Increased Dietary Manganese Impairs Neutrophil Extracellular Trap Formation Rendering Neutrophils Ineffective at Combating Staphylococcus aureus

ABSTRACT

Dietary metals can modify the risk to infection. Previously, we demonstrated that heightened dietary manganese (Mn) during systemic Staphylococcus aureus infection increases S. aureus virulence. However, immune cells also operate in these same environments and the effect of dietary Mn on neutrophil function in vivo has not been assessed. This study reveals that increased concentrations of Mn impairs mitochondrial respiration and superoxide production in neutrophils responding to S. aureus. As a result, high Mn accelerates primary degranulation, while impairing suicidal neutrophil extracellular trap (NET) formation, which decreases bactericidal activity. In vivo, elevated dietary Mn accumulated extracellularly in the heart, indicating that excess Mn may be more bioavailable in the heart. Coinciding with this phenotype, neutrophil function in the heart was most impacted by a high Mn diet, as neutrophils produced lower levels of mitochondrial superoxide and underwent less suicidal NET formation. Consistent with an ineffective neutrophil response when mice are on a high Mn diet, S. aureus burdens were increased in the heart and mice were more susceptible to systemic infection. Therefore, elevated dietary Mn not only affects S. aureus but also renders neutrophils less capable of restricting staphylococcal infection.

INTRODUCTION

Staphylococcus aureus asymptomatically colonizes 20–30% of the population (1–4); however, once the barrier to the host is breached, S. aureus can infect nearly every niche in the body. As such, S. aureus is the leading cause of bacterial endocarditis (5–7) and second most frequent agent of bloodstream infections (8–10). Highlighting S. aureus as a serious global health threat, S. aureus causes nearly 400,000 hospitalizations each year with direct annual costs estimated at $1.7 billion (11). A better understanding of risk factors associated with staphylococcal infections will aid in developing more efficacious treatment strategies.

Transition metals such as zinc, manganese (Mn), iron, and copper are required cofactors for many enzymes involved in essential processes for life. S. aureus encodes two Mn acquisitions systems, MntH and MntABC (12), that are required for colonization of the kidney and liver (13), emphasizing the critical role Mn plays in S. aureus pathogenesis. Multiple risk factors that increase Mn concentrations in the serum and tissue, including intravenous drug use (14), intravenous catheters (15), and chronic liver disease (16, 17) are also associated with a higher risk of S. aureus bacteremia and endocarditis (17, 18). Consistent with these findings, we previously demonstrated that increasing levels of dietary Mn renders mice more susceptible to systemic staphylococcal infections with increased bacterial burdens in the heart (19). We previously showed that this heightened susceptibility to S. aureus was due to increased bioavailability of Mn increasing S. aureus virulence (19–21). However, because immune cells must also operate in these same environments, we hypothesized that changes in Mn concentrations in the tissues may play an underappreciated but integral role in regulating immune cell function during infection (22).

Mn availability enhances the attachment of neutrophils to different extracellular matrices (23, 24), suggesting that Mn may influence cellular motility and chemotaxis. In addition, increasing extracellular Mn also directly influences antimicrobial processes in response to Escherichia coli by heightening the respiratory burst (25) and degranulation (26), which increases neutrophil-mediated killing of E. coli in vitro (25). Whether similar phenotypes are observed in vivo or in response to other bacterial pathogens, such as S. aureus, remains unexplored.

This study reveals that increasing extracellular concentrations of Mn causes accumulation of Mn within neutrophils, coinciding with enhanced total reactive oxygen species (ROS) but impaired mitochondrial superoxide (O₂-) production in response to S. aureus. The generation of mitochondrial O₂- plays a critical role in promoting primary degranulation or suicidal neutrophil extracellular trap (NET) formation (27–29). Consistent with this, we show that increased concentrations of extracellular Mn impair suicidal NET formation, while enhancing primary degranulation. However, the Mn-induced alteration in antimicrobial function rendered neutrophils less bactericidal to S. aureus. In vivo, high dietary Mn accumulated intracellularly within the tissues of the kidney and liver, but in the heart, accumulation was in the extracellular milieu, indicating that excess Mn may be more available to recruited immune cells in the heart. Coinciding with this phenotype, the effect of high dietary Mn on neutrophil function was most prominent in the heart, as neutrophils produced lower levels of mitochondrial O₂- and underwent less suicidal NET formation but increased primary degranulation during staphylococcal infection. Consistent with impaired bactericidal activity, S. aureus burdens were increased in the heart, and mice were more susceptible to systemic infection.

RESULTS

Excess Mn accumulates within neutrophils and impairs mitochondrial O₂- production.

Mice on a high Mn diet are more susceptible to systemic staphylococcal infection resulting in higher bacterial burdens within the heart (19). Since neutrophils play a critical role in restricting S. aureus burdens in the heart (27–29), excess Mn could influence neutrophil function. Therefore, neutrophils were isolated from the bone marrow of C57BL6/J mice and cultured in media supplemented with excess MnCl₂. After 4 h, neutrophils were lysed, and cellular lysates were subjected to inductively coupled plasma mass spectrometry (ICP-MS). Consistent with increased concentrations of Mn in the media, neutrophils accumulated higher levels of Mn intracellularly (Fig. 1). In addition, copper levels were decreased slightly as Mn levels increased, while excess Mn had no effect on intracellular levels of iron and zinc. Thus, accumulation of Mn within neutrophils may suggest that high Mn environments influence neutrophil function.

FIG 1.

Neutrophils accumulate Mn intracellularly. Neutrophils were pretreated with MnCl₂ for 4 h and cellular lysates were obtained. Manganese (⁵⁵Mn), iron (⁵⁶Fe), copper (⁶³Cu), and zinc (⁶⁶Zn) levels were quantified in the lysates by ICP-MS. Elemental abundance was normalized to sulfur (³⁴S). Each point represents neutrophils isolated from a single mouse (2 experiments, n = 4). One-way analysis of variance (ANOVA) with Tukey multiple-comparison test (*, P ≤ 0.05, ***, P ≤ 0.001, ns = not significant).

It is possible that the accumulation of Mn within neutrophils has an antioxidant effect by quenching O₂- (30, 31), which could influence ROS production by NADPH oxidase and O₂- production by mitochondrial respiration. Therefore, neutrophils were cultured with S. aureus, and ROS and mitochondrial O₂- production were quantified by flow cytometry. Consistent with previous observations in response to E. coli (25), neutrophils cultured in higher concentrations of Mn and stimulated with S. aureus produced more ROS than neutrophils cultured in standard media (Fig. 2A, Fig. S1A). In contrast, production of mitochondrial O₂- was significantly lower in neutrophils cultured in higher concentrations of Mn using MitoSOX (Fig. 2B, Fig. S1B), a dye that localizes to mitochondrial structures and fluoresces upon oxidation by O₂-. To determine whether lower levels of mitochondrial O₂- were the result of excess intracellular Mn mediating antioxidant activity or inhibiting electron transport chain function, mitochondrial membrane potential was quantified in neutrophils using the dye tetramethylrhodamine methyl ester (TMRM). TMRM was used at a quenching concentration so that decreases in fluorescence intensity correlate to an increase in membrane potential (32, 33). In response to S. aureus, neutrophils underwent a transient increase in membrane potential within 15 min, followed by a long-term decrease in membrane potential out to 120 min (Fig. 2C, Fig. S1C). However, neutrophils cultured in elevated concentrations of Mn maintained their membrane potential but did not undergo the transient increase or long-term decrease in membrane potential observed in neutrophils cultured in standard media (Fig. 2C).

FIG 2.

Increased concentrations of Mn disrupt mitochondrial membrane potential and O₂- production in response to S. aureus. Neutrophils were pretreated with MnCl₂ for 4 h, rotenone (Rot.; 0.5 μM) or MitoPQ (10 μM) for 20 min and stimulated with S. aureus (multiplicity of infection; MOI = 10) and (A) total ROS and (B, D) mitochondrial O₂- production, and (C) mitochondrial membrane potential (ΔΨ_m) were quantified by flow cytometry. (A) Median fluorescence intensity (MFI) normalized to unstimulated WT, (B, D) MitoSOX MFI normalized to MitoTracker MFI, (C) TMRM used at a quenching concentration. Each point represents neutrophils isolated from a single mouse (A, 4 experiments,0–120 min, 0 mM n = 9, 10 μM n = 5, 100 μM n = 4, 360 min n = 4; B, 3 experiments, 0–120 min n = 5, 360 min n = 4; C, 4 experiments, n = 5; D, 2 experiments, n = 4). Two-way ANOVA with Tukey multiple-comparison test (*, P ≤ 0.05, **, P ≤ 0.01, ***, P ≤ 0.001, ****, P ≤ 0.0001, ns = not significant).

To determine whether impaired electron transport chain activity accounts for less mitochondrial O₂- production, neutrophils were treated with rotenone or MitoPQ prior to culturing with S. aureus. The generation of O₂- during rotenone treatment results from impairing complex I activity (34), while MitoPQ is a version of paraquat that localizes to mitochondria and produces O₂- independent of electron transport (35). Treatment with rotenone or MitoPQ had an equivalent impact on mitochondrial O₂- production by neutrophils in response to S. aureus; however, in high Mn conditions, only MitoPQ restored production of mitochondrial O₂- to levels comparable to neutrophils in standard media (Fig. 2D). Since rotenone had no effect on mitochondrial O₂- production in the presence of high Mn, these results suggest that elevated concentrations of intracellular Mn inhibit mitochondrial O₂- production by dampening the activity of the electron transport chain.

