Abstract
Inorganic arsenic is an immunotoxic environmental contaminant to which millions of humans are chronically exposed. We recently demonstrated that human primary macrophages constituted a critical target for arsenic trioxide (As2O3), an inorganic trivalent form. To specify the effects of arsenic on macrophage phenotype, we investigated in the present study whether As2O3 could regulate the activity of NADPH oxidase, a major superoxide-generating enzymatic system in human phagocytes. Our results show that superoxide levels were significantly increased in a time-dependent manner in blood monocyte-derived macrophages treated with 1 μM As2O3 for 72 h. Concomitantly, As2O3 induced phosphorylation and membrane translocation of the NADPH oxidase subunit p47phox and it also increased translocation of Rac1 and p67phox. Apocynin, a selective inhibitor of NADPH oxidases, prevented both p47phox translocation and superoxide production. NADPH oxidase activation was preceded by phosphorylation of p38-kinase in As2O3-treated macrophages. The p38-kinase inhibitor SB-203580 prevented phosphorylation and translocation of p47phox and subsequent superoxide production. Pretreatment of macrophages with the Rho-kinase inhibitor Y-27632 was found to mimic inhibitory effects of SB-203580 and to prevent As2O3-induced phosphorylation of p38 kinase. Treatment with As2O3 also resulted in an increased secretion of the proinflammatory chemokine CCL18 that was fully inhibited by both apocynin and SB-203580. Taken together, our results demonstrate that As2O3 induced a marked activation of NADPH oxidase in human macrophages, likely through stimulation of a Rho-kinase/p38-kinase pathway, and which may contribute to some of the deleterious effects of inorganic arsenic on macrophage phenotype.
Inorganic arsenic (iAs)3 is an environmental toxicant to which millions of humans are exposed worldwide, mainly through contaminated drinking water (1). Chronic exposure to this metalloid is associated with skin lesions (2), vascular inflammatory diseases (3), and cancers (4). Particularly, epidemiological studies have demonstrated that iAs significantly increases incidence of skin and lung cancers (5), and favors development of atherosclerosis in humans (6). Although iAs is a well-known pro-oxidant metalloid (4), cellular and molecular mechanisms mediating its toxicity remain poorly understood. Recent studies indicate that some chronic effects of iAs might be related to immunotoxic properties.
The immunotoxicity of iAs has been studied mainly using sodium arsenite (NaAsO2) and arsenic trioxide (As2O3); when in aqueous solution, these two trivalent inorganic compounds form As(OH)3, the predominant trivalent specie to which humans are exposed through drinking water (7). iAs was first demonstrated to markedly alter physiology of lymphocytes, not only in vitro and in rodents, but also in chronically exposed humans (8, 9). In addition, we and others have recently demonstrated that monocytes/macrophages constituted critical targets for iAs (10, 11, 12, 13). Indeed, long-term treatment of mice with NaAsO2 was shown to significantly decrease adhesion property and phagocytic activity of splenic macrophages (11). We demonstrated that, in vitro, As2O3 also blocks differentiation of human blood monocytes into functional macrophages by down-regulating NF-κB-related survival pathways (12). Moreover, we found that noncytotoxic concentrations of As2O3 partially dedifferentiate human macrophages into monocytes (13). As2O3-treated macrophages notably display a decreased membrane expression of macrophagic markers and a monocytic-like round cell morphology resulting, at least in part, from sequential activation of a RhoA/Rho-kinase (ROCK) pathway and subsequent reorganization of actin cytoskeleton.
Cellular effects of iAs are frequently related to the increased production of reactive oxygen species (ROS) (4, 14); recent studies have reported that NADPH oxidase could, in part, mediate ROS formation in iAs-treated cells (15, 16). Interestingly, NADPH oxidase is a major superoxide-generating enzymatic system, first described in monocytes/macrophages (17). In such cells, stimulation of NADPH oxidase activity results from translocation of the p47phox, p67phox, and Rac1 cytoplasmic proteins to plasma membranes. These proteins associate with the cytochrome b558 subunit gp91phox and p22phox to form a catalytically active NADPH oxidase complex (18). In phagocytes, relocalization of p47phox/p67phox to the membrane is generally driven by phosphorylation of several serine residues in p47phox (18). Invalidation of the NCF1 gene, which encodes p47phox, or inhibition of serine p47phox phosphorylation, correlates with reduced superoxide production (19, 20). Several kinases have been demonstrated to increase phosphorylation of p47phox serine residues, including the ERK (ERK1/2) and p38 kinase (18, 21). Once produced, superoxides are released either outside or inside the cells and then rapidly converted into hydrogen peroxide (H2O2) (17). Besides their antimicrobial role, it is established that NADPH oxidase-derived superoxides can modify many cellular functions of macrophages, such as adhesion or gene expression (17, 18). This suggests that NADPH oxidase may be a major target of iAs in macrophages, putatively accounting for some of its deleterious effects on macrophagic phenotype. The present study was therefore designed to investigate effects of As2O3 on NADPH-oxidase expression and activity in human primary macrophages. Our results demonstrate that low micromolar concentrations of As2O3 (0.5–1 μM), in the range of iAs blood levels measured in chronically exposed humans (10–60 μg/l) (22, 23), induced phosphorylation and membrane translocation of p47phox and subsequent ROS formation through a ROCK/p38-kinase pathway. In addition, we show that NADPH oxidase-derived ROS were likely involved in metalloid-induced secretion of the chemokine CCL18, but not in As2O3-triggered morphological changes due to actin cytoskeleton reorganization.
Materials and Methods
Chemicals, reagents, and Abs
As2O3, NaAsO2, superoxide dismutase (SOD), catalase, and phalloidin-FITC were purchased from Sigma-Aldrich. Dihydrorhodamine 123 and dihydroethidium were obtained from Invitrogen-Molecular Probes (Interchim). Apocynin, PD-98059, SB-203580, SB-202474, NSC-23766, and Y-27632 were obtained from Calbiochem (VWR). GM-CSF (sp. act. 1.2 × 108 IU/mg) was obtained from Schering Plough. The c-Jun N-terminal kinase inhibitor 1 d-stereoisomer (D-JNKI1) was purchased from Alexis Biochemicals. Rabbit polyclonal Ab against Rac1 and mouse mAbs against p47phox and β-catenin were obtained from Santa Cruz Biotechnology (Tebu-bio). Mouse mAb against p67phox was obtained from BD Pharmingen (BD Biosciences). Rabbit polyclonal Abs against phospho- and total p38 kinase were purchased from Cell Signaling Technology (Ozyme). Detection of phosphoserine residues was performed with a specific rabbit polyclonal Ab obtained from Zymed Laboratories (Clinisciences).
Preparation of human macrophages
PBMC were first isolated from bloody buffy coats of healthy donors through Ficoll gradient centrifugation. Human monocytes were then prepared by a 2-h adhesion step, which routinely obtained >90% of adherent CD14-positive cells as assessed by immunostaining. To obtain macrophages, monocytic cells were next cultured for 6 days with 800 IU/ml GM-CSF, as previously reported (13). Differentiated macrophages were then treated with As2O3 in GM-CSF-free RPMI 1640 medium for the indicated time intervals. In some experiments, macrophages were first pretreated with various inhibitors for 2 h and then exposed to As2O3. Apocynin, PD-98059, NSC-23766, SB-203580, and SB- 202474 were used as stock solution in DMSO. Final concentrations of solvent in culture medium did not exceed 0.2% (v/v); control cultures received the same dose of solvent as for their treated counterparts. Y-27632, SOD, and catalase were dissolved in distilled water. For flow cytometric studies, macrophages were collected after a 15-min incubation at 37°C in PBS supplemented with 100 μM EDTA.
Preparation of mouse macrophages
Male Swiss mice (CERJ) were injected with 0.5 ml of 4% sterile thioglycolate (i.p.) for 4 days; then, peritoneal macrophages were harvested and cultured, as previously described (24). Briefly, thioglycolate-elicited macrophages were prepared by a 2-h adhesion step, which routinely obtained >85% of adherent CD11b-positive cells as assessed by immunostaining (25). Macrophages were next cultured for 48 h in the absence or presence of iAs and analyzed for ROS production. In some experiments, mice were cotreated for 4 days with thioglycolate and 0.9% NaCl (i.p.) or 8 mg/kg/day (i.p.) NaAsO2. After isolation, thioglycolate-elicited macrophages were directly used for flow cytometry analysis.
Measurement of ROS by flow cytometry
Detection of ROS was performed using the nonspecific dihydrorhodamine 123 dye and the superoxide-sensitive dihydroethidium dye (26). Dihydrorhodamine 123 can be oxidized into fluorescent rhodamine 123 by H2O2 and peroxynitrite, the anion formed when NO reacts with superoxide. Dihydroethidium, also called hydroethidine, is selectively oxidized by superoxides into fluorescent ethidium. After treatment with iAs in the absence or presence of inhibitors, macrophages were collected, washed, and incubated with 10 μM dihydrorhodamine or 5 μM dihydroethidium for 30 min at 37°C and then analyzed using a FACSCalibur flow cytometer (BD Biosciences). Fluorescence emission from oxidized dye was detected at 525 nm (FL1-H) and 605 nm (FL2-H) for dihydrorhodamine and dihydroethidine, respectively. Each measurement was conducted on 10,000 events and analyzed on Cell Quest software (BD Biosciences). In each experiment, a positive control for rhodamine 123 and ethidium detection was performed by incubating macrophages with H2O2 (1 mM) and menadione (50 μM), respectively. Superposition of control and H2O2 or menadione graphs allowed defining a gate for calculating the percentage of rhodamine- or ethidium-positive cells (27).
Membrane isolation
As2O3-treated macrophages were harvested, sonicated on ice in a buffer containing 100 mM KCl, 3 mM NaCl, 3.5 mM MgCl2, 10 mM HEPES, 1 mM EGTA, 10 μg/ml pepstatin, 10 μg/ml leupeptin, 0.5 mM PMSF; lysates were centrifuged at 600 × g for 10 min at 4°C to remove nuclei and unbroken cells. The supernatant was then ultracentrifuged at 100,000 × g for 1 h at 4°C. Membranes were washed in the same buffer, quantified (28), and resuspended in Laemmli sample buffer, before Western blot analysis.
Western blot immunoassays
After treatment, macrophages were harvested, centrifuged, washed with cold PBS, and lysed, as previously described (13). A total of 50 μg of each sample were heated for 5 min at 100°C, analyzed by 12% SDS-PAGE and electroblotted overnight onto nitrocellulose membranes (Bio-Rad). After blocking, membranes were hybridized with primary Ab overnight at 4°C, washed, and incubated with appropriate HRP-conjugated secondary Ab (DakoCytomation). Immunolabeled proteins were visualized by chemiluminescence. Densitometry data analysis was performed using Bio-Profil BiolD software.
Immunolocalization
Monocytes were first differentiated with GM-CSF into macrophages on glass coverslips for 6 days and then treated with As2O3. After washings, cells were fixed on coverslips with 3% paraformaldehyde in PBS for 30 min at 4°C and washed three times with PBS. Fixed cells were subsequently incubated with a blocking and permeabilizing solution (PBS, 2 mg/ml BSA, 0.2 mg/ml saponin) for 1 h at room temperature. Cells were incubated with mouse anti-p47phox Ab (BD Pharmingen) in a blocking-permeabilizing solution for 2 h at room temperature, washed in PBS, and then stained with rhodamine Red-X-conjugated anti-mouse Ab (Invitrogen) in the same solution for 1 h at room temperature. In some experiments, to detect F-actin filaments, cells were incubated with FITC-phalloidin (1.5 μM) in the same blocking and permeabilizing solution (13). Thereafter, cells were costained by a 15-min incubation in a blocking solution containing 0.25 μg/ml 4′,6′-diamidino-2-phenylindole (DAPI), a fluorescent dye specific for DNA. After washing, coverslips were mounted with PBS-glycerol-DABCO (Sigma-Aldrich). Fluorescent-labeled cells were captured with a DMRXA2 Leica microscope.
Immunoprecipitation
After treatment, macrophages were collected, washed in cold PBS, and lysed for 30 min on ice in a buffer containing 1× TBS (pH 7.6), 1% Triton X-100, 50 mM NaF, 1 mM PMSF, 1 mM sodium orthovanadate, and 1% protease inhibitor mixture. Lysates were then centrifuged twice at 13,000 × g for 10 min at 4°C. Preclearing was achieved on the resulting supernatants with 50 μl of Sepharose 6B (Sigma-Aldrich) for 2 h at 4°C. Then, after protein quantification (28), 1 mg of each sample was incubated overnight at 4°C in the presence of 5 μg of anti-p47phox Ab. The resulting immunocomplexes were precipitated with 50 μl of agarose G plus beads for 2 h at 4°C. After washings, phosphoserine levels were quantified by immunoblotting using a rabbit polyclonal anti-phosphoserine Ab.
Determination of cytokine levels
Levels of TNF-α, IL-8, and CCL18 in the supernatants of macrophage cultures were quantified using Duoset ELISA development system kits obtained from R&D Systems, as previously described (13).
Statistical analysis
The results are presented as means ± SEM. Significant differences were evaluated with the multirange Dunnett’s t test for experiments in which multiple comparisons were studied. Other differences were evaluated with the Student t test. Criterion of significance of the difference between means was p < 0.05.
