Cell-Cycle Changes and Oxidative Stress Response to Magnetite in

Article pubs.acs.org/crt

Cell-Cycle Changes and Oxidative Stress Response to Magnetite in A549 Human Lung Cells Mathias Könczöl,*,†,‡ Adilka Weiss,† Evi Stangenberg,† Richard Gminski,† Manuel Garcia-Kaü fer,† Reto Gieré,§ Irmgard Merfort,‡ and Volker Mersch-Sundermann† †

Department of Environmental Health Sciences, University Medical Center Freiburg, Freiburg, Germany Department of Pharmaceutical Biology and Biotechnology, University of Freiburg, Freiburg, Germany § Institute of Earth and Environmental Sciences, University of Freiburg, Freiburg, Germany ‡

S Supporting Information *

ABSTRACT: In a recent study, magnetite was investigated for its potential to induce toxic effects and influence signaling pathways. It was clearly demonstrated that ROS formation leads to mitochondrial damage and genotoxic effects in A549 cells. On the basis of these findings, we wanted to elucidate the origin of magnetite-mediated ROS formation and its influence on the cell cycle of A549 and H1299 human lung epithelial cells. Concentration- and size-dependent superoxide formation, measured by electron paramagnetic resonance (EPR), was observed. Furthermore, we could show that the GSH level decreased significantly after exposure to magnetite particles, while catalase (CAT) activity was increased. These effects were also dependent on particle size, albeit less pronounced than those observed with EPR. We were able to show that incubation of A549 cells prior to particle treatment with diphenyleneiodonium (DPI), a NADPH-oxidase (NOX) inhibitor, leads to decreased ROS formation, but this effect was not observed for the NOX inhibitor apocynin. Soluble iron does not contribute considerably to ROS production. Analysis of cell-cycle distribution revealed a pronounced sub-G1 peak, which cannot be linked to increased cell death. Western blot analysis did not show activation of p53 but upregulation of p21 in A549. Here, we were unexpectedly able to demonstrate that exposure to magnetite leads to p21mediated G1-like arrest. This has been reported previously only for low concentrations of microtubule stabilization drugs. Importantly, the arrested sub-G1 cells were viable and showed no caspase 3/7 activation.

■

INTRODUCTION Iron oxide magnetite (Fe3O4) is a major anthropogenic constituent of ambient particulate matter (PM) and an important component of many synthetic products.1−3 Besides the increasing presence of magnetite in PM derived from urban and industrial sources, magnetite nanoparticles (MNP) are used as contrast agents and in biomedical research as potential drug carriers.4 In a former study, magnetite was thoroughly investigated for its potential to cause cytotoxic and genotoxic effects, as well as for its influence on signaling pathways.5 There, it was demonstrated that exposure of A549 cells to magnetite micro- and nanoparticles leads to increased ROS formation and mitochondrial membrane depolarization, both of which were almost independent of particle size. Furthermore, the genotoxic effects were reduced by incubating A549 cells with different ROS inhibitors, and it was thus concluded that DNA-damaging effects and micronuclei (MN) induction are ROS-dependent. The ability of particles to trigger ROS formation in biological systems appears to be a key mechanism for their toxic potential.6,7 An imbalance between ROS production and intracellular defense mechanisms can induce oxidative stress followed by further reactions of ROS with components of cell membranes, e.g., lipids or proteins, or DNA.8 Oxidative stress © 2013 American Chemical Society

and secondary failures are known to be involved in the pathogenesis of many diseases, such as cancer, diabetes, respiratory problems, and in aging processes.9 The ROS pathway is initiated by the formation of the superoxide anion via the reduction of molecular oxygen. This process is mediated by enzymes, such as NADPH-oxidase (NOX) and xanthine oxidase, or nonenzymatically by redox-reactive compounds of the mitochondrial electron transport chain.10 After particle uptake, excessive formation of ROS has been shown to occur by activation of NOX or damage to cellular compartments, e.g., mitochondria.11−14 The family of NOX enzymes comprises a membrane-bound complex of oxidoreductases, which catalyzes the formation of the superoxide radical (O2.−) through univalent electron transfer.15 Particle uptake can lead to activation of NOX in keratinocytes and epithelial cells.12,14 Recently, the expression of four NOX isoforms was shown in A549 cells.16 Mitochondria serve not only as a principle supplier of ROS through their major biological function to run ATP synthesis by oxidative phosphorylation17 but also as a main target for cellular ROS. Even under physiological circumstances, electrons “leak” to O2, generating superoxide Received: December 14, 2012 Published: April 22, 2013 693

