Functional expression of system ${\text{x}}_{\text{c}}^{-}$ is upregulated by asbestos but not crystalline silica in murine macrophages

Abstract

Context

Inhalation of asbestos or silica is associated with chronic and progressive diseases, including fibrosis, cancer, and increased risk of systemic autoimmunity. Because there is a need for treatment options for these diseases, a better understanding of their mechanistic etiologies is essential. While oxidative stress in macrophages is an early consequence of these exposures, it may also serve as a signaling mechanism involved in downstream immune dysregulation. The system ${\text{x}}_{\text{c}}^{-}$ exchange protein is induced by oxidative stress, and exchanges equimolor levels of extracellular cystine for intracellular glutamate. Cystine is subsequently reduced to cysteine, the rate-limiting precursor for glutathione synthesis.

Objective

As the primary transporter responsible for cystine/glutamate exchange on macrophages, system ${\text{x}}_{\text{c}}^{-}$ was hypothesized to be inducible in response to asbestos and silica, and to increase viability through protection from oxidative stress.

Results

When challenged with amphibole asbestos, but not crystalline silica, RAW 264.7 macrophages increased expression of xCT and the rate of cystine/glutamate exchange in sodium-free conditions. This upregulation was prevented with N-acetylcysteine, implicating oxidative stress. Cystine protected the macrophages from asbestos-induced oxidative stress and cell death, supporting the hypothesis that imported cystine was used for synthesis of cellular antioxidants. System ${\text{x}}_{\text{c}}^{-}$ inhibitors, glutamate and S-4-carboxyphenylglycine ((S)-4-CPG), significantly increased oxidative stress and cell death of asbestos-treated macrophages.

Conclusion

System ${\text{x}}_{\text{c}}^{-}$ plays a critical role in survival of macrophages exposed to asbestos, but not silica. These data demonstrate a very early difference in the cellular response to these silicates that may have important downstream implications in the pathologic outcome of exposure.

Introduction

Macrophages are key in orchestrating immune responses to inhaled material, using signals sent to lymphocytes that help tailor an appropriate response dependent on their recognition of various types of pathogen through specialized receptors. In the case of inhaled dusts such as silica and asbestos, however, there remain questions about (a) how the macrophage recognizes and distinguishes these materials and (b) what the subtle signals are that lead to the distinct disease outcomes following exposure. While much is known about macrophage-derived signals that influence immune function, there are many subtle regulatory steps that are still not understood. Interaction of macrophages with asbestos and silica causes production of reactive oxygen species (ROS) (Blake et al., 2007; Shen et al., 2001; Shukla et al., 2003). Some of the macrophages simply die from the oxidative damage (Blake et al., 2007; Shen et al., 2001), but some survive and impact ultimate immune/inflammatory outcomes. Macrophages play several roles in activating the immune system, primarily through the production of response-tailoring cytokines and lipid mediators, and these activities are sensitive to redox signals (Murata et al., 2002; Peterson et al., 1998; Pfau et al., 2004; Schneider et al., 2005).

System ${\text{x}}_{\text{c}}^{-}$ is an amino acid transport system expressed on many cell types. It appears to be the major glutamate/cystine antiporter expressed on murine macrophages (Watanabe & Bannai, 1987), although another glutamate transport system (system ${\text{x}}_{\text{AG}}^{-}$)has also been reported on human macrophages (Rimaniol et al., 2000). Transport via system ${\text{x}}_{\text{c}}^{-}$ is sodium-independent, unlike other anionic amino acid transporters such as system ${\text{x}}_{\text{AG}}^{-}$,but is dependent on chloride (Waniewski & Martin, 1984). It consists of two subunits xCT, the light chain; and a heavy chain, 4F2 (CD98), forming a functional heterodimer (Sato et al., 1999) (Figure 7). Although transport is reversible (dependent on amino acid concentration gradients), in macrophages undergoing oxidative stress, System ${\text{x}}_{\text{C}}^{-}$ exchanges intracellular glutamate for extracellular cystine. Following import, cystine is reduced to cysteine, which may serve not only as an antioxidant itself (Banjac et al., 2008), but also acts as the rate-limiting precursor for the synthesis of glutathione, the predominant cellular thiol-antioxidant (Meister & Anderson, 1983).

Figure 7.

A simplified representation of the system ${\text{x}}_{\text{c}}^{-}$ transporter and exchange of amino acids for the purpose of cystine import and protection from oxidative stress. System ${\text{x}}_{\text{c}}^{-}$ is a heterodimer of the proteins xCT and CD98, which is the constitutively expressed heavy chain (4F2).