Excess Mn accelerates primary degranulation and impairs suicidal NET formation in response to S. aureus.

The production of mitochondrial O₂- plays a critical role upstream of primary degranulation and suicidal NET formation in neutrophils (27–29). Therefore, neutrophils were stimulated with S. aureus and suicidal NET formation was quantified by a flow cytometry-based technique that has been previously described (Fig. S1D; 28). Neutrophils cultured in higher levels of Mn exhibited a significant decrease in suicidal NET formation in response to S. aureus compared to neutrophils in standard media (Fig. 3A, Fig. S1E). The impaired suicidal NET formation by neutrophils in high Mn was the result of decreased mitochondrial O₂- production as MitoPQ treatment was sufficient to restore suicidal NET formation to WT levels (Fig. 3B). Furthermore, similar to what has previously been observed using phorbol myristate acetate (PMA) stimulation (26), culturing neutrophils in higher levels of Mn caused an acceleration in primary degranulation in response to S. aureus, as quantified by the abundance of CD63 on the surface of the cell by flow cytometry (Fig. 3C, Fig. S1F). These findings demonstrate that increased Mn causes neutrophils to undergo accelerated primary degranulation and impaired suicidal NET formation in response to S. aureus.

FIG 3.

Increased concentrations of Mn accelerate primary degranulation and impair suicidal NET formation. Neutrophils were pretreated with MnCl₂ for 4 h, rotenone (Rot.; 0.5 μM) or MitoPQ (10 μM) for 20 min and stimulated with S. aureus (MOI = 10) and (A-B) the percentage of neutrophils (Ly6G⁺CD11b⁺) undergoing suicidal NET formation (Dead: extracellular dsDNA⁺MPO⁺H3Cit⁺) and (C) primary degranulation (surface CD63; sCD63) by neutrophils was quantified by flow cytometry. (C) MFI normalized by an isotype control. Each point represents neutrophils isolated from a single mouse (A, 3 experiments,0–240 min n = 5, 360 min n = 4; B, 2 experiments, n = 4; C, 4 experiments,0–120 min, 0 mM n = 8, 10 μM n = 4, 100 μM n = 5, 360 min n = 4). Two-way ANOVA with Tukey multiple-comparison test (*, P ≤ 0.05, **, P ≤ 0.01, ****, P ≤ 0.0001, ns = not significant).

Impairing suicidal NET formation decreases bactericidal activity of neutrophils in response to S. aureus (28, 29, 36, 37). To determine whether the Mn-induced change in neutrophil function alters bactericidal activity, neutrophils were cultured with S. aureus and bacterial survival was quantified by measuring CFU (CFU). Increased concentrations of Mn had no effect on the capacity of neutrophils to restrict bacterial survival over the first 2 h; however, at 6 h neutrophils cultured in higher concentrations of Mn failed to restrict S. aureus survival to the same extent as neutrophils cultured in standard media (Fig. 4A). The observed differences were not the result of Mn directly influencing bacterial replication as bacterial growth was comparable across Mn concentrations in the absence of immune cells. Furthermore, the lower bactericidal activity of neutrophils in high Mn was the result of decreased mitochondrial O₂- production as preloading neutrophils with MitoPQ was sufficient to restore bactericidal activity to levels comparable to WT (Fig. 4B).

FIG 4.

Increased concentrations of Mn impairs neutrophils bactericidal activity. (A) Neutrophils were pretreated with MnCl₂ for 4 h and stimulated with S. aureus (MOI = 1) and S. aureus (SA) growth was quantified by CFU (CFU) dilution spot plating. Percent survival of S. aureus calculated relative to S. aureus growth in the absence of immune cells. Each point represents the mean result (technical triplicate) of immune cells isolated from a single mouse (4 experiments, n = 4). (B) Neutrophils pretreated with MnCl₂ for 4 h were treated with rotenone (Rot.; 0.5 μM) or MitoPQ (10 μM) for 20 min. Neutrophils were washed, stimulated with S. aureus (MOI = 1) in media supplemented with MnCl₂, and S. aureus growth was quantified by CFU dilution spot plating. Each point represents the mean result (technical triplicate) of immune cells isolated from a single mouse (2 experiments, n = 4). (C) Neutrophils (Neut.) or S. aureus were pretreated with MnCl₂ for 4 h, washed, and cultured together (MOI = 1) for 6 h in standard culture media. S. aureus growth was quantified by CFU dilution spot plating. Percent survival of S. aureus calculated relative to S. aureus growth in the absence of immune cells. Each point represents the mean result (technical triplicate) of immune cells isolated from a single mouse (2 experiments, n = 3). Two-way ANOVA with Tukey multiple-comparison test (*, P ≤ 0.05, ****, P ≤ 0.0001, ns = not significant).

To determine whether high Mn impairs neutrophil function or enhances S. aureus fitness, neutrophils or bacteria were preloaded with Mn prior to the culturing in standard culture media. Loading S. aureus with high Mn prior to culturing with neutrophils marginally increased S. aureus survival; however, the result did not achieve statistical significance (Fig. 4C). In contrast, loading neutrophils with high Mn prior to culturing with S. aureus significantly increased S. aureus survival, consistent with decreased bactericidal activity (Fig. 4C). Since the generation of ROS and primary degranulation occur within the first 2 h of engaging S. aureus and suicidal NET formation occurs after 2 h, these results are consistent with the decreased long-term bactericidal activity observed in neutrophils cultured in higher concentrations of Mn reflecting impaired suicidal NET formation.

Mice on a high Mn diet exhibit reduced suicidal NET formation in the heart.

Mice on a high Mn diet accumulated increased concentrations of Mn in their tissue prior to and during infection (Fig. S2A; 19), but whether the metal is bioavailable to recruited neutrophils is unclear. Mice were placed on a high Mn diet for 6 weeks (wk) followed by systemic infection with S. aureus by retroorbital inoculation. Four days postinfection (dpi), tissues were isolated from the mice, gently homogenized, and metal abundance was quantified in the supernatant (extracellular) and cell pellet (intracellular) by ICP-MS. Mice on a high Mn diet accumulated a higher abundance of Mn intracellularly in the liver and kidney during infection (Fig. 5A); however, Mn accumulation was extracellular in the heart (Fig. 5B). These results are consistent with Mn being more bioavailable to recruited neutrophils during infection in the heart compared to other tissues, such as the kidney and liver where Mn is stored intracellularly within the tissue.

FIG 5.

Increased dietary Mn accumulates extracellularly in the heart during staphylococcal infection. Mice were on standard or high Mn diet 6 wk prior to systemic infection with S. aureus (CFU = 2 × 10⁷). At 4 dpi, tissues were homogenized and pelleted. The supernatant was isolated and the cellular pellet lysed. The abundance of Mn in the (A) lysates (intracellular; intra.) and (B) supernatant (extracellular; extra) were quantified by ICP-MS. Mn abundance was normalized to sulfur (³⁴S). Each point represents a single mouse (2 infections, control, n = 8; high Mn, n = 14). Unpaired t test (**, P ≤ 0.01, ***, P ≤ 0.001, ****, P ≤ 0.0001, ns = not significant).

Previous infections of mice on a high Mn diet used the S. aureus strain Newman (19), which has been laboratory adapted over many decades (38), and compared to more recent clinical isolates, may differ in virulence. Similar to what was observed with Newman (19), mice on a high Mn diet that were infected with the methicillin resistant S. aureus (MRSA) strain USA300 (LAC) showed increased mortality (Fig. 6A), despite losing less weight (Fig. S2B). In addition, bacterial burdens were increased specifically in the heart of mice on a high Mn diet, 4 dpi compared to mice on a standard diet (Fig. 6B). Significant mortality was not observed past 4 dpi as surviving mice tended to resolve the infection. These results demonstrate that a high Mn diet renders mice susceptible to systemic MRSA infection.

FIG 6.

Impaired suicidal NET formation in the heart of mice on a high Mn diet coincides with increased susceptibility to staphylococcal infection. Mice were on standard or high Mn diet 6 wk prior to systemic infection (inf.) with S. aureus (CFU = 2 × 10⁷). (A) During the infection, mouse survival was monitored. Each point represents the percentage of living mice (mock n = 5; control inf. n = 11, high Mn inf. n = 20). (B–F) At 4 dpi, organs were homogenized and (B) CFU were enumerated using dilution spot plating (limit of detection = LoD) and (C) ROS and (D) mitochondrial O₂- production by neutrophils (Ly6G⁺CD11b⁺), (E) primary degranulation (sCD63) by neutrophils, and (F) percentage of neutrophils undergoing suicidal NETosis (dead: extracellular dsDNA⁺MPO⁺H3Cit⁺). (C) MFI normalized to uninfected WT, (D) mitoSOX MFI was normalized by mitoTracker MFI, and (E) MFI normalized by an isotype control antibody. Each point represents a single mouse (2 infections, control n = 8; high Mn n = 14). (A) Log-rank (Mantel-Cox) test or (B–F) unpaired t test (*, P ≤ 0.05, **, P ≤ 0.01, ***, P ≤ 0.001, ****, P ≤ 0.0001, ns = not significant).

The accumulation of Mn within cultured neutrophils enhanced generation of ROS by NADPH oxidase while quenching mitochondrial O₂- production (Fig. 2A and B). To determine whether similar phenotypes observed in vitro were present during infection of mice fed a high Mn diet, mice were infected and ROS and mitochondrial O₂- production were quantified in neutrophils by flow cytometry at 4 dpi. Total ROS production by neutrophils was comparable in the liver, kidney, and heart between mice on standard and high Mn diets (Fig. 6C, Fig. S2C); but, mitochondrial O₂- production was decreased in neutrophils in the liver and heart of mice on a high Mn diet (Fig. 6D, Fig. S2D). These results suggest that a high Mn diet suppresses the capacity of neutrophils to produce mitochondrial O₂- during infection.

Mitochondrial O₂- production influences primary degranulation and suicidal NET formation (27–29). Consistent to what was observed in vitro (Fig. 3C), neutrophils in the liver and heart of mice on a high Mn diet have enhanced primary degranulation (Fig. 6E, Fig. S2E) compared to mice on a normal diet, which coincides with lower production of mitochondrial O₂- (Fig. 6D). However, the percentage of neutrophils undergoing suicidal NET formation was only significantly decreased in the heart of mice on a high Mn diet (Fig. 6F, Fig. S2F), corresponding with increased bacterial burdens (Fig. 6B). Suicidal NET formation was lower in the liver but it did not achieve statistical significance (Fig. 6F), while mitochondrial O₂- production, primary degranulation, and suicidal NET formation in the kidney was comparable between mice on standard and high Mn diets (Fig. 6D to F). These results suggest that a high Mn diet suppresses neutrophil antibacterial function in the heart by preventing suicidal NET formation, while the effects of a high Mn diet have a less severe impact on the bactericidal activity of neutrophils in the liver and kidney. As a result, mice on a high Mn diet are more susceptible to systemic staphylococcal infection with increased bacterial burdens in the heart.

DISCUSSION

We previously observed that a high Mn diet causes increases in mortality following S. aureus infection, in part by enhancing bacterial virulence. Here, we expand on those initial findings and show that a high Mn diet renders mice more susceptible to systemic staphylococcal infection by negatively impacting neutrophil antibacterial activity. We show that in cultured neutrophils, excess Mn accumulates intracellularly (Fig. 1) and alters mitochondrial respiration and O₂- production in response to S. aureus (Fig. 2B to D). As a result, increased Mn accelerates primary degranulation, inhibits suicidal NET formation (Fig. 3), and impairs bactericidal activity (Fig. 4); a phenotype that is most prevalent during staphylococcal infection in the heart (Fig. 6). The effect of a high Mn diet on neutrophil function in the heart correlates with an accumulation of extracellular Mn (Fig. 5B), which suggests that excess Mn is more bioavailable to recruited neutrophils in the heart compared to other organs such as the kidney and liver. These data demonstrate the deleterious effects of excess dietary Mn on neutrophil function during infection and provide further evidence that elevated dietary Mn heighten susceptibility of the host to S. aureus.

We have previously demonstrated that mitochondria within neutrophils impact antibacterial function and bactericidal activity in response to multiple bacterial pathogens, including S. aureus (27–29). Since then, alternative roles for neutrophil mitochondria in facilitating phagocytosis and direct oxidative stress upon S. aureus have been described (39). These findings highlight the complex role that mitochondria play within neutrophils during infection. The accumulation of Mn within mitochondria has detrimental effects on energy metabolism (40–46) and oxidative phosphorylation (42, 44, 46, 47). Similarly, we demonstrate that increased accumulation of intracellular Mn (Fig. 1) coincides with decreased mitochondrial O₂- production (Fig. 2B) and dampened membrane polarization dynamics (Fig. 2C) in response to S. aureus. The disruption of oxidative phosphorylation is potentially the result of Mn binding readily to calcium, magnesium, and iron sites at many steps in the overall process of oxidative phosphorylation. Excess Mn preferentially inhibits ATP synthase activity in mitochondria from liver and heart tissue, while complex II is preferentially inhibited in brain mitochondria (47). This suggests that how Mn metabolically influences host-pathogen interactions can be tissue specific. We provide evidence that the bioavailability of Mn varies across tissues, where in the heart, Mn is found extracellularly, where it may be more accessible to recruited neutrophils (Fig. 5).