dx.doi.org/10.1021/tx300503q | Chem. Res. Toxicol. 2013, 26, 693−702

Chemical Research in Toxicology

Article

Table 1. Characteristics of the Investigated Magnetite Particles According to Könczöl et al.5 particles Fe3O4, micro (Bulk) Fe3O4, nano (MNP) a

characteristics euhedral crystals, sharp plates, irregularly broken to slightly rounded particles rounded to slightly elongated

mean diametera

mean diameterb

surface area [m2/g]

0.2−10 μm

0.36−12.21 μm (50% are 98%, nominally 20−60 nm) were supplied by Sigma-Aldrich (Taufkirchen, Germany). To rule out the possibility of endotoxin contaminations causing false-positive results in the assays, the Limulus amoebocyte lysate test was performed. Briefly, particle suspensions of 0.64 (double the highest concentration used for the bioassays), 0.32, and 0.064 mg/mL in deionized water were sonicated for 20 min and centrifuged at 4,000g for 10 min, and the supernatants were tested according to European Pharmacopoeia 7.3. No endotoxin contamination was detected. Both bulk magnetite and MNP were characterized in detail as shown in Table 1 and in our previous study.5 Preparation of Particle Suspensions. Particle suspensions were freshly prepared before each experiment. Particles were suspended in FBS-free culture medium supplemented with 1% L-glutamine and 1% penicillin/streptomycin to a concentration of 5 mg/mL. These suspensions were sonicated for 20 min at 40 °C in an ultrasonic water bath (Sonorex Bandelin, Berlin, Germany) to ensure homogeneous suspensions. Subsequently, the suspensions were diluted in FBS-free culture medium to obtain the required concentrations for each bioassay. Cells were washed with PBS, and culture medium containing 10% FBS was provided before adding the particle suspensions. In each bioassay, the volume of culture medium was adjusted to obtain the same corresponding concentration of particles in μg/mL. The concentrations 10, 50, 100, and 200 μg/cm2 refer to 16, 80, 160, and 320 μg/mL in all bioassays, except the caspase 3/7 assay, where 10 and 100 μg/cm2 refers to 32 and 320 μg/mL. Cell Culture and Cell Treatment. For all experiments, the human lung epithelial cell lines A549 (obtained from Deutsche Sammlung von Mikroorganismen and Zellkulturen GmbH, Braunschweig, Germany) and H1299 (kindly provided by Dr. Malte Kriegs, Hamburg, Germany) were used. A549 cells, derived from an individual with alveolar cell carcinoma,24 have been used extensively to assess type-II cell function because many characteristics of normal type-II cells are retained. For a more thorough investigation of cell cycle-related effects with regard to p53, the H1299 cell line was included. These cells, derived from the lymph node, do not express p53 due to a homozygous partial deletion of the TP53 gene.25 Both cell lines were cultured in RPMI culture medium supplemented with 10% FBS, 1% L-glutamine, and 1% penicillin/streptomycin in a humidified incubator at 37 °C with 5 vol% CO2. For the experiments, cells were trypsinized at 80−90% confluency, seeded into well tissue plates (Greiner, Frickenhausen, Germany), and exposed at 37 °C in a humidified atmosphere containing 5 vol% CO2. Analysis of Intracellular ROS Production. The level of intracellular oxidative stress was measured by EPR. For these experiments, 4 × 105 cells were seeded in 12-well plates, incubated with 2 mL of culture medium for 24 h, and exposed to the particle samples in 2 mL of culture medium containing 5% FBS. After exposure, cells were washed with KHB and incubated for 30 min with 500 μL of KHB containing the spin probe CMH (200 μM). Thereafter, cells were kept on ice until the EPR measurement was performed. EPR (e-scan EPR, model NOX-E.11-ESR, Noxygen

MATERIALS AND METHODS

Chemicals and Reagents. Apocynin and potassium oxide (KO2) were obtained from Alfa Aesar (Karlsruhe, Germany). Diphenyleneiodonium (DPI) was purchased from Enzo Life Science (Lörrach, Germany), and rotenone was obtained from Sigma-Aldrich (Taufkirchen, Germany). Stock solutions were made with DMSO. RPMI 1640 medium was obtained from Invitrogen (Darmstadt, Germany). CAT, L-glutamine, fetal bovine serum (FBS), trypsine/ EDTA, and phosphate-buffered saline (PBS) were purchased from PAA (Parsching, Austria). Propidiumiodide (PI) and DMSO were obtained from Sigma-Aldrich (Taufkirchen, Germany). Krebs−Hepesbuffer (KHB) and 1-hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethyl694



Article

suspensions for 24 h. After adding 100 μL of assay reagent, the plate was incubated for 2 h, followed by luminescence measurement. Camptothecin 10 μM (CPT) was used as the positive control. Particle samples without cells were included in each experiment to rule out the possibility that the observed effects are due to interactions between the magnetite particles and assay components. Each experiment was performed independently in triplicate. Western Blot. The influence of the particles on p53 and p21 protein expression was determined by Western blot analysis. For this purpose, 5 × 105 cells were seeded into 6-well plates in 6 mL of culture medium and incubated for 24 h before the medium was changed to 6 mL of starvation medium without FBS for another 16 h. After exposure in 6 mL of culture medium containing 5% FBS, cells were further processed as described in detail by Könczöl et al.5 Antigen detection was done using antibodies against p53 (Sigma Aldrich, Taufkirchen, Germany), p21 (Cell signaling Technology, Danvers, USA), and β-actin (Sigma Aldrich, Taufkirchen, Germany), and an appropriate horseradish peroxidase-labeled secondary antibody. As detection reagent, ECL advanced chemiluminescence detection reagent (Amersham Pharmacia Biotech, Pittsburgh, USA) was used. Chemiluminescence was measured by using the Analyzer Universal hood (Bio Rad, München) and the Quantify one 4.6.6 Software. Statistical Analysis. All data presented are given as the mean ± standard deviation (SD) or standard error of the mean (SEM) of at least three independent experiments. Differences among treatments compared to solvent controls were evaluated by using either one-way ANOVA followed by Dunnett’s posthoc pairwise comparisons or the unpaired two-tailed Student’s t test with unequal variance as appropriate. A difference was considered significant at p < 0.05.

Science Transfer & Diagnostics GmbH, Elzach, Germany) was carried out using the following parameters: scans, 5; attenuator, 3 dB; sweep time, 5.24 s; and center field, 3460.546. A cell-free CMH sample was used to assess autoxidation of the spin probe. As the positive control, KO2 was added to this CMH sample, leading to complete oxidation of the spin probe. Additionally, all particle samples were incubated in culture medium without cells for 24 h to assess the influence of culture medium and serum on the particles. These results were subtracted from the corresponding cellular samples to describe only the ROS formation derived from A549 cells. Another acellular analysis was carried out where particles were directly incubated with CMH for 30 min in a cell-free environment to elucidate possible interactions between the spin probe and the particles. Further experiments with various ROS inhibitors were carried out to clarify the source of the produced superoxide. Therefore, cells were preincubated with DPI (1, 5 μM), apocynin (100 μM), rotenone (10, 50 nM), or phenathroline (1 μM) for 30 min before exposure to particle suspensions. Each experiment was performed at least three times independently. GSH Level. Analysis of intracellular GSH level was performed using monochlorobimane (MCB) dye with Glutathione Cell-Based Detection Kit (Cayman, Ann Arbor, USA). MCB reacts specifically with GSH to form a fluorescent product.26 For this purpose, 2 × 104 A549 cells were seeded in 200 μL of culture medium in 96-well plates and incubated for 24 h. After washing with PBS, cells were treated with 200 μL of the particle suspensions for 6 and 24 h. The assay was further processed according to the supplier’s instructions. Briefly, the supernatant was discarded, 100 μL of lysis buffer (Cayman, Ann Arbor, USA) was added, and the plate was shaken on an orbital shaker for 15 min at room temperature. The plate was centrifuged at 1000 rpm for 10 min, and 90 μL of the supernatants was transferred to a black 96well plate. After adding 10 μL of MCB solution, the plate was incubated for 2 h at RT. Fluorescence was measured at 384 nm (excitation) and 460 nm (emission). As a positive control, H2O2 (200 μM) was used. Analysis of CAT Activity. The activity of CAT was analyzed according to Barillet et al.27 Therefore, 5 × 105 cells were seeded into 6-well plates in 6 mL of culture medium and incubated for 24 h before exposure to particle samples in 6 mL of culture medium containing 5% FBS. Thereafter, cells were washed and scraped in PBS and collected by centrifugation at 1500 rpm for 5 min at 4 °C before cell lysis in Hepes-buffer containing 1% phenylmethanesulfonyl fluoride by freezing in liquid nitrogen and thawing at 37 °C three times. They were then centrifuged (14000 rpm, 10 min, 4 °C), supernatants were collected, and their protein concentration was measured. CAT activity was measured by adding 50 μL of cell lysate to 10 mM H2O2 and reading absorbance at 240 nm for 2 min. Each experiment was performed independently in triplicate. Cell-Cycle Measurement. Cells (2.5 × 105) were seeded in 2 mL of culture medium in 12-well plates, incubated for 24 h, and exposed to the particle samples in 2 mL of culture medium containing 5% FBS. After exposure, cells were washed with PBS, trypsinized, and collected by centrifugation at 1500 rpm for 5 min at 4 °C. Cells were fixed in 2 mL of ice cold 70% ethanol and kept at 4 °C until analysis. For flow cytometry, cells were centrifuged at 1500 rpm for 10 min and resuspended in PBS containing 0.1 mg/mL RNase and 0.25 mg/mL PI. Cell-cycle distribution was examined by measuring the DNA content using a flow cytometer (FACScalibur, BD biosciences, San Jose, USA). As interference analysis, particle suspensions were added to untreated cells directly before the trypsinization procedure, and the samples were further processed as described. FACS analysis was carried out, and the obtained values were subtracted from the corresponding samples that had been incubated for 24 h with the particle suspensions. More details on this calculation can be found in Supporting Information. Each experiment was performed independently in triplicate. Caspase Assay. Caspase 3/7 activity in A549 and H1299 cells after exposure to magnetite was analyzed by the Caspase-Glo assay (Promega, Mannheim, Germany). Cells (2 ×104) were seeded in white 96-well plates in 200 μL of culture medium and incubated for 24 h. After washing with PBS, cells were treated with 100 μL of the particle

■

RESULTS ROS Production. The DCFH-DA assay is a valuable tool for the measurement of unspecific ROS formation. However, the superoxide radical barely reacts with the dye.28 Therefore, EPR analysis was performed to investigate particle-induced superoxide formation in an acellular environment and in A549. The acellular incubation of CMH with 10 and 100 μg/cm2 Bulk and MNP showed increased oxidation of the spin probe for both particle types (Figure 1). Compared with autoxidation, a pronounced but nonsignificant increase in CMH oxidation was observed at 100 μg/cm2 MNP, whereas the other samples caused only minor changes. After incubation of the particles in culture medium for 24 h before measuring their ROSproducing potential, these effects disappeared completely. In this case, exposure to Bulk and MNP led to a marginal, statistically nonsignificant decrease of the CMH oxidation compared to autoxidation. Comparing the different exposure conditions (direct vs incubation in culture medium), we observed a significant effect (p < 0.05) on the EPR signal for all particle samples except 100 μg/cm2 Bulk. Exposure of A549 cells to magnetite caused size- and concentration-dependent superoxide formation (Figure 1B). Exposure to 10 μg/cm2 Bulk for 24 h led to increased superoxide production (17414 ± 1577 A.U.) compared with the untreated control (7185 ± 2357 A.U.). This effect was more pronounced for 100 μg/cm2 (31932 ± 2738 A.U., p < 0.05). Superoxide production in A549 after incubation with MNP was even higher than that for Bulk. For MNP, the EPR signal increased significantly to 41471 ± 10221 A.U. (10 μg/ cm2, p < 0.01) and 66295 ± 5495 A.U. (100 μg/cm2, p < 0.001). To determine the origin of the radicals, A549 cells were preincubated with various inhibitors of ROS formation before exposure to Bulk or MNP. Pretreatment with rotenone, a complex-I inhibitor of the mitochondrial electron-transfer chain, or the iron chelator phenanthroline did not reduce the 695



Article

Figure 1. (A) Acellular ROS formation measured by EPR spectroscopy for two different exposure scenarios. Black bars show direct incubation of particle samples with CMH (200 μM) for 30 min, gray bars show results after 24-h incubation of Bulk and MNP in RPMI (5% FBS) before 30 min of incubation with CMH (200 μM). After incubation with CMH, 50 μL of supernatant was transferred to a microcapillary and analyzed for CMH oxidation. Samples were kept on ice until measurement. Results are expressed as arbitrary units (A.U.) and are compared to autoxidation of CMH (without particle treatment). For complete oxidation of CMH, a mean value of 5.57 × 106 A.U. was obtained by adding KO2. (B) Cellular ROS formation in A549 cells after exposure to Bulk and MNP for 24 h. Autooxidation of CMH was subtracted from untreated control cells; particlemediated oxidation in acellular environment after 24 h was subtracted from the corresponding cellular samples. All data represent the mean ± SEM from at least three independent experiments; *p < 0.05, **p < 0.01, and ***p < 0.001 versus untreated control by Student’s t test (Figure 1A) or one-way ANOVA followed by Dunnett’s posthoc test (Figure 1B).

Figure 2. ROS formation in A549 cells after preincubation with (A) rotenone (10, 50 nM) and phenanthroline (1 μM), and (B) apocynin (100 μM) and DPI (1, 5 μM) before exposure to Bulk and MNP for 24 h, measured by EPR spectroscopy. After exposure, cells were incubated with CMH (200 μM) for 30 min and kept on ice until measurement. 50 μL of supernatant was transferred to a microcapillary and analyzed for CMH oxidation. Results are expressed as fold change compared to untreated control cells (without particle treatment). Data represent the mean ± SEM from three independent experiments.

EPR signal significantly for cells exposed to Bulk or MNP after 24 h of incubation (Figure 2A). Pretreatment with apocynin, a nonspecific NOX inhibitor, did not reduce ROS formation as observed by EPR spectroscopy (Figure 2B), whereas preincubation with DPI, known to inhibit a wide range of flavoproteins including NOX and parts of the mitochondrial electron-transfer chain,29 led to a pronounced but not significant decrease in ROS production in A549 cells. GSH Status. Analysis of the GSH level in A549 after exposure to Bulk and MNP confirmed that the cells encountered oxidative stress (Figure 3). At 24 h, the GSH level increased notably at 10 μg/cm2 Bulk (123 ± 10%) as compared with the untreated control. Higher concentrations of Bulk led to a decrease in the GSH level, attaining 72% at 200 μg/cm2. Exposure to MNP reduced the GSH level at 100 μg/ cm2 (75 ± 3%), but higher concentrations did not cause an additional effect. Comparing the data of both samples, a sizedependency was observed for concentrations up to 100 μg/

Figure 3. Changes of GSH level induced by Bulk and MNP in human lung cells (A549) upon 24-h treatment. Results are expressed as the percentage of GSH compared to the untreated control. Data represent the mean ± SEM from four independent experiments. ***p < 0.001 versus the untreated control (one-way ANOVA followed by Dunnett’s posthoc test).

cm2, but this difference is leveled at 200 μg/cm2. Taken together, the GSH status had already decreased in a concentration-dependent manner after 6 h of incubation (data not shown). Prolonged exposure for 24 h did not lead to more pronounced effects. The positive control H2O2 (200 μM) reduced the GSH level in A549 cells after 24 h to 25 ± 10% (p < 0.001). 696