Early experiments on system ${\text{x}}_{\text{c}}^{-}$ illustrated that transport activity for cystine is induced by oxygen (Bannai et al., 1989). Accordingly, xCT expression decreases when cells are cultured under low oxygen conditions (Sato et al., 2001). Thus, oxygen affects system ${\text{x}}_{\text{c}}^{-}$ transporter expression and activity, and it follows logically that particles that cause oxidative stress in macrophages, such as silica and asbestos (Vallyathan et al., 1992) would likewise induce its activity.

The ultimate pathologies resulting from exposure to amphibole asbestos are unique when compared to those caused by crystalline silica (Maeda et al., 2010; Mossman & Churg, 1998), despite the fact that both are inhaled (impacting macrophages early on), both cause oxidative stress in macrophages, both elicit a variety of pro-inflammatory mediators, and both can induce apoptosis in macrophages (Table 1). Initially, we expected that silica and asbestos would cause similar upregulation of system ${\text{x}}_{\text{c}}^{-}$ due to these commonalities. The data presented here demonstrate that while silica did not upregulate system ${\text{x}}_{\text{c}}^{-}$, asbestos treatment resulted in increased functional expression of the transporter. Additionally, a GSH-precursor, N-acetylcysteine (NAC), was able to decrease ROS, prevent the increase in system ${\text{x}}_{\text{c}}^{-}$, and protect RAW macrophages from cell death. These data suggest that the oxidative conditions produced by asbestos exposure are able to enhance survival of macrophages through the upregulation of system ${\text{x}}_{\text{c}}^{-}$. In this way, system ${\text{x}}_{\text{c}}^{-}$ may serve as an initial response that could ultimately be responsible for a whole host of downstream signaling cascades as a result of asbestos exposure, as well as represent an attractive therapeutic target for early intervention in asbestos-induced disease.

Table 1.

Summary of effects of crystalline silica and amphibole asbestos.

	Amphibole asbestos	Crystalline silica	References
Pulmonary inflammation	Yes	Yes	Holley et al. (1992); Mossman and Churg (1998)
ROS production by macrophages	Yes	Yes	Blake et al. (2007); Vallyathan et al. (1992); Hu et al. (2006)
Apoptosis in macrophages	Yes	Yes	Hamilton et al. (2008); Shen et al. (2001); Blake et al. (2008); Shukla et al. (2003)
Superoxide production by macrophages	Yes, via PLC/PKC pathway	Mixed results	Holian et al. (1994); Premasekharan et al. (2011); Lim et al. (1997); Quinlan et al. (1998)
Lung Mn Superoxide Dismutase (MnSOD) expression & activity	Yes	No (some mixed results)	Holley et al. (1992); Janssen et al. (1992)
NALP3 Inflammasome activation	Yes	Yes	Dostert et al. (2008); Biswas et al. (2011)
Pro-fibrotic	Yes, asbestosis	Yes, silicosis	Mossman and Churg (1998) (Reviewed)
Radiography	Bilateral diffuse interstitial fibrosis, lower zones	Silicotic nodules, upper zones	Mossman and Churg (1998) (Reviewed)
Pleural disease	Common	Less common; found in advanced disease	Mossman and Churg (1998); Arakawa et al. (2005); Elshazley et al. (2011)
Carcinogenic	Yes, with mesothelioma	Yes, without mesothelioma	Maeda et al. (2010)
Anti-nuclear antibodies	Yes	Yes (some mixed results)	Pfau et al. (2005); Cooper et al. (2008); Zaghi et al. (2010)

Materials and methods

Chemicals, particulates and fibers

Libby amphibole asbestos, collected and characterized by the US Geological Survey, and purified crocidolite were kindly provided by the Center for Environmental Health Sciences (CEHS), University of Montana. Libby MT asbestos (6-Mix) has been called “6-Mix”, because it was collected from six sites around the mine area near Libby. It is amphibole asbestos a mixture of tremolite, amosite, richterite and other amphiboles. Crystalline silica and wollastonite were obtained from US Silica and the National Institutes of Standard and Technology (NIST), respectively, by the CEHS and kindly provided for these experiments. Fibers or silica were made up as stock suspensions at 1 mg/mL in sterile phosphate buffered saline (PBS) or medium and then sonicated in a water-bath sonicator (Branson Ultrasonics, Danbury, CT, USA) for at least 1 min before adding the suspension to the cell cultures on an equal mass basis. Most experiments were performed at 35 μg/cm², which was shown not to induce significant cell death by silica or asbestos, and most treatments were for 24 h, unless otherwise indicated.

System ${\text{x}}_{\text{c}}^{-}$ inhibitor, S-4-carboxyphenyl-glycine ((S)-4-CPG), was purchased from Tocris Bioscience (Ellisville, MO, USA). Lipopolysaccharide (LPS, from Salmonella minnesota), L-cystine, L-glutamic acid, N-acetylcysteine (NAC) and other chemicals were purchased from Sigma Chemical (St. Louis, MO, USA).

Cells and cell culture

RAW 264.7 macrophages were obtained from American type culture collection (ATCC, Rockville, MD, USA) and maintained at 37°C in Dulbecco’s modified eagle’s medium (DMEM) (CellGro Mediatech, Manassas, VA, USA) containing L-glutamine, 100 U/mL penicillin/streptomycin, and 5% fetal calf serum (FBS, Hyclone) in a humidified CO₂ incubator. For experiments, the cells were switched to serum free DMEM without cystine, cysteine or glutamate (Gibco/Invitrogen, Carlsbad, CA, USA) to which amino acids could be added individually at known concentrations. Sodium-free buffer consisted of 1.1 mM CaCl₂, 5.36 mM KCl, 0.77 mM KH₂PO₄, 0.71 mM MgSO₄, 137 mM choline chloride, 11 mM D-glucose, and 1 mM HEPES.

Flow cytometry

RAW 264.7 cells were harvested and counted, then blocked in PBS with 3% bovine serum albumin and Fc Block (anti-CD16/32, BD Biosciences, San Jose, CA, USA) for 10 min at room temperature. Approximately 1 × 10⁶ cells per sample were stained in 100 μL of blocking buffer with antibodies to xCT (Santa Cruz Biotech, clone Q-18) and CD98 (BD Biosciences). Secondary antibodies were anti-goat IgG conjugated with fluorescein isothiocyanate (FITC) and anti-rat conjugated with Phycoerythrin (PE), respectively (Molecular Probes/Invitrogen). Stained cells were washed thoroughly with cold PBS and analyzed immediately or fixed with 1% paraformaldehyde overnight. Staining was analyzed on a BD Biosciences (San Jose, CA, USA) FACS Calibur flow cytometer, with gates set on live cells, and a positive/negative cut-off gate was based on cells stained with secondary antibody alone. Data analysis was performed using Cell Quest software (BD Biosciences), with data expressed as percent positive, or median fluorescence intensity (MFI) of the entire population.

System ${\text{x}}_{\text{C}}^{-}$ activity (L-Glu transport assay)

RAW 264.7 cells were plated at 5 × 10⁵ cells/well in 12-well plates containing RPMI 1640 growth media supplemented with 5% FBS. Cultures were maintained at 37°C and 5% CO₂ for 24-h post-plating until cells were 80–90% confluent. The cultures were then treated with indicated amounts of asbestos, silica, and LPS for 12, 24, or 48 h. System ${\text{x}}_{\text{C}}^{-}$ activity was assayed using a protocol previously described with a few modifications (Warren et al., 2004). Briefly, on the day of the assay, plates were removed from the incubator, growth media (RPMI with 5% FBS) was removed from the 12-well culture plates and replaced with a Na⁺-free buffer (mM), CaCl₂ (1.1), KCl (5.36), KH₂PO₄ (0.77), MgSO₄ (0.71), choline chloride (137.5), D-glucose (11), HEPES (10), pH 7.4. Following 5 min pre-incubation in Na⁺-free buffer (30°C), the buffer was aspirated and replaced with buffer containing 100 μM [³H]-L-glutamate (New England Nuclear). Following 5-min incubation at 30°C, the assays were terminated by three sequential 1-mL washes with ice-cold buffer and then the cells were dissolved in 1 mL of 0.4 M NaOH for 24 h. An aliquot (200 μL) was transferred into a 5-mL glass scintillation vial and neutralized with 5 μL glacial acetic acid, followed by the addition of 3.5 mL Liquiscint scintillation fluid (National Diagnostics) to each sample. Incorporation of [³H]-L-glutamate was quantified by liquid scintillation counting (LSC, Beckman LS 6500). All transport activity rates were normalized for protein determined via the BCA method (Pierce). Values are reported as mean pmol/min/mg protein ± SEM and are corrected for nonspecific uptake (e.g. leakage and binding) by subtracting the amount of [³H]-L-glutamate accumulated at 4°C.

Viability assay

Viability was assessed using the CellTiter Blue reagent from Promega. RAW 264.7 cells were plated at 1 × 10⁶ cells/mL in 96-well cell culture plates (100 μL/well) in media lacking cystine, cysteine and glutamate, unless supplemented, and challenged with indicated amounts (μg/cm²) of particles or fibers, in the presence or absence of amino acids, NAC, or (S)-4-CPG for 24 h. Triton-X-100 at 1% was used as a positive control for cell death. Following the challenge, 20 μL/well of CellTiter Blue dye was added to each well and the plate was tapped gently to mix. The plate was incubated at 37°C for 1 h and read on a Synergy HT microtiter plate reader (BioTek Instruments, Winooski, VT, USA) at 570 nm.

Glutamate assay

The amount of glutamate that was transported out of the cells was measured using the Amplex®-Red Glutamic Acid/Glutamate Oxidase assay kit (Molecular Probes/Invitrogen). The assay is based on the oxidation of glutamate by glutamate oxidase, producing alpha-ketoglutarate, NH₃, and H₂O₂. The H₂O₂ reacts with the Amplex®-Red reagent to produce the fluorescent product, risorufin. A standard curve from 0 to 20 μM was prepared using the kit’s glutamate stock. The positive control was 10 μM H₂O₂ in reaction buffer. Macrophages were cultured in 24-well plates, in media lacking glutamate, but containing cystine as the substrate for glutamate transport. Media from the challenged RAW 264.7 cells was diluted in the kit’s reaction buffer and placed in the wells of an opaque 96-well plate, 50 μL/ well. 50 μL of Amplex®-Red working solution was added to each well, and the plate was incubated at 37°C for 30 min to 1 hour. Fluorescence was detected at 590 nm using the Synergy plate reader. Concentrations of L-glutamate were calculated from the standard curve.

DCFDA assay

Dichlorofluorescein diacetate (DCFDA, Molecular Probes/ Invitrogen) was used as a measure of general oxidative stress. The cells were plated in media lacking cystine, cysteine or glutamate, and challenged in 96-well opaque cell culture plates as described above, with or without supplemented NAC, cystine, or glutamate. Following the challenge period, the media was removed and 50 μL of 10μM DCFDA was added to each well. Wells containing no cells served as a negative control, and cells treated with 10 μM hydrogen peroxide served as the positive control. The plate was incubated for designated times at 37°C, and then read at 520 nm on the microtiter plate reader.

Glutathione assay

Cellular GSH was measured using the GSH-Glo^™ Glutathione assay from Promega following the manufacturer’s instructions. Briefly, after challenge with asbestos in media containing cystine as the substrate for system ${\text{x}}_{\text{c}}^{-}$, the media was removed from the macrophages, and the Luciferin-NT and glutathione S-transferase (GST) were added to the reaction buffer and the wells. After 30-min incubation, the Luciferin Detection Reagent (containing luciferase) was added to the plate, incubated for an additional 15 min and then read on the luminescence setting of the microtiter plate reader. In the presence of GSH, the Luciferin-NT is reduced by GST to Luciferin, which then fluoresces in the presence of ATP and Luciferase.

Statistical analyses

All experiments were repeated at least three times, and representative data is shown. Bar graphs represent mean values in each group, and error bars are standard error of the mean (SEM). Statistical analyses were performed using Excel or Kaleidograph software (Synergy Software, Reading PA, USA). Unpaired two-tailed t-tests were used wherever appropriate, for numerical measurements of a mean against a single control mean. One-way ANOVA was used to compare treatment groups with one independent variable, with a Tukey post-hoc test. Statistical significance was assigned as p < 0.05, with significance at p < 0.01 indicated with double asterisk or letters as indicated in the legends.

Results

Increased expression of xCT subunit by 6-Mix asbestos

Exposure to 6-Mix asbestos or to lipopolysaccharide (LPS) increased the expression of xCT on RAW cells as determined by flow cytometry (Figure 1A). Treatment with silica or wollastonite (a non-fibrogenic control fiber) did not significantly upregulate the constitutive level of expression. LPS was used as a positive control, since it has been shown to increase system ${\text{x}}_{\text{c}}^{-}$ activity (Sato et al., 1995). CD98 was expressed at a high level on over 90% of RAW cells regardless of treatment, indicating that its expression is constitutive on these cells (data not shown). A representative flow cytometry histogram demonstrating upregulation of xCT following 6-Mix treatment on RAW macrophages is shown in Figure 1B.

Figure 1.

(A) Expression of xCT on RAW 264.7 cells by flow cytometry, showing percent positive (above secondary antibody alone). Cells were treated with 35 μg/cm² of 6-Mix asbestos, silica, wollastonite, or 1 μg/mL LPS for 24 h. *p < 0.05 compared to no treatment, n = 3 in each group. (B) Representative histogram shows untreated cells (gray) at 14.2% positive and 6-Mix asbestos (black line) is 67.7% positive. The M1 gate is set on background staining with the secondary antibody only.

Increased system ${\text{x}}_{\text{c}}^{-}$ transport activity following 6-Mix exposure

The increased expression of the xCT subunit induced by 6-Mix also led to increased Na+-independent [³H]-L-Glu transport, indicating that the upregulated expression yields a functional system ${\text{x}}_{\text{c}}^{-}$ (Figure 2A). [³H]-L-Glu transport activity increased in a dose-dependent manner at both 12 and 24 h, in response to asbestos. However, no increase in transport was seen with silica exposure (Figure 2B), consistent with the lack of surface expression upregulation from control values. In fact, there appears to be a decrease in the transporter activity at the higher concentrations (Figure 2B). We hypothesized that this might be due to increased stress and/or cell death, which are addressed below.

Figure 2.

(A) Asbestos increased [³H]-L-glutamate transport in a dose dependent manner (10, 50 100 μg/cm²), using a radiolabeled glutamate uptake assay, as described in “Materials and Methods”. (B) Using the same technique as in A, silica time course/dose response for [³H]-L-glutamate transport. *p < 0.05, **p < 0.01, using t-test compared to control, or one-way ANOVA, n = 3 in each group.

Because the cystine transport model assay used for Figure 2 forces glutamate import (due to application of high extracellular glutamate), we also tested the physiologically relevant efflux of glutamate out of the macrophages using the Amplex®-Red assay. Glutamate measured in the macrophage media following exposure to 6-Mix was significantly increased, and this effect was inhibited by addition of (S)-4-CPG, the selective system ${\text{x}}_{\text{c}}^{-}$ inhibitor (Figure 3).

Figure 3.

Asbestos increased L-glutamate in the media of RAW 264.7 cells. The RAW cells were cultured in media free of glutamate for 2 days in the absence or presence of 35 μg/cm² asbestos, 1 μg/mL LPS, and (S)-4-CPG (500μM). The media was collected, centrifuged to remove any cells or debris, and assayed for L-glutamic acid using the Amplex®-Red glutamic acid assay. L-Glutamate was quantified against a standard curve using a fluorescence microtiter plate reader, ex 530 nm/em 590 nm. *p < 0.05 compared to no treatment, n = 5 in each group; **p < 0.01 compared to no CPG with same treatment. CPG, carboxyphenylglycine.

NAC prevents asbestos-induced ROS and system ${\text{x}}_{\text{c}}^{-}$ expression

To demonstrate that the upregulation of system ${\text{x}}_{\text{c}}^{-}$ was due to ROS following 6-Mix exposure, RAW cells exposed to 6-Mix were first assayed for oxidative stress using the ROS-indicator dye, dichlorodihydrofluorescein diacetate (DCFDA). Indeed, as shown in Figure 4A, the fluorescence of cells loaded with 6-Mix asbestos and DCFDA increased above controls, indicating that ROS production is linked to asbestos exposure, consistent with earlier studies (Blake et al., 2007). This increase in fluorescence was prevented by providing N-acetylcysteine (NAC, 2.5 mM). NAC pretreatment also prevented the increase in system ${\text{x}}_{\text{c}}^{-}$ expression (Figure 4B). These data demonstrate (albeit indirectly) that system ${\text{x}}_{\text{c}}^{-}$ induction is linked to ROS generation following asbestos exposure. This is consistent with reports that xCT expression is regulated, at least in part, by redox signaling (Bannai et al., 1989; Sato et al., 2001).

Figure 4.

RAW 264.7 cells treated for 24 h with 35 μg/cm² 6-Mix increased ROS, as measured by DCFDA fluorescence. (A) 6-Mix asbestos increased ROS in RAW cells, which was inhibited by addition of NAC (2.5 mM) or cystine (100 μM). n = 6 in each group, ^ap< 0.05 compared to no treatment, ^bp < 0.05 compared to 6-Mix asbestos only. (B) Percent positive for xCT expression with asbestos treatment, in the presence or absence of NAC (2.5 mM). n = 4 in each group, *p < 0.05 compared to no treatment. ROS, reactive oxygen species; DCFDA, dichlorofluorescein diacetate; NAC, N-acetylcysteine.

System ${\text{x}}_{\text{c}}^{-}$ inhibition increases asbestos-induced ROS and GSH depletion

The synthesis of intracellular GSH levels in macrophages is highly dependent upon the transport of cystine via system ${\text{x}}_{\text{c}}^{-}$ (Sato et al., 2001). Inhibition of cystine transport by the non-transportable inhibitor (S)-4-CPG increased the asbestos-induced ROS (Figure 5A). As expected, inhibition of system ${\text{x}}_{\text{c}}^{-}$ also led to a dramatic depletion in the intracellular levels of GSH (Figure 5B). Thus, through its link to GSH production, system ${\text{x}}_{\text{c}}^{-}$ limits the production of ROS in macrophages exposed to asbestos and, when inhibited, leads to increased ROS.

Figure 5.

(A) ROS production increased with addition of system ${\text{x}}_{\text{c}}^{-}$ inhibitor (S)-4-CPG (500 μM) in the presence of 6-Mix asbestos. (B) Glutathione was measured following 24 h treatment with 35 μg/cm² 6-Mix, in the presence or absence of NAC or (S)-4-CPG. n = 4 in each group, ^ap < 0.05 compared to no treatment, ^bp < 0.05 compared to asbestos treatment. ROS, reactive oxygen species; NAC, N-acetylcysteine; (S)-4-CPG, S-4-carboxyphenylglycine.

System ${\text{x}}_{\text{c}}^{-}$ activity helps protect macrophages from asbestos-induced cell death

To better understand the influence of system ${\text{x}}_{\text{c}}^{-}$ on macrophage viability in response to ROS, asbestos exposure both with and without GSH-precursors and transport-inhibitors was examined. Exposure to 6-Mix asbestos in the absence of cystine resulted in a loss of macrophage viability in a dose-dependent fashion starting at 25 μg/ cm² (Figure 6A, solid line with diamonds). Pretreatment with NAC (2.5 mM) resulted in the highest level of protection of RAW cells from asbestos-induced cell death, and cystine (2.5 mM) protected up to 50 μg/cm² asbestos. Conversely, extracellular glutamate increased the toxicity of asbestos (Figure 6A). We hypothesize that this is due to inhibition of the system ${\text{x}}_{\text{c}}^{-}$ transporter, since glutamate is a substrate inhibitor of cystine transport (Patel et al., 2004). Although addition of NAC helped to protect the macrophages from silica-induced cell death, cystine did not, providing further evidence that macrophages do not use system ${\text{x}}_{\text{c}}^{-}$ to protect from silica-induced cytotoxicity (Figure 6B), and thus do not use this pathway to replenish intracellular cysteine. Figure 6C shows that while supplemental cystine was protective for asbestos-induced cell death, blocking system ${\text{x}}_{\text{c}}^{-}$ with a specific inhibitor led to cell death even in the presence of cystine.

Figure 6.

(A) Supplementing with cystine protected the cells from asbestos toxicity (measured by CellTiter Blue) up to 50 μg/cm². RAW 264.7 cells were plated in media containing no cystine or glutamate, unless otherwise indicated. Pretreatment with N-acetylcysteine protected the cells at all concentrations, relative to cells treated with asbestos or silica only. (B) Cystine or glutamate had no effect on silica-induced cytotoxicity. (C) In a separate experiment, cystine is again shown to be protective from asbestos cytotoxicity, and this protection is lost by addition of (S)-4-CPG. For A, B, n = 8, *p < 0.05 compared to asbestos (or silica) only at the same concentration of asbestos or silica. For C, n = 3, ^ap < 0.01 compared to no treatment, ^bp < 0.01 compared to black bars (no antioxidant added) by unpaired t-test. (S)-4-CPG, S-4-carboxyphenylglycine.

Discussion

This study reports a novel function for the system ${\text{x}}_{\text{c}}^{-}$ glutamate/cystine transporter in protecting macrophages from oxidative stress caused by amphibole asbestos. System ${\text{x}}_{\text{c}}^{-}$ is an amino acid antiporter that exchanges intracellular glutamate for extracellular cystine in macrophages (Figure 7). Import of cystine is critical for maintenance of redox status in many cell types. Cystine is reduced intracellularly to cysteine, which can serve as a redox modifier itself (Banjac et al., 2008), or as a precursor to glutathione. Glutathione (GSH), a major intracellular antioxidant, is composed of glutamate, glycine and cysteine, with the bioavailability of cysteine being the limiting factor for glutathione synthesis (Townsend et al., 2003). Transport by system ${\text{x}}_{\text{c}}^{-}$ is sodium-independent, unlike other glutamate transporters such as system ${\text{x}}_{\text{AG}}^{-}$, but is dependent on chloride (Waniewski & Martin, 1984). It is also ATP-independent, driven by concentration gradients of the amino acids. This study shows that RAW 264.7 cells express the heavy chain of system ${\text{x}}_{\text{c}}^{-}$, 4F2 (CD98) constitutively, and that xCT, which is the actual transporter, is inducible on these cells by amphibole asbestos (6-Mix) and LPS. The complete heterodimer is required for transport function (Sato et al., 1999). Since outside the central nervous system the normal gradients for cystine and glutamate strongly drive the system in the direction of cystine import, protein expression of xCT appears to be the limiting factor for the rate of transport activity. Early experiments on the system ${\text{x}}_{\text{c}}^{-}$ illustrated that transport activity for cystine is induced by oxygen (Bannai et al., 1989), and Sato et al., (2001) have shown that under low oxygen conditions, xCT expression is reduced compared to normal cell culture oxygen conditions. Thus, oxygen affects system ${\text{x}}_{\text{c}}^{-}$ transporter expression and activity, and it follows logically that particles that cause oxidative stress in macrophages, such as silica and asbestos (Vallyathan et al., 1992), would likewise induce its activity. Indeed, amphibole asbestos (Libby MT 6-Mix) increased protein expression of xCT, and caused induction of system ${\text{x}}_{\text{c}}^{-}$ activity in a dose-dependent manner. Because the increased expression was inhibited by addition of cell-permeable cysteine (N-acetylacysteine, NAC), which can serve as an antioxidant (Voisin et al, 1987), this induction appears to be related to oxidative stress.

Unexpectedly, silica did not induce increased expression and transport activity, despite strong evidence of oxidative stress and glutathione depletion by silica (Hamilton et al., 2008; Hu et al., 2006). It is possible that the activation of cell death signaling pathways predominated with silica under these conditions, since it has been shown that cytokine release and cytotoxicity following silica exposure are independently regulated, and depend on time and dose considerations (Claudio et al., 1995). Our data show that starting around 50 μg/ cm², silica did reduce cell viability in a dose-dependent fashion in the absence of cystine. However, the cells were not protected by supplementation by cystine or cell death exacerbated by glutamate, ruling out a role for system ${\text{x}}_{\text{c}}^{-}$. It is possible, therefore, that silica induces cell death through specific receptors (Claudio et al., 1995; Hamilton et al., 2008; Hamilton et al., 2006) which then lead to upregulation of other antioxidant pathways since pharmacological mimics for catalase and superoxide dismutase (SOD) blocked silica-induced cytokine production (Premasekharan et al., 2011). In fact, restoration of glutathione levels in macrophages can occur through a variety of anti-oxidant mechanisms. In vivo, the profile of antioxidant enzyme expression is different for silica versus amphibole asbestos exposure, suggesting different lung defense mechanisms to these dusts (Janssen et al., 1992). All of this is consistent with our data showing that NAC protected the macrophages from silica-induced cytotoxicity at concentrations above 50 μg/cm². However, apparently maintenance of cellular GSH in response to silica and subsequent ROS does not include uptake of cystine, since supplementation with cystine along with silica challenge did not protect the cells from silica cytotoxicity as it did for asbestos. Therefore, the system x_c⁻ is not induced by silica exposure, and in fact amino acid transport appeared to be somewhat inhibited in silica-treated macrophages (Figure 2). It is very unlikely that silica led to cystine uptake by another transporter for three reasons: (a) cystine was not protective for silica cytotoxicity in our system, (b) system x_c⁻ has been shown to transport the vast majority of cystine needed by murine macrophages (Watanabe & Bannai, 1987), and (c) our activity assays were performed in sodium-free buffer. In fact, the other glutamate transport system shown to be expressed on macrophages, system ${\text{x}}_{\text{AG}}^{-}$ does not transport cystine and is sodium-dependent (Rimaniol et al., 2000, 2001). Therefore, our data showing that silica-exposed macrophages do not use the system ${\text{x}}_{\text{c}}^{-}$ as an antioxidant system, unlike asbestos-exposed macrophages, are novel and interesting in view of differences in the antioxidant responses and ultimate outcomes of these two exposures. Table 1, while not intended to be a complete review of the literature, summarizes some of the similarities and differences between responses to crystalline silica and amphibole asbestos.

The amount of glutamate found in media from untreated cells (Figure 3) is consistent with the moderate level of system ${\text{x}}_{\text{c}}^{-}$ expression in cultured untreated macrophages (Figure 1). Watanabe and Bannai (1987) have shown that simply culturing cells in a typical cell culture incubator (with near atmospheric oxygen partial pressure as opposed to physiological levels in vivo) induces expression of the transporter. Because even the moderate level of glutamate in the cell-conditioned media from untreated cells was inhibited by pre-treatment with (S)-4-CPG (a selective non-substrate inhibitor) (Figure 3), system ${\text{x}}_{\text{c}}^{-}$ clearly helps protect macrophages from the oxygen levels in culture by providing adequate cystine for maintenance of redox balance. Further evidence of protection from oxidative stress by system ${\text{x}}_{\text{c}}^{-}$ activity comes from pretreatment with (S)-4-CPG, which led to significantly increased ROS and depleted GSH in the presence of amphibole asbestos (Figure 5).

The pharmacology of system ${\text{x}}_{\text{c}}^{-}$ continues to be defined. Because glutamate acts at a number of receptors and transporters in the CNS and elsewhere, a certain degree of cross-reactivity is associated with all pharmacological agents developed thus far (Bridges et al., 2012). (S)-4-CPG, while originally developed as a group I metabotropic receptor antagonist (Bedingfield et al., 1995), is now one of the most potent system ${\text{x}}_{\text{c}}^{-}$ inhibitors characterized to date (Ye et al.,1999; Patel et al., 2004). Unlike many of the other known substrate inhibitors of system ${\text{x}}_{\text{c}}^{-}$, (S)-4-CPG is ideal for inhibition studies as it essentially acts as a non-substrate (very weak substrate), allowing for inhibition of the cystine/glutamate exchange without leading to increased glutamate efflux (Patel et al., 2004). In this study, used purely as a pharmacological agent to block cystine or glutamate entry over brief periods, (S)-4-CPG allowed us to examine the short-term effects of near complete transporter inhibition at acute time points (Patel et al., 2004; Seib et al., 2011).

The ability of system ${\text{x}}_{\text{c}}^{-}$ to protect macrophages from asbestos cytotoxicity was examined by culturing the cells with asbestos in the presence or absence of NAC, cystine, glutamate, or (S)-4-CPG (Figure 6). NAC maintained viability up to very high concentrations of asbestos, but cystine was protective (relative to cells in the absence of any cystine/cysteine) only up to 50 μg/cm². There are several possibilities for the loss of protection above that asbestos level. First, other cell death pathways or simple membrane disruption may be activated at higher levels of asbestos exposure in culture. Second, the amount of ROS generated by the asbestos may exceed the rate of cystine transport in these conditions. It has been shown that L-glutamine is also required for maximal activity of system ${\text{x}}_{\text{c}}^{-}$, as it is the primary source of intracellular glutamate (Piani and Fontana, 1994). In our experimental media, no L-glutamine was provided, mostly because we did not want extracellular conversion of glutamine to glutamate to confound glutamate measurements. However, this likely led to the cell death at higher concentrations of asbestos despite the presence of cystine. Third, assuming the cells did have adequate intracellular glutamate, export of that glutamate into the culture medium would eventually diminish the gradient driving the transporter and thus inhibit the ability to import cystine. Also, supplementing the cells with glutamate increased the cytotoxicity of asbestos at all concentrations, again likely due to its role as a transport inhibitor. Addition of (S)-4-CPG prevented cystine supplementation from protecting the cells, although NAC was able to do so even in the presence of the inhibitor (Figure 6c). NAC is cell permeable and does not require a transporter, and is not affected by blocking system ${\text{x}}_{\text{c}}^{-}$. These data demonstrate the critical role of cystine/cysteine in protection of cells from asbestos toxicity and emphasizes the importance of system ${\text{x}}_{\text{c}}^{-}$ in that protection.

The fact that both LPS and asbestos upregulate system ${\text{x}}_{\text{c}}^{-}$ may provide clues to the signaling involved. While it is possible that the 6-Mix asbestos used in this study was contaminated LPS, it was not detectable by PyroGene® Recombinant Factor C endotoxin assay (Cambrex, Walkersville, MD, USA), which is sensitive to 0.01 EU/ mL (data not shown). LPS signals through a complex that includes the toll-like receptor 4 (TLR-4), and also through non-receptor mechanisms. Both silica and asbestos are putative “ligands” for pattern recognition receptors such as scavenger receptors and TLRs (Hamilton et al., 2008; Resnick et al., 1993). While crystalline silica has been shown to signal through scavenger receptors such as MARCO (Hamilton et al., 2006), amphibole asbestos may signal through TLR4 (Pfau, unpublished data). Interestingly, others have reported effects that are similar for amphibole asbestos and LPS, but different for silica, such as induction of iNOS and of superoxide anion production through protein kinase C by asbestos and/or LPS but not silica (Lim et al., 1997; Quinlan et al., 1998), and induction of tyrosine phosphorylation by silica (Holian et al., 1994). These differences remain to be confirmed and clarified, and the current study provides a new pathway where differences in the effects of silica and asbestos may lead to a much better understanding of the molecular mechanisms that govern the ultimate outcomes of exposure.

Conclusions

In conclusion, system ${\text{x}}_{\text{c}}^{-}$ is functionally upregulated in cultured macrophages by amphibole asbestos exposure, apparently in response to oxidative stress. The uptake of cystine is associated with decreased oxidative stress and increased survival of the macrophages. Silica exposure did not induce system ${\text{x}}_{\text{c}}^{-}$, suggesting that silica-exposed cells may use alternate strategies for dealing with the oxidative stress of the exposure. Therefore, system ${\text{x}}_{\text{c}}^{-}$ may play an integral role in regulating immune outcomes following asbestos exposure by enhancing macrophage survival. More research is needed to determine the specific down-stream events impacted by system ${\text{x}}_{\text{c}}^{-}$, particularly cytokine expression that might dysregulate the overall immune response. Use of siRNA to selectively down-regulate xCT is an approach being explored for this purpose.

No previous study has specifically explored the role of system ${\text{x}}_{\text{c}}^{-}$ in macrophages exposed to asbestos, nor compared this with responses to silica. Therefore, this novel data reveals a mechanism through which cells trigger survival mechanisms and downstream signaling cascades that can affect ultimate outcomes of exposure. Because oxidative stress is an important component of the pathology resulting from asbestos exposure in humans, system ${\text{x}}_{\text{c}}^{-}$ may be an attractive therapeutic target for early intervention in asbestos-induced disease.