Our data are consistent with elevated concentrations of Mn broadly dampening changes in mitochondrial membrane potential in neutrophils responding to S. aureus (Fig. 2C and D). While this would dramatically affect ATP production in other cell types, neutrophils do not use mitochondria to generate energy, as most of the ATP is derived from glycolysis (48–50). Instead, ATP generated from oxidative phosphorylation plays a critical role in regulating neutrophil activation and antibacterial functions (48). In addition, generation of mitochondrial O₂-, a by-product of oxidative phosphorylation, also plays an equally important role in regulating neutrophil function (28, 29). We demonstrate that increased Mn disrupts electron transport thereby preventing mitochondrial O₂- production and leading to accelerated primary degranulation; however, the rapid decrease in surface CD63 may reflect that releasing primary granules under conditions of high Mn results in neutrophil death (Fig. 3C). In addition, high Mn also impairs suicidal NET formation (Fig. 3A and B), which is particularly important in combating staphylococcal infections in the heart (Fig. 6). However, similar phenotypes may not be observed with other bacterial pathogens, as increased concentrations of Mn render neutrophils more bactericidal to E. coli (25). This suggests that whether high dietary Mn is beneficial or detrimental to neutrophil function may depend on the bacterial pathogen during infection. A clearer understanding on how nutrient metals shape immune cell and pathogen biology at the site of infection and how these interactions change depending on the tissue requires further study.

Shared risk factors between increased Mn and the development of bacterial infections provides clinical relevance to these findings. Risk factors such as intravenous drug use, intravenous catheters, and chronic liver disease increases Mn levels in the patient (14–16), coinciding with a higher risk of S. aureus endocarditis and bacteremia (17, 18). We and others have shown previously that acquisition of Mn is critical for the virulence of S. aureus (13, 19–21, 51), which can increase susceptibility to infection (19). Here, we build further on this model by demonstrating that excess Mn limits suicidal NET formation by neutrophils (Fig. 3A and 5F). NETs play a critical role in facilitating direct antibacterial activity during staphylococcal infections (29, 36, 37) and augmenting macrophage bactericidal activity (27, 28). These findings support the model that limiting Mn supplementation can be beneficial to the host by reducing S. aureus virulence and promoting a more bactericidal neutrophil response.

MATERIALS AND METHODS

Mice.

C57BL/6J were purchased from Jackson Laboratory (JAX mice stock no. 000664) and maintained in an accredited animal facility. Mice were housed 2–5 to a cage in specific pathogen-free conditions and randomly assigned to experimental groups. Food and water were provided ad libitum. For diet experiments, diets were synthesized by Dyets Inc. (Bethlehem, PA) using the AIN-93M standardized diet (52). The base diet was formulated as Mn and zinc-free with Mn and zinc carbonate salts added to achieve the following concentrations: control: 10 ppm Mn, 29 ppm zinc; high Mn: 500 ppm Mn, 29 ppm zinc. 6-week-old female mice were placed on custom-synthesized diets for 6 weeks prior to infection and maintained throughout infection.

Reagents.

Antibodies specific for Ly6G (1A8), and rabbit IgG were from Biolegend; CD11b (M1/70) was from Tonbo; CD63 was from BD Biosciences; myeloperoxidase (2D4), histone citrulline, neutrophil elastase, and dsDNA (3519 DNA) were from Abcam. Streptavidin Alexa 488 was from Biolegend. Live/dead stain, dihydrorhodamine 123 (DHR123), TMRM, MitoSOX red, and MitoTracker deep red were from Invitrogen; Helix NP Blue was from Biolegend. Murine Fc-blocking antibody (2.4G2) was from Tonbo. Rotenone was from Sigma and MitoPQ was from Abcam. Paraformaldehyde (PFA) was from Electron Microscopy Sciences. Histopaque 1119 and Histopaque 1077 were from Sigma. DMEM and PBS were from Gibco and fetal bovine serum (FBS) from Atlanta biologicals were used for all tissue culture.

Bacterial strains.

All experiments used the bacterial strain USA300 (LAC) (53). Bacteria were maintained as a −80°C stock and S. aureus was streaked onto plates containing tryptic soy agar (TSA; 2% agar) 2 days prior to each experiment. S. aureus plates were grown at 37°C overnight. Single colonies of S. aureus were transferred to liquid cultures containing tryptic soy broth (TSB) and grown overnight at 37°C with 180 rpm shaking. Prior to in vitro experiments, overnight cultures were diluted 1:10 into non-heat-inactivated FBS for 2 h on ice.

Animal infections.

Retroorbital infections were performed as described (28, 29, 54). S. aureus was streaked from frozen stocks onto plates containing TSA 2 days prior to infection. Single colonies were grown overnight in TSB, and subcultured 1:100 into 5 ml TSB on the following day. S. aureus was grown to the mid-exponential phase, washed, and resuspended in 100 μl ice-cold PBS.10–13-week-old female mice were anesthetized prior to infection with intraperitoneal injection of 2,2,2-tribromoethanol diluted in PBS. Systemic infections were induced by intravenous injection of the retroorbital sinus with 2 × 10⁷ CFU S. aureus. Mice were monitored for 4 days prior to humane euthanasia by inhalation of CO₂ at 4 days postinfection.

Organ harvest for CFU and neutrophil characterization.

Organs were processed as previously described (28, 29). Single-cell suspensions were generated from tissue of kidney, liver, and hearts by processing organs through a 5 μm filter into FACs media (PBS, 2% FBS, 0.02% NaAz). Blood was flushed from the chambers of the heart to ensure immune cells were isolated from the tissue. Heart tissue was cut into smaller pieces to aid in processing through the filter. Aliquots from each organ homogenate were removed for CFU dilution spot plating to quantify bacterial burdens. Organ homogenates were pelleted, and the supernatant was removed to quantify metal abundance by ICP-MS. The remaining organ homogenate was either lysed for ICP-MS analysis, or resuspended and red blood cell-lysed per manufacturer’s instructions (Biolegend). The resulting single cell suspensions were divided into separate aliquots in FACS media to be stained ex vivo for flow cytometry.

Neutrophil isolation.

Neutrophils were isolated as previously described (28, 29). Single-cell suspensions of bone marrow were prepared from the tibias and femurs of mice. Polymorphonuclear granulocytes were isolated by density centrifugation to obtain neutrophils at the interface between layers of Histopaque 1119 and Histopaque 1077. Prior to any experiments, isolated neutrophils were rested on ice for at least 1 h in D10 media (DMEM, 10% FBS) before being transferred to ultra-low cluster round bottom 96-well plates (Costar) to incubate (37°C, 5% CO₂) for 4 h in D10 supplemented with 0, 10, or 100 μM MnCl₂. The isolated cells were 85–95% neutrophils (CD11b⁺Ly6G⁺).

Flow cytometry.

Flow cytometry was conducted as previously described (28, 29). All data were collected using a BD LSRII flow cytometer with FACSDIVA software and analyzed using FlowJo (FlowJo LLC, Ashland, OR). All flow cytometry samples were gated side scatter height (SSC-H) by side scatter area (SSC-A) followed by forward scatter height (FSC-H) by forward scatter area (FSC-A) to remove doublet populations. The singlet population was gated SSC-A by FSC-A to isolate the granulocyte population. The resulting cell population was then assessed for assay-specific fluorescent markers as described below:

Exogenous treatments: Rotenone (0.5 μM) or MitoPQ (10 μM) was added 20 min prior to stimulation with S. aureus.

Degranulation flow: Single-cell suspensions isolated from the tissue or isolated neutrophils (referred to as cells for the remainder of the flow cytometry section) were stained 20 min prior to fixation with a live/dead dye to monitor membrane integrity. Cells were fixed with room temperature 4% paraformaldehyde and incubated for 15 min on ice. Cells were pelleted, aspirated, and resuspended in Fc block diluted in FACS media for 30 min on ice. Cells were washed and stained with anti-CD11b, -Ly6G, and -CD63 (primary) or appropriate isotype controls diluted in FACS media for 30 min on ice. Following staining, cells were washed, and resuspended in FACS media, and maintained at 4°C until analysis. Cells were gated for live cells (Live/Dead-negative) and then neutrophils (CD11b- and Ly6G-positive). The median fluorescence intensity (MFI) was quantified for surface CD63 relative to staining with an isotype control antibody.

Suicidal NET flow: As described above, cells were live/dead stained, but Helix NP Blue (2.5 nM) was also added to stain for extracellular DNA structures. Cells were fixed and blocked as described above. Cells were washed and stained for neutrophils (anti-CD11b and -Ly6G), extracellular myeloperoxidase (MPO), and citrullinated histones (H3Cit) in FACS media for 30 min on ice. Cells were washed and stained with fluorescent streptavidin and anti-rabbit IgG diluted in FACS media for 30 min on ice. Cells were washed following staining, resuspended in FACS, and maintained at 4°C until analysis. Cells were gated for neutrophils and then live or dead cells. Neutrophils that are positive for extracellular dsDNA, MPO, and H3Cit with permeabilized cell membranes were defined as having undergone suicidal NET formation (representative gating in Fig. S1D). The percentage of neutrophils undergoing suicidal NET formation was quantified relative to the total number of neutrophils.

Total ROS or mitochondrial O₂- flow: As described in ‘Degranulation Flow’, cells were live/dead stained, but DHR123 (3 μM) was also added to the culture to quantify total ROS production or MitoSOX red (5 μM) and MitoTracker deep red (500 nM) were added to quantify mitochondrial O₂- production. As described in ‘Degranulation Flow’, cells were fixed, blocked and stained for neutrophils (anti-CD11b and -Ly6G). Cells were washed following staining, resuspended in FACS media, and maintained at 4°C until analysis. Cells were gated for live cells and then neutrophils. For total ROS production, the DHR123 MFI was quantified relative to unstimulated neutrophils (in vitro assays) or mock treated mice (in vivo infections). For mitochondrial O₂-, mitoSOX MFI was normalized to mitoTracker MFI.

Membrane potential flow: Neutrophils were cultured in phenol red-free D10 media supplemented MnCl₂. Two hr prior to experiment, neutrophils were stained with MitoTracker deep red (500 nM) for 15 min at 37°C. Medium was removed and fresh D10 medium, supplemented with MnCl₂ and tetramethylrhodamine methyl ester (TMRM) at a quenching concentration (200 nM; 32, 33), was added to the cells for 2 h at 37°C. Cells were stimulated with S. aureus and at appropriate time points, unfixed cells were analyzed by flow cytometry. The TMRM MFI was normalized to mitoTracker MFI.

Metal quantification by ICP-MS.

Neutrophils or cell pellets from tissue were digested at 65°C for 24 h in 100 μl of OPTIMA grade nitric acid (Thermo Fisher Scientific) and 25 μl of ultratrace hydrogen peroxide (Thermo Fisher Scientific). After digestion, 875 μl of Ultrapure water (Invitrogen) was added to the samples to decrease the nitric acid concentration to below 10% (vol/vol). Supernatant from the tissues were analyzed without acid digestion. An Agilent 7700 inductively coupled plasma mass spectrometer (Agilent, Santa Clara, CA) attached to a Teledyne CETAC Technologies ASX-560 autosampler (Teledyne CETAC Technologies, Omaha, NE) was used for elemental quantification. Cell Entrance = −40 V, Cell Exit = −60 V, Plate Bias = −60 V, OctP Bias = −18 V, and collision cell Helium Flow = 4.5 ml/min were fixed during the analysis. Optimal voltages for Extract 2, Omega Bias, Omega Lens, OctP RF, and Deflect were determined empirically at the beginning of each analysis. Element calibration curves were generated using ARISTAR ICP Standard Mix (VWR, Radnor, PA). A peristaltic pump with 0.5 mm internal diameter tubing through a MicroMist borosilicate glass nebulizer (Agilent) was used to introduce the samples. Samples were initially up taken at 0.5 rps for 30 sec followed by 30 sec at 0.1 rps to stabilize the signal, and analyzed in spectrum mode at 0.1 rps collecting three points across each peak and performing three replicates of 100 sweeps for each element analyzed. Sampling probe and tubing were rinsed for 20 sec at 0.5 rps with 2% nitric acid between every sample. Agilent Mass Hunter Workstation Software version A.01.02 was used to acquire and analyze the data.

S. aureus killing assay.

S. aureus killing assays were conducted as previously described (28, 29). Neutrophils were cultured with S. aureus at a multiplicity of infection (MOI) of 1 (100,000 CFU per 100,000 immune cells) in D10 media (37°C, 5% CO₂) supplemented with Mn (Fig. 4A). In S. aureus killing assays where neutrophils or S. aureus were preloaded with Mn (Fig. 4B), neutrophils or S. aureus were cultured in D10 media supplemented with 100 μM Mn for 4 h at (37°C, 5% CO₂). Cells were washed and neutrophils were cultured with S. aureus (MOI = 1) in standard D10 media (37°C, 5% CO₂). After 0.5, 2, and 6 h, cultures were serially diluted and spot-plated onto TSA. Percent growth was quantified by dividing the CFU (CFU) of the S. aureus-immune cell cultures by the bacteria alone cultures.

Statistics.

Specific statistical details can be found in the corresponding figure legend for each experiment. For applicable figures, error bars represent standard error of the mean across mice. The specific number of replicates is noted in the corresponding figure legend with a minimum of three experimental replicates were performed for each assay. All P values were calculated using a one- or two-way analysis of variance (ANOVA) with a Sidak’s or Tukey multiple-comparison test), unpaired t test, log-rank (Mantel-Cox), or parts of a whole analysis test when applicable. Statistical work was performed using Prism 6 software (GraphPad) and significance is indicated on the graphs as follows: *, P ≤ 0.05, **, P ≤ 0.01, ***, P ≤ 0.001, ****, P ≤ 0.0001, ns = not significant.