Role of NF-κB in the oxidative stress-induced lung inflammatory response to iron and selenium at ambient levels

Abstract

Metals enriched in ambient air fine particulate matter (PM_2.5) are thought to contribute to the pathogenesis of PM_2.5-induced inflammatory lung diseases. An important mechanism involved in metal-induced lung injury involves increased oxidative stress due to generation of reactive oxygen species. The redox sensitive transcription factor, nuclear factor kappa B (NF-κB) converts extracellular oxidative stress signals into changes in expression of genes associated with diverse cellular activities. The purpose of this study was to determine the mechanism by which exposure to Fe or Se, at environmentally relevant concentrations, leads to an increased release of chemokines by cultured human lung epithelial cells (A549). We tested the hypothesis that NF-κB signaling pathway is involved in the metal induced IL-8 and MCP-1 release by Fe and Se. Exposure to Fe or Se induced an enhanced release of chemokines at 6 and 24 h, and mediated nuclear translocation of NF-κB. Levels of chemokines in response to Fe were significantly suppressed in the presence of BMS-345541, a specific inhibitor of NF-κB. Similar effects were seen in response to Se, indicating the involvement of NF-κB in the metal-induced chemokine release, while not affecting the AP-1 c-Jun-DNA binding activity. Overall, results indicate that both Fe and Se, at ambient levels, possess the potential for inducing lung inflammation via an oxidative stress pathway in lung epithelial cells.

---

Introduction

Atmospheric particulate matter (PM) constitutes an important component of air pollution and exposure to inhaled PM has been recognized to be causally associated with the induction of inflammatory lung diseases and airway injury.^1–6 Although diverse components contributing to the total ambient PM have been implicated in causing adverse health effects, accumulating evidence suggests a strong relationship between exposure to transition metal constituents of PM and increased incidence of respiratory diseases.^7–9 Epithelial cells lining the respiratory tract are frequent targets for reactive oxygen species (ROS) produced in response to inhaled metal components of atmospheric dust. These cells functionally respond to ROS-exposure by releasing various pro-inflammatory mediators, including cytokines and chemokines.^10,11 Additionally, ROS have been viewed as intracellular molecules that regulate diverse signal transduction cascades leading to changes in expression of certain transcription factors such as nuclear factor kappa B (NF-κB).^12,13

NF-κB is a prototype of a family of eukaryotic transcription factors that play a pivotal role in many cellular responses to diverse environmental changes, leading to up-regulation of gene transcription. The prototypical activated form, a heterodimer of p65/p50 subunits, is widely expressed in various cell types, and is the ubiquitous and biologically active form of NF-κB.^14,15 In most unstimulated cells, NF-κB is sequestered in the cytosol bound in an inactive form to an inhibitory protein called “IκB”.^16,17 In the classical NF-κB signaling pathway, activation of NF-κB is mediated by signal-induced phosphorylation of IκB, followed by ubiquitination, and proteolytic degradation of IκB by the 26S proteasome.¹⁷ Degradation and loss of the inhibitory IκB enables dissociation of the NF-κB-IκB complex, allowing translocation of free NF-κB to the nucleus where it binds to specific regulatory sites in the promoter region on cognate DNA. The promoters of genes encoding for cytokines/chemokines have binding sites for NF-κB.

Among the various metal elements enriching the fine fraction of ambient PM, the redox-active metal iron (Fe) and the essential trace element selenium (Se), have been identified in significant amounts,^18,19 and their toxicities have been shown to involve the generation of intracellular ROS.^20,21 Since exposure to metal rich PM_2.5 is causally linked with lung inflammation, we sought to explore the ability of Fe or Se, at their ambient concentrations, to induce inflammatory effects through expression of chemokines MCP-1 and IL-8 via the regulation of the transcription factor NF-κB. We tested the hypothesis that exposure to environmentally relevant levels of Fe or Se promotes increased NF-κB-DNA binding and up-regulates the NF-κB-driven expression of MCP-1 and IL-8 in A549 epithelial cells.

The primary objective of our study was to examine the role of NF-κB in the induction of these inflammatory mediators by Fe or Se. Although previous studies have examined the effects of these elements on intracellular signaling pathways and their inflammatory responses, none have looked at these effects at their environmentally relevant concentrations. Therefore, an important aspect of our study was to determine whether exposure to Fe or Se, at ambient levels, caused the oxidative activation of NF-κB with the resulting increase in chemokine production.

Materials and methods

All chemicals were reagent grade and used as received from the manufacturer. Roswell Park Memorial Institute Tissue Culture Medium 1640 (RPMI 1640), heat-inactivated fetal bovine serum (FBS), penicillin/streptomycin, L-glutamine, sodium-pyruvate, and 0.25% trypsin-ethylenediaminetetra-acetic disodium salt (EDTA) were purchased from Invitrogen (Grand Island, NY, USA). Trypan blue solution (0.4%), Hank's balanced salt solution 1× (HBSS), hydrogen peroxide solution (30% w/w), ferric citrate, and selenium dioxide were purchased from Sigma (SIGMA-ALDRICH, St. Louis, MO, USA). Stocks of ferric citrate (50 mM) and selenium dioxide (10 mM) were prepared using sterile water (Baxter Corporation), and final working concentrations were diluted from the main stock. The cytokine ELISA kits for MCP-1 and IL-8 were purchased from Quantikine Immunoassay R&D Systems (Minneapolis, MN, USA). The TransAM™ NF-κB p65 Transcription Factor Assay Kits (CAT # 40097) and TransAM™ AP-1 c-Jun Transcription Factor Assay Kits (CAT # 46096) were purchased from Active Motif, CA, USA.

Calculation of Fe and Se exposure concentrations

The environmentally relevant concentrations of Fe and Se used in our study were calculated based on ambient levels of these elements observed in PM_2.5 samples collected in Baltimore, MD in 2002. Briefly, ambient Baltimore air was drawn into a self-contained Semi-continuous Elements in Aerosol System at a rate of 90 liters per minute through an inlet that excluded particulates greater than 2.5 μm in aerodynamic diameter. Concentrations of Fe and Se were determined from collected samples using a Perkin-Elmer simultaneous multi-element electrothermal atomic absorption spectrometer (SIMAA 6000, PE, Norwalk, CT) equipped with a transversely heated graphite atomizer and a longitudinal Zeeman-effect background corrector. Peak concentrations of Fe and Se in air that were associated with hourly PM_2.5 fractions were determined to be approximately 800 ng m⁻³ and 10 ng m⁻³, respectively. In vitro exposure concentrations for the current study were calculated by multiplying ambient peak values by the hourly volume of air collected (5.4 m³), accounting for the molecular weight of each element, and dividing by the total volume of cell culture medium into which air samples were dissolved previously (1 ml). Based on these calculations, Fe and Se were tested in the current study at the relevant environmental concentrations of 77 μM and 0.85 μM, respectively.

Cell culture

Human lung adenocarcinoma cell line (A549) was purchased from American Type Culture Collection (ATCC CCL-185) and maintained in RPMI 1640 supplemented with 10% heat-inactivated FBS, 100 U ml⁻¹ penicillin/streptomycin, 1 mM sodium pyruvate, and 2 mM L-glutamine at 37 °C in a 5% CO₂ atmosphere. Cells were seeded at a density of 0.5 × 10⁶ cells per 500 μl per well in 24-well culture plates.

Measurement of cell viability

Cell viability was estimated using the Alamar Blue assay. Metabolically active cells readily reduce the non-toxic, non-fluorescent dye into a highly fluorescent compound, whereas, non-viable cells, devoid of metabolic ability, fail to reduce the dye and hence the assay is a true measure of viable cells. Briefly, cultured A549 epithelial cells were seeded in 24-well plates (0.5 × 10⁶ cells per 500 μl per well) and exposed to Fe (77 μM) or Se (0.85 μM). Following exposure for 24 h, cell supernatants were removed and replenished with equal volume of cell culture medium containing the Alamar Blue dye (9.1%). Cells were then incubated at 37 °C in 5% CO₂ for 3 h, followed by measurement of the fluorescence intensity using a cytofluor (Multi-well plate reader, Series 4000, PerSeptive Biosystems) at wavelengths 530 nm–580 nm. The percent viability was calculated comparing the absorbance of exposed cultures with that of the non-metal exposed cultures, using the formula:

Dichlorofluorescein (H₂DCF) oxidation assay

Intracellular generation of ROS was measured using the probe 2′-7′-dichlorofluorescein diacetate (DCFH-DA),²² and the assay was performed as described.²³ Briefly, DCFH-DA was mixed with cell culture medium to obtain a final concentration of 50 μM, followed by addition to cells and incubation for 30 min at 37 °C in 5% CO₂. Upon washing twice with HBSS, cells were exposed to medium containing metal solutions (500 μl per well in 24-well plates) for 0, 0.5, 2, 4, 6, and 24 h time intervals. Fluorescence intensity (arising from the oxidation of DCFH-DA to the fluorescent DCF moiety) was quantified at the indicated time points using a Multi-well fluorescent plate reader (Series 4000, PerSeptive Biosystems) at excitation and emission wavelengths of 485 nm and 530 nm, respectively.

Enzyme linked immunosorbent assay (ELISA) for measuring pro-inflammatory mediators

Previously frozen (−70 °C) aliquots of cell culture supernatants treated with Fe or Se as well as the positive control, hydrogen peroxide (H₂O₂), were thawed and levels of IL-8 and MP-1 in the supernatants were analyzed. The quantitative ELISA was performed to measure levels of chemokine according to the manufacturer's instructions (Quantikine Colorimetric Sandwich ELISAs, R&D Systems, Inc.).

Cellular fractionation

Nuclear and cytoplasmic fractions from the treated and non-treated cells were prepared using the Nuclear Extraction Kit (Active Motif, CA, USA). Briefly, cell monolayers (approx. 5–6 × 10⁶ in 100 mm tissue-culture dishes) were washed with ice-cold PBS (Ca⁺² and Mg⁺² free) containing phosphatase inhibitors (provided in kit), removed from the dish by gentle scraping, and transferred to pre-chilled tubes for centrifugation for 7–8 min at 500 rpm (centrifuge pre-cooled at 4 °C). Following centrifugation, the supernatant was discarded and the cell pellet was kept on ice.

Cytoplasmic fraction preparation: cell pellets were resuspended in 1× hypotonic buffer (provided in kit) and transferred to pre-chilled microcentrifuge tubes for 15 min incubation on ice. Cell suspensions were then mixed with detergent (provided in kit) and centrifuged for 30 s at 14 000g (centrifuge pre-cooled at 4 °C). Supernatants were transferred into pre-chilled eppendorf tubes and stored at −80 °C as a source of cytoplasmic fraction. Cell pellet was used for nuclear extract isolation.

Nuclear fraction collection: nuclear pellets were resuspended in 50 μl Complete Lysis Buffer (10 mM DTT, Lysis Buffer AM1, Protease Inhibitor Cocktail) by pipetting up and down and vortexing for 10 s. Suspensions were incubated for 30 min on ice on a rocking platform at 150 rpm. Following a 30 s vortexing step, suspensions were centrifuged for 10 min at 14 000g (centrifuge pre-cooled at 4 °C). Supernatants were then transferred into pre-chilled eppendorf tubes and stored at −80 °C as a source of nuclear fraction. The extracted cytoplasmic and nuclear proteins were quantified in reference to bovine serum albumin standards using Bio-Rad Protein Assay (Bio-Rad Laboratories). The absorbance maximum was measured at a wavelength of 595 nm.

Western blotting

Western blot analyses were performed using nuclear and cytoplasmic extracts from exposed and non-exposed cells. Briefly, protein extracts from Fe or Se exposed and non-exposed A549 cells were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel and electro-transferred to nitrocellulose membrane (BioRad). The membranes were probed with anti-NF-κB p65 (RelA) primary antibody (Cell Signaling Technology, Inc. # 3034) at a 1 : 1000 dilution overnight at 4 °C. Membranes were incubated with goat anti-rabbit horseradish peroxidase-(HRP) conjugated IgG secondary antibody (1 : 5000) for 1 h at room temperature. Protein bands on membranes were visualized by chemiluminescence using ECL plus chemiluminescence detection reagents (GE Healthcare, Amersham # RPN2209) and exposed to X-ray film in dark room. Membranes were stripped (Western blotting stripping buffer, Piercenet Biotechnology, Inc., Rockford, IL), for reprobing with control antibodies (LDH, Lamin B and β-actin, Chemicon). Densitometry of the appropriate band was performed using image quantitator (ChemiDoc XRS, BioRad). Individual treatment values were normalized to β-actin and expressed as a percentage of the control mean.

Transcription factor assay

The activation of DNA-binding protein (NF-κB) and activated protein-1 (AP-1) were measured using the TransAM™ NF-κB p65 (RelA) and the TransAM™ AP-1 c-Jun Transcription Factor Assay Kits, respectively (Active Motif, CA, USA). Briefly, complete binding buffer (CBB) (DTT, Herring sperm DNA, binding buffer AM2) was added to each well containing the NF-κB consensus binding nucleotide sequence (5′-GGACTTTCC-3′) or an oligonucleotide containing a TRE (5′-TGAGATCA-3′) sequence specific for AP-1 c-Jun dimer. Wells dedicated for competitive binding experiments contained 20 pmol of the wild-type or mutated consensus oligonucleotide mixed with CBB (above). Nuclear extracts (7.5 μg ml⁻¹) from treated and non-treated cells previously diluted in Complete Lysis Buffer (CLB) (10 mM DTT, Lysis Buffer AM1, Protease Inhibitor Cocktail), were then added to each well. Following an incubation period of 1 h at room temperature with mild agitation (100 rpm on a rocking platform), wells were washed three times with 1× wash buffer. A primary antibody directed against NF-κB p65 subunit or c-Jun (a major component of AP-1) (1 : 1000 dilution in 1× antibody binding buffer) was added to each well and the plate was incubated for 60 min at room temperature without agitation. A second wash step was then carried out followed by addition of the secondary antibody diluted in HRP (1 : 1000 dilution in 1× Antibody Binding Buffer) for an incubation period of 1 h at room temperature without agitation. Wells were washed again and developing solution (provided in kit) was added to each well for acquiring colorimetric reaction (incubation time 30 s–5 min). The colorimetric reaction was stopped using stop solution, and the optical density was measured using a spectrophotometer (Spectra MAX 250, Molecular Devices) at a wavelength of 450 nm.

Pre-treatment with BMS-345541

In order to elucidate the role of NF-κB in the mechanism of metal-induced cytokine/chemokine production, cultures were treated with the NF-κB inhibitor, BMS-345541N-(1,8-dimethylimidazo[1,2-a]quinoxalin-4-yl)-1,2-ethanediamine hydrochloride (BMS-345541) (50 μM) for 1 h prior to metal exposure. For ELISA studies, supernatants collected from BMS-345541-pre-treated A549 cultures were used. Similarly, in the immunoblotting experiments, cultures were pre-treated with BMS-345541 for 1 h prior to metal stimulation. Extracted cytoplasmic and nuclear lysates were subsequently used for WB analysis. Similarly, nuclear extracts from BMS-345541 pre-treated cultures were used for DNA binding experiments using the TransAM kits as described earlier.

Statistical analyses

All data were expressed as mean ± SD. Individual experiments were performed in triplicates. Simple descriptive statistical data analysis was performed using Microsoft Excel Software (Microsoft version 2000). The software package Instat 2 (Graph Pad Prism, version 5) (Graph Pad Software, San Diego, CA) was used for data that required greater statistical analysis. Significance of interaction among different treatment groups for different parameters at each time point was assessed using one-way analysis of variance (ANOVA) with post hoc Bonferroni test. The value of p ≤ 0.05 was considered as statistically significant.

Results

Generation of intracellular ROS in A549 cells by Fe and Se

The effects of Fe or Se on intracellular oxidation in the A549 cells were assessed using the probe DCFH-DA. Cells were loaded with the probe prior to metal exposure and readings were taken thereafter at intervals from 30 min to 24 h of exposure. The results were expressed as mean DCF-fluorescence intensity values. As shown in Fig. 1A, 24 h exposure of A549 cells to Fe resulted in approximately 4-fold increase in DCF fluorescence compared with untreated control cells incubated in cell culture medium alone (p < 0.001). Moreover, Fe-induced DCF fluorescence readings at 8 h and 24 h were even higher than those observed in the positive control samples treated with H₂O₂. Similarly, exposure to Se (Fig. 1B) induced a significantly greater increase in ROS (2-fold increase at 24 h) compared with non-treated control cells (p < 0.001).

Fig. 1 Increased intracellular oxidative stress in response to treatment with Fe or Se. A549 were loaded with 2′,7′-dichlorofluorescein diacetate (DCFH-DA) for 30 min prior to metal-exposure, (a) Fe (77 μM), or (b) Se (0.85 μM) for up to 24 h. DCF-fluorescence intensity was measured using a cytofluor, and readings were taken at time intervals shown. Each bar represents mean ± SD of three or more separate experiments (n ≥ 3). H₂O₂ (500 μM) was used as a positive control. Asterisks indicate statistically significant increase in DCF fluorescence over media controls (untreated cells) (*** = p < 0.001).

Effect of Fe- or Se on IL-8 and MCP-1 release by A459 epithelial cells

To assess the ability of Fe or Se, at low, non-cytotoxic concentrations, to mediate inflammatory responses in lung epithelial cells, A549 cultures were treated with Fe (77 μM) or Se (0.85 μM) for 6 h and 24 h and the levels of IL-8 and MCP-1 protein in cell-free supernatants were determined by ELISA. As shown in Fig. 2A, incubation with Fe caused a time-dependent IL-8 release by A549 cells. Fe increased IL-8 production over controls (non-treated cells) by ∼4.5-fold (p < 0.05) and 7-fold (p < 0.001) after cell exposure for 6 h and 24 h, respectively. Exposure to Se resulted in a 6-fold increase in the levels of secreted IL-8 compared with medium-treated controls (p < 0.001) at both the time points tested. Results also demonstrate that IL-8 release induced by the positive control tested in this study, H₂O₂ was significantly higher (8-fold) than the non-treated cells (p < 0.001).

Fig. 2 Expression of IL-8 and MCP-1 in A549 epithelial cells. Cells were treated with Fe or Se. Supernatants from treated and untreated cultures were collected at the end of 6 h and 24 h post-exposure. Protein levels in the cell supernatants were determined using ELISA kits as mentioned in Materials and methods. H₂O₂ was used as a positive control. Data are expressed as mean ± SD of three or more individual experiments with each sample (treatment) ran in triplicates. Significant differences between groups were determined by analysis of variance (ANOVA) followed by Bonferroni post-test to compare significance of differences within means. Asterisks indicate statistically significant release of IL-8 or MCP-1 over the controls (untreated cells) (* = p < 0.05; *** = p < 0.001).

Both Fe and Se stimulated release of MCP-1 protein was significantly higher compared with controls (Fig. 2B). Cells exposed to Fe demonstrated ∼5-fold and ∼6-fold increase in the production of MCP-1 over medium-treated control cells at 6 h and 24 h time periods, respectively (p < 0.001). In comparison to H₂O₂, the 6 h reading for Fe was not statistically different (p > 0.05), although this effect at 24 h achieved statistical significance (p < 0.05). Treatment of A549 cells with Se led to ∼4- and ∼5-fold increase at 6 h (p < 0.001), and 24 h (p < 0.001) in MCP-1 release compared with medium-treated controls at both the time intervals tested, respectively.

Requirement of NF-κB in the release of chemokine by Fe or Se

In order to determine if NF-κB indeed plays a role in Fe- or Se-induced chemokine release, A549 cells were pre-incubated with the NF-κB-inhibitor, BMS-345441 (50 μM) for 1 h prior to and until the end of Fe- or Se-exposure. Supernatants were collected at 6 h and 24 h post-stimulation, and IL-8/MCP-1 levels were determined by ELISA. Fig. 3A and B show that addition of BMS-345541 to cell cultures inhibited IL-8 and MCP-1 production by A549 cells treated with Fe or Se, respectively. The mean MCP-1 levels released in Fe-stimulated cells not pretreated with BMS-345541 were noticed to be 4-fold and 3.5-fold higher (p < 0.001) at 6 h and 24 h, respectively, than those in the BMS-345541 pretreated cells exposed to Fe. Similarly, the levels of MCP-1 in Se-treated cells without pre-incubation with BMS-345541 were observed to be approximately 4-fold and 3-fold higher (p < 0.001) at 6 h and 24 h respectively, than those in the BMS-345541 pretreated cells.

Fig. 3 Effect of pretreatment with BMS-345541 on the release of IL-8 and MCP-1 following treatment with Fe or Se in the A549 cells. Cells were pre-treated with BMS-345541 (50 μM) 1 h prior to and until end of treatment with Fe or Se for 6 h or 24 h and supernatants from exposed and non-exposed cultures were analyzed for IL-8 (3a) and MCP-1 (3b) by ELISA. H₂O₂ was used as a positive control. Each histogram represents the mean ± SD of at least three separate experiments. Data with a statistically derived p-value of 0.05 or less was considered to be significant. Asterisks indicate statistically significant increase in IL-8 or MCP-1 release in the metal-exposed cells compared with the BMS-345541 pretreated group exposed to metals (*** = p < 0.001).

Fe or Se induce nuclear translocation of Rel A (p65)

To characterize the role of NF-κB in the metal-induced chemokine release, cytoplasmic and nuclear protein fractions from Fe- or Se-treated and non-treated A549 cells were subjected to immunoblotting using specific antibody against p65 (RelA), a subunit of NF-κB. Fig. 4A shows the presence of the RelA protein band in the cytosolic and nuclear fractions, and rapid accumulation of RelA in the nucleus as early as 0.5 h after exposure to Fe. The nuclear accumulation of RelA reached a plateau at 2 h, and by 6 h post-treatment, levels of nuclear RelA declined to their basal levels that were observed at 0 h. The disappearance of RelA in the cytoplasm at 2 h coincided with a peak increase in the band-intensity in the nucleus at 2 h. Levels of RelA gradually re-accumulated in the cytosol at 6 h, and this observation correlated with a significant reduction in the density of the corresponding nuclear RelA band. The amount of nuclear RelA protein was quantified via densitometry scanning of the chemiluminescence bands from three different experiments. In response to Fe, expression of nuclear RelA at 2 h (peak accumulation) was observed to be 2.5-fold higher than in the non-stimulated control cells. As expected, the cytosolic and nuclear fractions derived from H₂O₂-stimulated cells (positive control) revealed marked nuclear translocation of RelA.

Fig. 4 Time course of activation of NF-κB in response to Fe or Se in A549 epithelial cells. A549 cells were treated with Fe (4a) or Se (4b) at the indicated time points (h). Cytoplasmic and nuclear extracts were prepared as described in Materials and methods. Nuclear protein extracts (40 μg) were fractionated on a 12% SDS-PAGE, and electro-transferred to a nitrocellulose membrane. Western blot analysis was performed using specific antibody against the NF-κB-dimer, RelA. For loading control, we used anti-LDH antibody, anti-lamin B antibody and anti-β antibody. Treatment with H₂O₂ (500 μM) for 1 h served as the positive control. Illustrated blots are representative of three individual experiments (n = 3). Histograms on the right illustrate densitometry scanning results of nuclear RelA levels for respective treatments. Asterisks indicate statistically significant increase in metal-induced NF-κB activation at the respective time points over the controls at time 0 (** = p < 0.01).

Exposure to Se induced an increase in the levels of nuclear RelA over its background levels (untreated control levels) with peak accumulation at 1 h and 6 h (Fig. 4B). Se treatment demonstrated a biphasic pattern of increase in the nuclear band intensity at 1 h and 6 h time intervals. The early rise in nuclear RelA expression observed at 1 h preceded a decrease at 2 h, followed by a second peak at 6 h after treatment. Levels of nuclear RelA at 8 h exposure were observed to be equivalent to the baseline levels shown at 0 h. Densitometry scanning of nuclear RelA band intensity in response to Se was observed to be approximately 2.5-fold higher (at 1 h and 6 h of exposure) than in the non-stimulated control cells.

Immunoblot analyses of cellular and nuclear fractions with antibodies against the cytoplasmic constitutive protein LDH, the nuclear constitutive protein lamin B, and the cytoplasmic-nuclear constitutive protein β-actin were used to confirm the purity of the cytosolic and nuclear fractions and to control for protein loading. As shown in Fig. 4A and B, LDH proteins were not detected in the nuclei, whereas, lamin B band was detected in the nuclear extracts only, indicating that contamination of the two compartments did not occur during separation of the cytoplasmic and nuclear proteins. These results indicate that nuclear translocation of RelA is a specific intracellular signaling event as a result of Fe- or Se-exposure, and could not be attributed to differences in protein loading.

BMS-345541 inhibits Fe- or Se-induced nuclear translocation of NF-κB

To examine a possible role of NF-κB in the mechanism of Fe- or Se-induced IL-8 and MCP-1 release, we used the NF-κB inhibitor, BMS-345541, to suppress nuclear translocation of RelA in the metal-stimulated and unstimulated cultures. A549 cells were pretreated with BMS-345541 for 1 h, followed by cell treatment with Fe or Se. The time course for individual treatment selected for protein extractions were similar to the exposure time periods at which a peak increase in the nuclear expression of NF-κB was observed (as noted from immunoblot experiments). Cytosolic and nuclear extracts were prepared and assayed for the presence of RelA proteins by immunoblotting. As observed in Fig. 5A and B, pretreatment of A549 cells with BMS-345541 prior to exposure to Fe or Se significantly inhibited the translocation of the RelA from the cytosol to the nucleus. Inhibition in the nuclear accumulation of RelA-band coincided with an increase in the RelA cytoplasmic band intensity, implying that RelA was retained in the cytoplasm and there was no translocation of RelA to the nucleus in the presence of BMS-345541.

Fig. 5 Western blot images showing the inhibition of NF-κB pathway by the IKK/NF-κB inhibitor BMS-345541. A549 cells were pre-incubated with BMS-345541 (50 μM) for 1 h and stimulated with Fe (5a) or Se (5b). Cytoplasmic and nuclear extracts were prepared as described in Materials and methods. Extracts were fractionated on a 12% SDS-PAGE, and electro-transferred to a nitrocellulose membrane. Western Blot analysis was performed using specific antibody against the NF-κB-dimer, RelA. For loading control, we used anti-LDH antibody, anti-lamin B antibody and anti-β antibody. Illustrated blots are representative of three individual experiments (n = 3). Histograms on the right illustrate densitometry scanning results of nuclear RelA levels for respective treatments. Asterisks indicate statistically significant increase in the activation of NF-κB in the metal-exposed cultures compared with the metal-exposed cultures pre-treated with BMS-345541 (** = p < 0.01).

Exposure to Fe or Se increase NF-κB-DNA binding in A549 cells

The TransAM transcription factor assay for RelA was conducted to determine whether Fe- or Se-induced IL-8 and MCP-1 expression in A549 cells correlated with the binding of NF-κB to its consensus DNA elements. Nuclear protein extracts isolated from Fe- or Se-treated and untreated A549 cells were incubated with an oligonucleotide containing the NF-κB consensus binding site (5′-GGGACTTTCC-3′) and treated with specific antibody against the NF-κB member, RelA. The exposure time periods at which a peak increase in the nuclear expression of NF-κB was observed (as noted from immunoblot experiments), were considered for assessing the NF-κB-DNA binding activity. As seen in Fig. 6A, low levels of constitutive NF-κB-DNA binding activity were present in extracts from the untreated control A549 cells. NF-κB-DNA binding activity increased over background levels within 2 h following incubation with Fe or Se. Exposure to Fe enhanced the binding activity by ∼1.5-fold (p < 0.01) increase over basal levels that were observed in the medium-treated control cells. Similarly, Se induced a 1.5-fold increase in NF-κB-DNA-binding activity compared with control cells (p < 0.001) at 6 h of exposure. This inducible binding activity was specific as DNA binding was abolished in the presence of wild type (WT) consensus oligonucleotide, which was used as competitor for DNA binding. Conversely, the NF-κB-DNA binding activity was not affected in the presence of the mutated (MUT) consensus oligonucleotide (Fig. 6A and B).

Fig. 6 NF-κB-DNA binding activity in A549 cells exposed to Fe and Se at ambient concentrations. Nuclear extract from Fe or Se treated and non-treated cells was incubated with an oligonucleotide containing the NF-κB consensus binding site (5′-GGGACTTTCC-3′). The antibody against the p65 (RelA) subunit of NF-κB protein was added to confirm protein nature of DNA–protein complexes. H₂O₂ was used as a positive control. Each histogram represents the mean ± SD of at least three separate experiments. 6b and 6c illustrate the characterization of NF-κB complexes in the oligonucleotide competition assays. The wild type (WT) consensus oligonucleotide was used as a competitor for DNA binding and prevented NF-κB binding to the probe. The mutated (MUT) consensus oligonucleotide showed no effect on Fe- or Se-induced NF-κB binding. Asterisks indicate statistically significant NF-κB-DNA binding activity over the controls (untreated cells) (** = p < 0.01; *** = p < 0.001).

Requirement of NF-κB in Fe or Se induced chemokine release in A549 cells

To assess whether effect of Fe or Se on NF-κB-DNA binding activity was specifically related to NF-κB activation and to confirm the participation of NF-κB in inducing this response, nuclear protein were isolated from Fe- or Se-treated A549 cells pretreated with or without BMS-345541, and the NF-κB-DNA binding activity was measured. BMS-345541 by itself had no effect on cell viability as tested by the Alamar Blue assay for cytotoxicity (data not shown). NF-κB-DNA binding activity in the presence of BMS-345541 in the metal challenged cultures was equal to the baseline binding activity as seen in the unexposed control cells (Fig. 7). BMS-345541 blocked the NF-κB-DNA binding activity significantly in cells exposed to Fe and Se as compared with Fe- or Se-treated cells not pretreated with BMS-345541 (p < 0.001). Additionally, the increase in NF-κB-DNA binding activity in response to H₂O₂ stimulation was effectively inhibited in cells preincubated with the NF-κB inhibitor, BMS-345541.

Fig. 7 Effects of BMS-345541 on Fe- or Se-induced increase in NF-κB-DNA binding activity in A549 cells. A549 cells were pre-incubated with BMS-345541 (50 μM) prior to metal exposure. Nuclear extract from Fe or Se treated and non-treated cells was incubated with an oligonucleotide containing the NF-κB consensus binding site (5′-GGGACTTTCC-3′). The antibody against the p65 (RelA) subunit of NF-κB protein was added to confirm protein nature of DNA–protein complexes. H₂O₂ was used as a positive control. