, we reasoned that overexpression of aconitase would restore susceptibility in stimulated macrophages. To test this, J774 macrophages were infected with a S. aureus strain overexpressing aconitase (34). Overexpression of aconitase in S. aureus during macrophage infection was sufficient to sensitize S. aureus to antibiotic killing (Fig. 3c and d; Fig. S2b and c), indicating that host-mediated inactivation of aconitase is the primary mediator of tolerance in macrophages.

Next, we aimed to directly measure peroxynitrite-mediated damage of aconitase. Nitration of tyrosine residues is an established method that can be used as a fingerprint of peroxynitrite-mediated damage of proteins (35, 36). It should be noted that studies using purified mitochondrial aconitase suggest that peroxynitrite-mediated disruption of the [4Fe-4S]²⁺ cluster of aconitase is most likely responsible for enzyme inactivation, although nitration of tyrosine residues and oxidation of cysteine residues may contribute to loss of activity (33, 37–39). We infected J774 macrophages with S. aureus cells expressing a FLAG-tagged aconitase (34). Following infection, bacterial aconitase was purified by immunoprecipitation (IP) of the FLAG tag, followed by Western blotting (WB) for 3-nitrotyrosine (YNO₂) (Fig. 3e). Compared to that from unstimulated macrophages (correlating to decreased ROS production [7]), S. aureus aconitase purified from stimulated macrophages exhibited increased levels of YNO₂ (Fig. 3e). Taken together, these data suggest that peroxynitrite-mediated damage of S. aureus aconitase is sufficient to induce rifampicin tolerance.

Peroxynitrite induces antibiotic tolerance in clinical S. aureus bacteremia isolates.

Using a collection of S. aureus clinical isolates from bacteremia patients, we identified major variations in antibiotic tolerance under in vitro conditions, despite the isolates exhibiting similar MICs to rifampicin (Fig. 4a; Fig. S3a). Using 3 low-persister clinical isolates (BC1263, BC1266, and BC1272) and 3 high-persister clinical isolates (BC1267, BC1271, and BC1274) identified in an in vitro screen (Fig. 4a), we examined antibiotic tolerance following phagocytosis by stimulated or unstimulated macrophages. Strikingly, all bacteremia isolates were highly induced for rifampicin tolerance in stimulated macrophages, and multiple log-scale differences in persister formation identified in vitro were no longer evident (Fig. 4b; Fig. S3b to g).

FIG 4.

Antibiotic tolerance is driven by host cell interactions. (a) Percent survival of S. aureus clinical bacteremia isolates following 24 h of rifampicin treatment in vitro. (b) Percent survival of S. aureus clinical bacteremia isolates after internalization by unstimulated (black) J774 macrophages or J774 macrophages stimulated overnight with LPS + IFN-γ (red). rifampicin (10 μg/ml) was added at time zero, after 30 min of incubation with S. aureus to allow for phagocytosis. S. aureus treated with rifampicin for 4 h was compared to the untreated control. Data are representative of n = 2 (a) or n = 3 (b) biological samples, and all experiments were repeated 3 times on separate days. Error bars represent the SD. Statistical significance was determined using the unpaired two-tailed Student’s t test.

While numerous mechanisms of antibiotic tolerance and persister formation have been established in vitro, we were interested in determining the relative importance of these mechanisms in the survival of phagocytosed S. aureus cells exposed to antibiotics. Activation of the stringent response (SR) was shown to contribute to the formation of intracellular S. aureus persisters in unstimulated macrophages (8), and TA modules were shown to contribute to antibiotic tolerance in S. aureus biofilms (10). We infected unstimulated and stimulated J774 macrophages with an SR-deficient RSH (RelA/SpoT homolog) synthase domain mutant (Δrsh_syn) (40) and its wild-type strain HG001 control, as well as a triple-TA module mutant (Δ3TA) (12) and its wild-type strain Newman control. Interestingly, stimulated macrophages significantly induced tolerance of rifampicin in both the Δrsh_syn and Δ3TA mutants (Fig. S4). Stimulated macrophages also induced tolerance of moxifloxacin in the Δrsh_syn mutant similarly to the wild type (Fig. S4e). Together, these data indicate that host-derived ROS-mediated tolerance is the dominant driver of antibiotic tolerance within stimulated macrophages.

DISCUSSION

Within minutes of entering the blood, macrophages engulf S. aureus (19). The initial proinflammatory macrophage response is critical for controlling S. aureus infection. However, if the infection establishes, we find that high levels of ROS can induce antibiotic tolerance, which may contribute to treatment failure. Specifically, we identify peroxynitrite, the reaction product of superoxide and nitric oxide, to be the main driver of ROS-mediated antibiotic tolerance. Peroxynitrite is a potent reactive species capable of nitrating and oxidizing various cellular components, notably Fe-S cluster-containing enzymes (24, 35). We find that peroxynitrite-mediated damage of S. aureus aconitase leads to collapse of the TCA cycle, reduction of ATP, and ultimately, entrance into a metabolic state that is incompatible with antibiotic killing. In support of our findings, Huemer et al. recently demonstrated that antibiotic-tolerant S. aureus recovered from a patient abscess displayed reduced levels of ATP and aconitase activity, despite increased transcription and cytosolic accumulation of aconitase (9).

Why peroxynitrite drives the formation of antibiotic-tolerant S. aureus in macrophages more strongly than other reactive species (i.e., hydrogen peroxide) may be attributed to multiple factors. First, the formation of peroxynitrite from superoxide and nitric oxide is extremely efficient, occurring an order of magnitude faster than the conversion of superoxide into hydrogen peroxide, suggesting that the levels of peroxynitrite are likely extremely high during oxidative burst (24). Second, since the phagolysosome is highly acidic, peroxynitrite will be rapidly protonated and, thus, able to freely pass through the bacterial membrane (24). Additionally, aconitase is characteristically highly sensitive to oxidative damage and the ability of peroxynitrite to efficiently cause aconitase damage and TCA cycle collapse in mammals is well documented (33, 37–39, 41). Aside from damage to aconitase, additional nonlethal bacterial cell modifications induced by peroxynitrite may have importance in host-driven antibiotic tolerance in S. aureus, as peroxynitrite reacts with thiols and sulfur-containing moieties 3 orders of magnitude faster than hydrogen peroxide (42, 43).

Scavenging peroxynitrite exogenously or repairing peroxynitrite-mediated damage has been shown to improve disease outcome in a variety of diseases, including ischemic stroke and septic shock (44–46). There are both direct and indirect methods for resolving peroxynitrite-mediated damage. Thiols, such as glutathione and bacillithiol, directly scavenge peroxynitrite via reduction reactions, whereas ascorbic and uric acid indirectly resolve damage by inhibiting tyrosine nitration (47, 48). Metallic porphyrins, like FeTPPS and WW-85, facilitate decomposition of peroxynitrite by catalyzing the isomerization of peroxynitrite into nitrate (44, 46, 49). WW-85 has been shown to diminish the need for arginine vasopressin treatment during methicillin-resistant S. aureus (MRSA) septic shock, improving disease outcome (46).

Historically, mechanisms underlying antibiotic-tolerant and persister cell formation have been studied under in vitro conditions using pure bacterial cultures. Although numerous mechanisms contributing to persister formation have been identified this way, whether or not these mechanisms contribute to antibiotic tolerance in stimulated macrophages is unknown. We find that previously identified intrinsic bacterial mechanisms underlying persister formation are negated by the powerful tolerance-inducing effects of macrophage-derived ROS. Furthermore, using a collection of clinical isolates, we find that despite variability in persister formation in vitro, all bacteremia strains examined were highly tolerant in stimulated macrophages. Altogether, these data suggest that host-derived peroxynitrite is capable of indiscriminately inducing antibiotic tolerance in an array of genetically diverse S. aureus isolates.

Overall, our results identify host-derived peroxynitrite as a major contributor to antibiotic tolerance during S. aureus infection of macrophages. These findings suggest that acute modulation of peroxynitrite via inhibitors of peroxynitrite formation or peroxynitrite decomposition catalysts may represent a viable therapeutic strategy for improving antibiotic efficacy against recalcitrant S. aureus infection.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

S. aureus strains HG003, HG003 acnA::erm (7), Newman, Newman ΔmazEF Δaxe1-txe1 Δaxe2-txe2 (abbreviated Δ3TA) (12), HG001, and HG001Δrsh_syn (40) were routinely cultured in Mueller-Hinton broth (MHB) at 37°C and 225 rpm. Strains harboring plasmids pEPSA5 empty vector or pEPSA5::acnA_FLAG (34) were grown in the presence of 10 μg/ml chloramphenicol and induced with 0.2% or 2% xylose where indicated. S. aureus bacteremia isolates were obtained under an IRB exemption from a preexisting collection. Bacteremia isolates were cultured in MHB at 37°C and 225 rpm.

Macrophage growth and infection.

Macrophage growth and infection was performed as described previously (7). Briefly, J774A.1 murine macrophage-like cells (ATCC TIB-67) were cultured in Dulbecco’s modified essential medium, high glucose (DMEM) (Gibco), supplemented with 10% fetal bovine serum (FBS) (Millipore), nonessential amino acids (NEAA) (Gibco), sodium pyruvate (Gibco), and l-glutamine (Gibco) in a humidified incubator at 37°C and 5% CO₂. For infections, macrophages were seeded at a density of 2 × 10⁵ per well in 24-well plates in minimum essential medium (MEM) (Gibco) supplemented with 10% FBS and l-glutamine. For ATP assays, macrophages were seeded at 4 × 10⁵ cells/ml in MEM in 6-well plates. For IP-WB, macrophages were seeded at 6.5 × 10⁵ cells/ml in MEM in 10-cm tissue culture-treated petri dishes. To stimulate the macrophages, 500 ng/ml lipopolysaccharide (LPS) from Escherichia coli strain O55:B5 (Sigma) and 20 ng/ml recombinant murine interferon gamma (IFN-γ) (Peprotech) were added to supplemented MEM overnight (50).

S. aureus HG003, HG001, and Newman wild-type or mutant strains were used to infect macrophages at a multiplicity of infection (MOI) of 50. Where indicated, macrophages were pretreated with 10 μM VAS2870 (VAS) (Cayman Chemicals), 2 mM L-NAME (l-NG-nitroarginine methyl ester) (Santa Cruz Biotechnology), or 200 μM FeTPPS [Fe(III)5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrinato chloride] (Cayman Chemicals) for 1 h prior to infection. Plates were centrifuged at 1,000 × g for 2 min to bring bacteria into contact with the macrophages. After 30 min or 2 h (where indicated), the cells were washed once with phosphate-buffered saline (PBS), fresh medium containing 30 μg/ml gentamicin (Fisher) was added to kill extracellular bacteria (51), and 10 μg/ml rifampicin (Fisher) or 3 μg/ml (50× MIC) moxifloxacin (Fisher) was added to the appropriate wells. At indicated times, macrophages were lysed with 0.1% Triton X-100 to release the bacteria. PBS was added to each well, and lysates were resuspended by pipetting, serially diluted in 1% NaCl, and plated to enumerate surviving bacteria. The percentage of survival after rifampicin or moxifloxacin treatment was determined by comparing survivors after 4 h of antibiotic treatment to survivors of the corresponding untreated culture at the 4-h time point. The average values and standard deviations from 3 biological replicates are shown (n = 3). Cell lines were obtained from UNC Lineberger Comprehensive Cancer Center’s Tissue Culture Facility. We did not authenticate or test cells for mycoplasma contamination. Statistical significance was calculated using Student’s t test (unpaired, two-tailed) or one-way analysis of variance (ANOVA) with Sidak’s multiple-comparison test as described in the figure legends.

Data corresponding to this method was not presented in the submitted manuscript.

Antibiotic survival assays.

S. aureus strain HG003 and clinical isolates were cultured aerobically in MHB (Oxoid) at 37°C with shaking at 225 rpm for ∼16 h. Stationary cultures were diluted 1:100 in MHB and grown to mid-exponential phase. An aliquot was plated to enumerate CFU (time zero) before the addition of antibiotics. Where indicated, the culture was incubated with 0.3 U/liter xanthine oxidase (Calbiochem), 1 mM DEA NONOate (Cayman Chemical), 2.5 mM H₂O₂, or 2.5 mM peroxynitrite (Cayman Chemical) for 20 min prior to antibiotic challenge. For the experiment whose results are shown in Fig. 1c, 0.1 mM, 1 mM, 2.5 mM, 5 mM, or 7 mM H₂O₂ or peroxynitrite was used. rifampicin was added at concentrations similar to the maximum concentration of drug in serum (C_max) in humans (10 μg/ml) (52). At indicated times, an aliquot was removed and washed with 1% NaCl. Cells were serially diluted and plated on tryptic soy agar (TSA) to enumerate survivors. The average values and standard deviations from 3 biological replicates are shown (n = 3). Statistical significance was calculated using Student’s t test (unpaired, two-tailed) or one-way ANOVA with Sidak’s multiple-comparison test as described in the figure legends.

Growth curves.

S. aureus strain HG003 was cultured aerobically in MHB (Oxoid) at 37°C with shaking at 225 rpm for ∼16 h. Stationary cultures were diluted 1:100 in MHB and grown to ∼2 × 10⁸ CFU/ml. Growth curves were established by CFU at indicated times after the addition of stresses (0.3 U/liter xanthine oxidase, 1 mM DEA NONOate, 2.5 mM H₂O₂, or 2.5 mM peroxynitrite).

ATP assays.

HG003 was grown to ∼2 × 10⁸ CFU/ml in MHB. Where indicated, cells were exposed to 0.3 U/liter xanthine oxidase, 1 mM DEA NONOate, 2.5 mM H₂O₂, or 2.5 mM peroxynitrite. After 0.5 h, ATP levels were measured in 100-μl amounts of cells as described previously (12), using a BacTiter-Glo kit (Promega) according to the manufacturer’s instructions. For measurement of ATP following internalization in macrophages, unstimulated and stimulated macrophages were infected as described above for 0.5 h, followed by lysis and resuspension of S. aureus. Uninfected controls of both stimulated and unstimulated macrophages were used for background normalization. Following resuspension in PBS, cells were centrifuged at 300 × g for 5 min to collect cellular debris and then at 10,000 × g for 3 min to pellet bacteria. Supernatant was removed, and cells were washed once in PBS and resuspended in a final volume of 200 μl PBS. ATP levels were measured in 100-μl amounts of cells as described above and normalized to CFU counts. The average values and standard deviations from 3 biological replicates are shown (n = 3). Statistical significance was calculated using Student’s t test (unpaired, two-tailed) or one-way ANOVA with Sidak’s multiple-comparison test as described in the figure legends.

Aconitase activity.

HG003 or mutant strains were grown to mid-exponential phase in MHB and incubated with or without 2.5 mM peroxynitrite. After 0.5 h, 2-ml amounts of cells were pelleted, resuspended in 200 μl assay buffer, and lysed with 50 μg/ml lysostaphin at 37°C for 5 min. Samples were pelleted, and the supernatant was assayed for aconitase activity using an aconitase assay kit (Cayman Chemical) without activation according to the manufacturer’s instructions. Aconitase activity was normalized to the CFU count. The average values and standard deviations from 3 biological replicates are shown (n = 3). Statistical significance was calculated using one-way ANOVA with Sidak’s multiple-comparison test as described in the figure legends.

Immunoprecipitation and Western blot analysis of AcnA-YNO₂.

Unstimulated and stimulated J774 macrophages were seeded at 6.5 × 10⁵ cells/ml in supplemented MEM in 10-cm treated dishes and incubated overnight at 37°C, 5% CO₂. S. aureus strain HG003 acnA::erm/pEPSA5_acnA_FLAG was grown in MHB medium for ∼20 h. After 20 h, cultures were induced with 0.2% xylose for 1.5 h, pelleted, and resuspended in PBS. Macrophages were infected at an MOI of 100. Plates were incubated for 30 min at 37°C, 5% CO₂. After 30 min, the cells were washed 2 times with PBS and fresh medium containing 30 μg/ml gentamicin (Fisher) was added to kill extracellular bacteria (51). After 4 h of incubation, cells were washed 3 times with PBS. Five milliliters of 0.1% Triton X-100 was added to each well for 5 min at 37°C to selectively lyse the macrophages and release the bacteria. PBS was added to each plate, and lysates were suspended by pipetting and transferred to conical tubes. Samples were centrifuged at 10,000 × g for 1 min to pellet bacteria and resuspended in 500 μl PBS. Bacterial cells were lysed with lysostaphin (10 μg/ml) at 37°C for 10 min or until confluent lysis was observed. After lysis, 2× protein inhibitor complex (PIC) (Sigma) in PBS was added, and protein levels were normalized by the Bradford assay. Samples were mixed with 75 μl mouse anti-FLAG antibody-coupled magnetic beads (Sigma) and incubated at 4°C for 1 h. After 1 h, beads were washed 4 times for 5 min each with 1× PIC in PBS. Samples were boiled in SDS-reducing buffer and run on a 4 to 12% bis-Tris acrylamide gel (Invitrogen). Protein was transferred onto a polyvinylidene difluoride (PVDF) membrane, and YNO₂ was detected using mouse monoclonal anti-YNO₂ antibodies (1:500; Santa Cruz Biotechnology) with goat anti-mouse IgG-horseradish peroxidase (HRP) secondary antibody (1:10,000; Cayman Chemical). Western blots were analyzed and quantified by densitometry using FIJI.

Statistical information.

Statistical method and sample size (n) are indicated in the methods for each experiment. Statistical analysis was performed using Excel (Microsoft) or Prism 8 (GraphPad) software.

Data availability.

Additional data that support the findings of this study are available from the corresponding author upon request (brian_conlon@med.unc.edu).