Article

Bulk led to an activation of 1.22 ± 0.l3-fold compared to the untreated control, whereas for the same concentration of MNP a CAT activity of 1.34 ± 0.16 was observed. For higher concentrations of Bulk and MNP (100 μg/cm2), CAT activity increased to 1.37 ± 0.09-fold and 1.59 ± 0.13-fold (p < 0.05), respectively. Cell-Cycle Distribution. Cell-cycle measurements were carried out to determine whether the genotoxic effects observed after exposure of A549 cells to magnetite5 are associated with changes in cell-cycle distribution. First, it could be demonstrated once more that fluorescence-based assays can be influenced by the presence of particles, as already stated by various studies.30,31 In our case, Bulk influenced the cell-cycle measurements at 50 μg/cm2, whereas MNP interfered with flow cytometry at 100 μg/cm2. Detailed data on interferences can be found in Supporting Information (Figure S2). To obtain accurate data, the results of the interference analysis were subtracted from the data obtained after particle incubation for 24 h. The experiments revealed a concentration-dependent increase of the sub-G1 population after exposure of A549 cells to magnetite, while the G1 peak decreased concomitantly (Figure 5A). Notably, these changes in cell-cycle progression were not dependent on the size of the particles. To investigate whether these effects observed in A549 cells are influenced by p53, a key mediator of cell-cycle regulation, H1299 cells lacking TP53, were exposed to 100 μg/cm2 of Bulk and MNP. Here, both magnetite samples caused significant effects on the sub-G1

CAT Activity. To determine whether magnetite particles induced changes in cellular defense mechanisms, the activity of CAT in A549 cells was analyzed. Exposure of A549 to Bulk and MNP for 24 h increased the CAT activity in a concentrationand size-dependent manner (Figure 4). Exposure to 10 μg/cm2

Figure 4. Activity of CAT in human lung cells (A549) upon 24-h treatment with Bulk and MNP. Results are expressed as CAT activity per mg protein compared to the untreated control. Data represent the mean ± SEM from four independent experiments. ***p < 0.05 versus the untreated control (one-way ANOVA followed by Dunnett’s posthoc test).

Figure 5. DNA content indicating changes in cell-cycle progression induced by Bulk and MNP in human lung cells. Cells were stained with 0.25 mg/ mL PI for 0.5 h and analyzed by flow cytometry. (A) A549 and (B) H1299 cells after exposure to different concentrations of Bulk and MNP compared to untreated control cells. Data represent the mean ± SD from three independent experiments. (C) Representative histograms of A549 and H1299 cells upon 24-h treatment with 100 μg/cm2 Bulk or MNP showing increased sub-G1 populations. (D) Caspase 3/7 activity in A549 and H1299 cells after 24-h exposure to suspensions containing Bulk or MNP. Data are expressed as caspase activity compared to the untreated control. CPT (10 μM) was used as the positive control. Each bar represents the mean ± SEM of three independent experiments. *p < 0.05 and ***p < 0.001 versus the untreated control (one-way ANOVA followed by Dunnett’s posthoc test). 697



Article

studies mentioned above, we therefore assume that FBScontaining media interact with magnetite particles, leading to inactivation of reactive surfaces. In A549 cells, MNP show a more pronounced effect compared to the same concentrations of Bulk (Figure 1B). Since MNP exhibit a larger specific surface area compared to that of Bulk particles at the same mass concentration, we investigated different dose metrics and plotted the EPR data against the surface area or the total number of the investigated particles (see Supporting Information, Figure S1). Indeed, this approach does not necessarily reflect the conditions in culture medium (diffusion and agglomeration state) but may provide interesting data with regard to the effective exposure metrics of particles. However, we clearly demonstrated that superoxide production in A549 cells after exposure to magnetite is size-dependent. These results are consistent with those of Zhu et al.,33 who observed significant changes in superoxide level at 12.5 μg/cm2 MNP, suggesting that low concentrations of magnetite are sufficient to form superoxide in A549. According to the pathway of ROS production and clearance, superoxide is enzymatically converted to H2O2 and subsequently degraded by GSH and CAT.40 Therefore, we expected this size-dependent effect also for GSH levels and CAT activity in A549 cells after exposure to Bulk and MNP. Alternatively, in the presence of reduced transition metals (e.g., Fe2+), H2O2 may be converted into the highly reactive hydroxyl radical, which can be measured with the DCFH-DA assay.41 Interestingly, compared to the significant effects observed for the superoxide formation in A549 cells (Figure 1B), the size-dependency for the GSH levels and CAT activity after 24-h treatment with Bulk and MNP was considerably less pronounced (Figures 3 and 4). Furthermore, no size dependency was seen for magnetite-mediated ROS production in the DCFH-DA assay in our previous study, where all magnetite samples exhibited similar ROS-production patterns.5 The reason for this lack of size dependency may be addressed by the following explanations. First, it is known that DCFH-DA reacts poorly with superoxide and H2O2 and that conversion of these ROS leads to the generation of species that can be detected by this assay.28 The presence of superoxide in combination with catalytically active molecules or surfaces may induce the Haber−Weiss reaction, which leads to the formation of hydroxyl radicals. Assuming that this radical formation still is size-dependent, we would have expected to illustrate this effect in the DCFH-DA assay. Another explanation could be the upregulation of antioxidative mechanisms after exposure to the particles, referred to as the hierarchical oxidative stress model.42 In this model, small amounts of ROS trigger the upregulation of the transcription factor NRF2, which is critically involved in the induction of endogenous antioxidants such as CAT activity and GSH homeostasis,43,44 which would account for the slight increase in GSH for MNP 10 μg/cm2 (6 h, data not shown) and Bulk 10 μg/cm2 (24 h, Figure 3). EPR spectroscopy revealed a significantly increased superoxide level already at 10 μg/cm2 Bulk or MNP, but the GSH level actually increased, and DCF fluorescence was at the control level for this concentration. As stated above, we assume that the slight increase in ROS formation caused by a lower concentration of magnetite (10 μg/cm2) induced upregulation of antioxidative defense mechanisms, as shown for GSH levels and CAT activity, leading to an efficient degradation of the radicals. As a result, the abundance of ROS may be reduced to levels below detection by DCFH-DA, which would explain why DCF

population in H1299 (Figure 5B), which were more pronounced than in A549 cells (Figure 5C). Similar to the effects in A549 cells, however, the effects in H1299 cells were not dependent on particle size. Caspase 3/7 Assay. An increased sub-G1 peak is often linked to cytotoxic, especially apoptotic, events. Cell viability, however, was not affected significantly.5 Nevertheless, the caspase 3/7 assay was performed to rule out the possibility of the observed effects having been caused by the activation of proapototic pathways. The results from the caspase 3/7 assay clearly demonstrated that no significant caspase activation occurred in A549 or H1299 cells after exposure to Bulk and MNP (Figure 5D). Protein Levels of p53 and p21. The cell-cycle data suggest that the studied cells are arrested in the G1 phase after exposure to magnetite for 24 h. Therefore, Western blots were performed to investigate the influence of p53 and p21 on the observed effects. A549 cells were treated with Bulk or MNP 100 μg/cm2 up to 24 h. After the investigation of different time points, no significant upregulation of p53 was observed (see exemplary data after 24-h exposure in Figure 6). However, after 24 h of incubation with Bulk and MNP, we observed an activation of p21 in A549 cells, which showed no sizedependency (Figure 6).

Figure 6. Activation of p53 and p21 by Bulk and MNP in A549 cells after 24 h measured by Western blot analysis. Lane 1 shows 16-h treatment with PTX (6 nM) as the positive control, and lane 2 shows untreated cells after 24 h. One representative Western blot is shown.

■

DISCUSSION An increasing number of studies dealing with particle-mediated toxic effects have identified oxidative stress as a key mechanism for consecutive cellular responses.7,32,33 In this study, EPR spectroscopy was performed to assess superoxide production of magnetite in a cell-free environment and in A549 cells. Additionally, the intracellular GSH and CAT status were investigated as sensitive markers for oxidative stress. Acellular ROS measurement showed that incubation of CMH with Bulk and MNP led to a concentration- and size-dependent increase of the EPR signal, suggesting that magnetite has the potential to generate ROS in a cell-free environment. Bulk induced negligible amounts of ROS, which is consistent with the study of Fubini and Mollo,34 whereas the signal for MNP was more pronounced. Interestingly, this effect was not observed when incubating the particles with culture medium containing FBS (Figure 1A). In recent years, studies dealing with particle− protein interactions have become more frequent.35−37 The socalled protein corona not only suppresses toxic effects by coating particle surfaces but also influences particle uptake or signal transduction pathways.38,39 In accordance with the 698



Article

fluorescence did not differ from untreated control cells for these concentrations. For higher concentrations, additional defense mechanisms (superoxide dismutases and peroxidases) may be upregulated in A549 cells.45,46 Especially, various peroxidases are responsible for the rapid intracellular detoxification of peroxides. Indeed, we could demonstrate, that for 100 μg/cm2 Bulk and MNP, the effects on GSH level and CAT activity were influenced in a size-dependent manner. However, compared to the superoxide formation in A549, these regulations may not be sufficient to explain the lack of size dependency in the DCFHDA assay. In this context, Gao et al. observed that, depending on pH, temperature, and size, MNP showed peroxidase-like activity in an acellular assay.47 While this intrinsic activity cannot be observed under conventional culture conditions, maximum activity is reached at pH 3.5−4.5, a particle size of 30 nm, and at significantly increased ROS production, which corresponds to conditions found after the incorporation of MNP into degrading vesicles, such as lysosomes of A549 cells.5 Considering the loss of size dependency during the ROS degradation process, we therefore suggest, that the intrinsic peroxidase-like activity of the MNP is responsible for this effect. Disproportionation of the superoxide radical results in the production of H2O2, which is further degraded by GSH and CAT. Additionally, the acidic environment of the phagolysosomes promotes the peroxidase-like activity of MNP itself, which catalyzes the reaction. Consequently, H2O2, due to its rapid degradation, may not be available for the Haber−Weiss reaction, thus leading to a less pronounced signal in the DCFHDA assay. To characterize the origin of the radical species, EPR spectroscopy was performed with various ROS inhibitors. According to our results, it is unlikely that soluble iron released from the magnetite particles triggered ROS formation since incubation with phenanthroline, an iron chelator, did not reduce the EPR signal caused by Bulk and MNP (Figure 2A). This result is consistent with the study of Könczöl et al.,5 in which exposure to the supernatants of Bulk and MNP did not lead to increased ROS formation measured by the DCFH-DA assay. Mitochondria are known to play a vital and vulnerable role in ROS production.11 Therefore, A549 cells were preincubated with rotenone, a complex-I inhibitor of the mitochondrial electron-transfer chain. This pretreatment did not prevent the cells from increased ROS formation in the EPR experiments (Figure 2A) or in the DCFH-DA assay (data not shown). Furthermore, the TEM observations reported in our previous study suggest that no magnetite particles reached the mitochondria or the nucleus,5 assuming that direct interaction with cell components is unlikely to be the main reason for the increased ROS formation. The same study showed that Bulk and MNP are engulfed by A549 cells via endocytosis and that they are located in phagolysosomes. NOX is a membranebound enzyme, which can be activated by particle uptake as a first-line defense mechanism.12,14 At present, no specific NOX inhibitors exist. Thus, our EPR experiments were performed with apocynin and DPI, the two most commonly used substances known to inhibit NOX.15 Apocynin did not decrease the EPR signal (Figure 2B) but effectively led to a reduced ROS formation as determined by the DCFH-DA assay (data not shown), which may be explained by its ability to increase the GSH level in A549 cells.48 In contrast, DPI was able to reduce the EPR signal in a concentration-dependent manner for both samples tested (Figure 2B). It is of note that DPI not only inhibits NOX but also a wide range of flavoproteins, e.g.,

complex-I proteins in mitochondria.29 In summary, the origin of enhanced superoxide production in A549 still cannot be elucidated clearly. It can only be speculated that endocytosis of magnetite particles may lead to the activation of NOX with subsequent formation of superoxide. Taking into account that magnetite induced moderate mitochondrial membrane depolarization in our previous study5 and that DPI may also interact with the mitochondrial electron-transport chain, we suggest that both sources may be involved in magnetite-mediated superoxide formation. Further studies are needed to clarify these effects in A549 cells. In our preceding study, we demonstrated that ROS formation is able to induce genotoxic effects in A549 cells, as measured by the Comet and MN assays, without showing any sign of increased cell death.5 Therefore, the DNA content was analyzed to determine whether exposure to magnetite particles has an impact on cell-cycle progression of A549 cells. The DNA content of the cells in the G1 phase (2C) is halved compared to that in the G2 phase (4C). Generally, DNA fragmentation, as a result of cell death, would lead to histograms with a sub-G1 population (

Cell-Cycle Changes and Oxidative Stress Response to Magnetite in

Recommend Documents