Results
As2O3 increases ROS production in human macrophages
We first investigated the effects on ROS levels of As2O3 using the nonspecific dye dihydrorhodamine 123. Fig. 1,A shows that during prolonged time cultures, the amount of rhodamine-positive cells in untreated macrophages was only slightly and not significantly altered. In contrast, 1 μM As2O3, a noncytotoxic concentration (13), markedly increased the percentage of positive cells in a time-dependent manner. iAs effect was detected at 16 h and was maximal after 72 h (42.9% of rhodamine-positive cells). To specify the nature of ROS, we next used the superoxide-specific dihydroethidium dye. As observed with dihydrorhodamine 123, 1 μM As2O3 significantly increased the percentage of ethidium-positive cells during a 3-day treatment. A total of 25.1% and 34.5% of positive cells were detected in macrophages exposed to iAs for 16 and 72 h, respectively (Fig. 1,B). Fig. 2,A demonstrates that the As2O3 effect was concentration dependent. To confirm alteration of superoxide levels by iAs, we then tested different antioxidants. Pretreatment of macrophages with SOD alone or SOD plus catalase totally prevented iAs effects (Fig. 2 B); interestingly, apocynin, an inhibitor of NADPH oxidases (29) and NSC-23766, a Rac1 inhibitor (30), also markedly reduced the percentage of ethidium-positive cells in As2O3-treated cultures; thus, these results suggest that NADPH oxidase may be involved in ROS production in human macrophages exposed to As2O3.
As2O3 increases ROS production in human primary macrophages. Cells were cultured in absence or presence of 1 μM As2O3 for indicated times and then incubated at 37°C with 10 μM dihydrorhodamine 123 (A) or 5 μM dihydroethidium (B) for 30 min. After washings, fluorescence related to ROS production was measured using a FACSCalibur flow cytometer. Results are expressed as mean ± SEM of at least four independent experiments. Typical flow cytometric graph obtained with positive controls (H2O2 or menadione; left) or As2O3 (1 μM, 72 h; right) are presented below each bar graph. ∗, p < 0.05 vs control.
As2O3 increases ROS production in human primary macrophages. Cells were cultured in absence or presence of 1 μM As2O3 for indicated times and then incubated at 37°C with 10 μM dihydrorhodamine 123 (A) or 5 μM dihydroethidium (B) for 30 min. After washings, fluorescence related to ROS production was measured using a FACSCalibur flow cytometer. Results are expressed as mean ± SEM of at least four independent experiments. Typical flow cytometric graph obtained with positive controls (H2O2 or menadione; left) or As2O3 (1 μM, 72 h; right) are presented below each bar graph. ∗, p < 0.05 vs control.
ROS production depends on As2O3 concentration and is prevented by antioxidants. A, Macrophages were treated with increasing concentrations of As2O3 for 72 h. B, Macrophages were pretreated or not with SOD (1000 IU/ml), SOD (1000 IU/ml) + catalase (2000 IU/ml), NSC-23766 (25 μM), or apocynin (100 μM), for 2 h and then cultured in the absence or presence of 1 μM As2O3 for 72 h. After treatment, macrophages were incubated with dihydrorhodamine 123 or dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. Results are expressed as means ± SEM of at least four independent experiments. ∗, p < 0.05 vs control.
ROS production depends on As2O3 concentration and is prevented by antioxidants. A, Macrophages were treated with increasing concentrations of As2O3 for 72 h. B, Macrophages were pretreated or not with SOD (1000 IU/ml), SOD (1000 IU/ml) + catalase (2000 IU/ml), NSC-23766 (25 μM), or apocynin (100 μM), for 2 h and then cultured in the absence or presence of 1 μM As2O3 for 72 h. After treatment, macrophages were incubated with dihydrorhodamine 123 or dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. Results are expressed as means ± SEM of at least four independent experiments. ∗, p < 0.05 vs control.
We also determined whether ex vivo or in vivo treatment with iAs could alter ROS production in mouse macrophages. Our results indicate that, in contrast to human macrophages, ex vivo treatment of mouse primary macrophages with non- or subtoxic concentrations (<4 μM) of As2O3 or NaAsO2 did not modify percentage of ethidium-positive macrophages (data not shown); in addition, treatment of mice for 4 days with NaAsO2, at the genotoxic dose of 8 mg/kg/day i.p. (31), did not seem to modify ROS production in peritoneal macrophages.
As2O3 activates NADPH oxidase in human macrophages
To study NADPH oxidase activation in As2O3-treated macrophages, we next analyzed membrane translocation of p47phox, p67phox, and Rac1 by immunoblotting. Fig. 3,A shows that membrane levels of these proteins were markedly increased in As2O3-treated macrophages. Total expression of p47phox in whole cell lysates was however not modified by iAs (Fig. 3,B). Translocation of p47phox from cytoplasm to plasma membrane was further assessed by immunolocalization. In untreated macrophages, expression of this protein appears to be weak and highly diffused, as generally observed in nonactivated cells (Fig. 3,C). Treatment of macrophages with 1 μM As2O3 increased p47phox membrane expression in a time-dependent manner; its relocalization was detectable at 16 h and was maximal after 72 h. As reported in our previous study (13), this figure also shows that As2O3 induced macrophage rounding; this effect was maximal at 72 h. Treatment of macrophages with apocynin prevented As2O3-induced p47phox translocation (Fig. 3,C). We then determined levels of phosphoserine p47phox by immunoprecipitation. Fig. 3 D shows that As2O3 significantly increased the phosphorylation level of serine residues in cells treated for 72 h.
As2O3 induces membrane translocation of p47phox, p67phox, and Rac1 proteins. Macrophages were pretreated or not with 100 μM apocynin for 2 h and then treated with 1 μM As2O3 for 72 h (A, B, and D) or indicated time intervals (C). After preparation of membranes (A) or whole cell lysates (B), protein expressions were analyzed by Western blot. Blots are representative of at least three independent experiments. C, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and were subsequently viewed by fluorescence microscopy (magnification, ×400). Scale bar, 10 μm. Data are representative of at least three independent experiments. D, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals was quantified using Bio1D software and normalized to p47phox; values in C are means ± SEM of three independent results; ∗, p < 0.05 vs control.
As2O3 induces membrane translocation of p47phox, p67phox, and Rac1 proteins. Macrophages were pretreated or not with 100 μM apocynin for 2 h and then treated with 1 μM As2O3 for 72 h (A, B, and D) or indicated time intervals (C). After preparation of membranes (A) or whole cell lysates (B), protein expressions were analyzed by Western blot. Blots are representative of at least three independent experiments. C, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and were subsequently viewed by fluorescence microscopy (magnification, ×400). Scale bar, 10 μm. Data are representative of at least three independent experiments. D, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals was quantified using Bio1D software and normalized to p47phox; values in C are means ± SEM of three independent results; ∗, p < 0.05 vs control.
As2O3-induced NADPH oxidase activation involves p38 kinase
We next analyzed signaling pathways controlling activation of NADPH oxidase in As2O3-treated macrophages by studying involvement of MAPKs which can regulate NADPH oxidase activity in phagocytes (21). Neither the ERK pathway inhibitor PD98059 nor the JNK inhibitor D-JNKI1 could prevent superoxide production in macrophages treated with iAs for 72 h (data not shown). In contrast, increasing concentrations of the specific p38-kinase inhibitor SB-203580 significantly reduced the percentage of ethidium-positive macrophages (Fig. 4,A); the inactive structural analog SB-202474, had however no effect in As2O3-treated cells. We also analyzed phosphorylation of p38 kinase which is necessary for its activation. As shown in Fig. 4,B, As2O3 increased phosphorylated p38-kinase levels in a time-dependent manner; this increase was detectable at 8 h and was maximal after 72 h. No variation was observed for shorter time exposure (data not shown). We further determined involvement of p38 kinase in NADPH oxidase activation by studying its role in phosphorylation and membrane translocation of p47phox. Our results demonstrate that pretreatment with SB-203580 totally inhibited phosphorylation (Fig. 4,C) and membrane relocalization (Fig. 4 D) of p47phox, in macrophages treated with 1 μM As2O3 for 72 h.
p38 kinase mediates phosphorylation and membrane translocation of p47phox and superoxide production induced by As2O3. Macrophages were pretreated or not with SB-203580, or its inactive structural analog SB-202474, for 2 h. Macrophages were next treated or not with 1 μM As2O3 for 72 h (A, C, and D) or indicated time intervals (B). A, After treatment, macrophages were incubated with dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. Results are expressed as means ± SEM of at least four experiments. B, Levels of phospho-p38 kinase (p-p38-kinase) in whole cell lysates were analyzed by Western blot. Equal gel loading and transfer efficiency were checked by protein hybridization with anti-p38 kinase Ab. Experiments were repeated at least three times with similar results. C, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals was quantified using Bio1D software and normalized to p47phox. Values are means ± SEM of three independent results; ∗, p < 0.05 vs control. D, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. Data are representative of at least three independent experiments.
p38 kinase mediates phosphorylation and membrane translocation of p47phox and superoxide production induced by As2O3. Macrophages were pretreated or not with SB-203580, or its inactive structural analog SB-202474, for 2 h. Macrophages were next treated or not with 1 μM As2O3 for 72 h (A, C, and D) or indicated time intervals (B). A, After treatment, macrophages were incubated with dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. Results are expressed as means ± SEM of at least four experiments. B, Levels of phospho-p38 kinase (p-p38-kinase) in whole cell lysates were analyzed by Western blot. Equal gel loading and transfer efficiency were checked by protein hybridization with anti-p38 kinase Ab. Experiments were repeated at least three times with similar results. C, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals was quantified using Bio1D software and normalized to p47phox. Values are means ± SEM of three independent results; ∗, p < 0.05 vs control. D, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. Data are representative of at least three independent experiments.
Inhibition of ROCK prevents As2O3-induced activation of NADPH oxidase
We previously reported that rounding of As2O3-treated macrophages was regulated by the ROCK (13). We thus determined whether this kinase could also be involved in NADPH oxidase activation in As2O3-treated macrophages. Pretreatment of cells with the ROCK inhibitor Y-27632 (32) markedly reduced p47phox phosphorylation (Fig. 5,A), p47phox membrane translocation (Fig. 5,B), and superoxide production (Fig. 5,C). Interestingly, Fig. 5 D shows that Y-27632 also inhibited As2O3-induced phosphorylation of p38 kinase which thus suggests that p38 kinase and NADPH oxidase could be downstream targets of ROCK in iAs-exposed macrophages.
The ROCK inhibitor Y-27632 prevents p38-kinase and NADPH oxidase activation. Macrophages were pretreated or not with the selective ROCK inhibitor Y-27632 (20 μM), and then cultured in the absence or presence of 1 μM As2O3 for 72 h. A, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals were quantified using Bio1D software and normalized to p47phox. B, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. C, After treatments, macrophages were incubated with dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. D, Levels of phospho-p38-kinase (p-p38-kinase) in whole cell lysates were analyzed by Western blot. Equal gel loading and transfer efficiency were checked by protein hybridization with anti-p38-kinase Ab. Results in A and C are expressed as means ± SEM of at least three independent experiments; ∗, p < 0.05. Data in B and D are representative of at least three independent experiments.
The ROCK inhibitor Y-27632 prevents p38-kinase and NADPH oxidase activation. Macrophages were pretreated or not with the selective ROCK inhibitor Y-27632 (20 μM), and then cultured in the absence or presence of 1 μM As2O3 for 72 h. A, Cells were lysed and immunoprecipitated with anti-p47phox Ab. Levels of phosphoserine (p-serine) and p47phox in immunocomplexes were analyzed by Western blot. Intensity of phosphoserine signals were quantified using Bio1D software and normalized to p47phox. B, After fixation and permeabilization, cells were costained with anti-p47phox Ab and DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. C, After treatments, macrophages were incubated with dihydroethidium for 30 min. Fluorescence related to superoxide production was measured using a FACSCalibur flow cytometer. D, Levels of phospho-p38-kinase (p-p38-kinase) in whole cell lysates were analyzed by Western blot. Equal gel loading and transfer efficiency were checked by protein hybridization with anti-p38-kinase Ab. Results in A and C are expressed as means ± SEM of at least three independent experiments; ∗, p < 0.05. Data in B and D are representative of at least three independent experiments.
Apocynin and SB-203580 do not prevent cytoskeleton reorganization but inhibit As2O3-induced CCL18 secretion
We next investigated potential involvement of p38 kinase and NADPH oxidase in cell rounding and cytoskeleton reorganization occurring in As2O3-treated macrophages. Cytoskeleton reorganization was evaluated by immunolocalization of F-actin. As expected, Fig. 6,A indicates that As2O3 (1 μM, 72 h) induced cell rounding and formation of actin ring. Neither apocynin nor SB-203580 could prevent metalloid effect; pretreatment of macrophages with SOD and catalase was also ineffective (data not shown). Thus, these results indicate that NADPH oxidase and ROS are not involved in morphological effects of As2O3. One major feature of macrophages is secretion of various cytokines (33). To determine whether iAs could alter this function, we then analyzed its effect on secretion of TNF-α, IL-8, and the proinflammatory chemokine CCL18. As2O3 altered neither TNF-α (Fig. 6,B) nor IL-8 (data not shown) secretions. In contrast, after a 72 h-treatment, it significantly increased that of CCL18. Interestingly, both apocynin and SB-203580 totally prevented iAs effects (Fig. 6 C); a basal level of CCL18 secretion was by contrast not modified by these inhibitors.
Apocynin and SB-203580 do not prevent F-actin reorganization but inhibit CCL18 secretion induced by As2O3. Macrophages were pretreated or not with 100 μM apocynin or 20 μM SB-203580 for 2 h and then cultured in the absence or presence of 1 μM As2O3 for 72 h (A and C) or indicated time intervals (B). (A) After treatments, cells were fixed, permeabilized, and immunostained with FITC-phalloidin to specifically detect F-actin. Cells were costained with DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. Corresponding macrophage phase-contrast microscopy is also presented (magnification, ×40). Data are representative of three independent experiments. B and C, After treatments, cytokine concentrations in supernatants were measured by ELISA. Results are expressed as means ± SEM of four independent experiments. ∗, p < 0.05 vs control.
Apocynin and SB-203580 do not prevent F-actin reorganization but inhibit CCL18 secretion induced by As2O3. Macrophages were pretreated or not with 100 μM apocynin or 20 μM SB-203580 for 2 h and then cultured in the absence or presence of 1 μM As2O3 for 72 h (A and C) or indicated time intervals (B). (A) After treatments, cells were fixed, permeabilized, and immunostained with FITC-phalloidin to specifically detect F-actin. Cells were costained with DAPI to detect nuclei and subsequently viewed by fluorescence microscopy (magnification ×400). Scale bar, 10 μm. Corresponding macrophage phase-contrast microscopy is also presented (magnification, ×40). Data are representative of three independent experiments. B and C, After treatments, cytokine concentrations in supernatants were measured by ELISA. Results are expressed as means ± SEM of four independent experiments. ∗, p < 0.05 vs control.
Discussion
In this study, we demonstrate for the first time that noncytotoxic concentrations of As2O3 induce a marked activation of NADPH oxidase in human primary macrophages likely through a ROCK/p38-kinase-dependent pathway.
Our results show that As2O3 induced a marked increase of superoxide levels in human blood monocyte-derived macrophages. Several arguments support the conclusion that such ROS may originate from NADPH oxidase: first, As2O3 induced membrane translocation of p47phox, p67phox, and Rac1, a key step in activating the NADPH oxidase complex in macrophages (17, 18); second, translocation of p47phox was totally prevented by apocynin; this plant-derived phenol is an inhibitor of NADPH oxidases which blocks enzymatic complex assembly in leukocytes (29); third, superoxide formation was totally inhibited by apocynin and NSC-23766, a Rac1 inhibitor. Finally, kinetics of ROS production and p47phox translocation were very similar. Unfortunately, such results could not be extended to mouse primary macrophages. Indeed, in contrast to human macrophages, mouse primary macrophages did not respond to similar ex vivo treatment with iAs. In addition, in vivo treatment of mice with iAs did not increase ROS production in macrophages. These results indicate that the mouse model may not be adequate to assess metalloid effects on NADPH oxidase-derived ROS, and thus suggest that interspecies diversity can influence iAs response. This also indirectly strengthens the interest in human primary cultures of macrophages as a model for studying NADPH oxidase, although in vitro findings do not necessarily reflect what occurs in vivo.
Besides phagocytic NADPH oxidase, iAs can also stimulate nonphagocytic NADPH oxidase in endothelial cells (16, 34), vascular smooth muscle cells (35), and in the human promyelocytic NB4 cell line (15); however, molecular mechanisms mediating its activation remain poorly understood. Chou et al. (15) have shown that prolonged treatment of NB4 cells (9 days) with 0.75 μM As2O3 markedly up-regulate mRNA levels and membrane expression of p47phox. In contrast, our results show that p47phox protein levels were similar in whole cell lysates of untreated and As2O3-treated cells; consequently, they suggest that iAs does not alter global NCF1 gene expression in macrophages. Qian et al. (16) have reported that iAs can induce serine phosphorylation of p47phox in vascular endothelial cells but they did not determine which kinase was involved. Our study demonstrates that As2O3 significantly increased the level of serine p47phox phosphorylation in human macrophages likely through p38 kinase. In fact, As2O3 markedly increased levels of phosphorylated p38 kinase. Its effect was detected as soon as 8 h and thus preceded p47phox translocation. In addition, the p38-kinase inhibitor SB-203580 totally inhibited phosphorylation and translocation of p47phox and superoxide production. The demonstration that, in vitro, p38 kinase can phosphorylate p47phox strengthens our hypothesis (36). Our results also suggest that p38-kinase activation in As2O3-treated macrophages involved the serine-threonine kinase ROCK; indeed, the ROCK inhibitor Y-27632 was found to prevent phosphorylation of p38 kinase and to mimic inhibitory effects of SB-203580 on NADPH oxidase activation: it markedly reduced phosphorylation of p47phox, its membrane translocation, and superoxide production. In addition, our previous work demonstrates that a RhoA-ROCK pathway was rapidly activated in response to As2O3; in particular, formation of RhoA-GTP, the active form of RhoA, was found to be maximal at 8 h and consequently to precede increase of the phospho-p38-kinase level. Thus, p38 kinase may be a downstream target of RhoA/ROCK in As2O3-treated macrophages; this idea is concordant with recent demonstrations of RhoA/ROCK-dependent activation of p38 kinase in different cell types, including macrophages (37, 38, 39). Interestingly, Cdc42, another small GTP-binding protein, was found to promote p47phox phosphorylation in iAs-treated vascular endothelial cells (16). Cdc42 is thought to activate NADPH oxidase by triggering actin filament reorganization (16). In As2O3-treated macrophages, neither the actin stabilizer jasplakinolide nor the actin disruptor cytochalasin D could prevent ROS production (A. Lemarie, unpublished data); thus, involvement of actin dynamic in metalloid activation of macrophagic NADPH oxidase is unlikely. Besides RhoA/ROCK, a direct role for ROS in late p38-kinase phosphorylation and subsequent NADPH oxidase activation could not be ruled out because ROS are well-known activators of this redox-sensitive kinase, notably in macrophages (40). In other respects, p47phox contains several cysteine residues which could contribute to a possible redox regulation of NADPH oxidase activity (41).
ROS elicit a wide spectrum of responses that notably depend upon the magnitude of their doses. High doses of ROS rapidly result in necrotic or apoptotic cell death (42). In contrast, intermediate and low doses of ROS directly modulate signaling cascades and can subsequently modify many cellular functions, including cell adhesion (43). As we previously reported that RhoA/ROCK controlled macrophage rounding and F-actin reorganization in response to As2O3 (13), we investigated the potential role of NADPH oxidase-derived superoxides in these phenotypic alterations; our results indicate that neither apocynin nor catalase/SOD could prevent actin ring formation in As2O3-treated macrophages. Thus, the present study does not support a role for ROS in morphological effects induced by iAs; it rather indicates that the RhoA/ROCK pathway, activated by iAs, likely mediates unrelated downstream effects such as NADPH oxidase activation and actin filament reorganization.
ROS regulate expression of several genes, notably proinflammatory genes (17, 44). Although iAs can increase levels of various cytokines in vivo (45), our results clearly demonstrate that As2O3 altered neither TNF-α production nor that of IL-8 in human macrophages. However, we show for the first time that iAs significantly increased secretion of CCL18, a recently described chemokine (46); moreover, both apocynin and SB-203580 inhibited As2O3-induced chemokine production which suggests involvement of NADPH oxidase in this effect. CCL18 is preferentially secreted by human APCs and primarily targets lymphocytes (46). Its production can be increased by allergens and LPS but signaling pathways controlling CCL18 up-regulation remain poorly understood (47). Like ROS, iAs generally regulates gene expression by altering activity of redox-sensitive transcription factors such as AP-1, NF-κB, or Sp1 (4, 14). Interestingly, the proximal promoter of CCL18 contains two putative AP-1 regulatory elements (48); thus, a role for AP-1 in As2O3-induced CCL18 expression could not be excluded.
Increased ROS production in iAs-treated macrophages may contribute to metalloid chronic effects, notably atherosclerosis, cancer, or immunotoxicity. Indeed, macrophages play a major role in development of atherogenesis through low-density lipoprotein oxidation and such effects are, at least in part, mediated by NADPH oxidase-derived superoxides (18, 49). In addition, during chronic inflammation, production of superoxides by recruited macrophages may drive carcinogenesis by damaging neighboring epithelial or stromal cells (50, 51). Moreover, ROS can induce cell proliferation and/or modify DNA methylation (51); these cellular and molecular effects could also been triggered by iAs (52, 53). Finally, up-regulation of CCL18 may contribute to iAs-mediated immunotoxicity, owing to the role played by this chemokine in immunity and allergy (47).
In conclusion, the present study demonstrates that iAs activates human macrophagic NADPH oxidase, a key ROS-generating system, which may contribute to chronic toxicity of this environmental contaminant.
Acknowledgments
We are grateful to Etablissement Français du Sang (Rennes, France) for providing us with blood buffy coats, and express thanks to the Institut Fédératif de Recherche 140 microscopy platform in Rennes. We thank Justin Monnier for helpful advice on manuscript copyediting.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by grants from Association pour la Recherche sur le cancer and Ligue Contre le Cancer (Comité d’Ille-et-Vilaine). A.L. and E.B. are recipients of a fellowship from Ligue Nationale Contre le Cancer and Ligue Contre le Cancer (Comité d’Ille-et-Vilaine), respectively.
Abbreviations used in this paper: iAs, inorganic arsenic; ROCK, Rho kinase; ROS, reactive oxygen species; SOD, superoxide dismutase; DAPI, 4′,6′-diamidino-2-phenylindole.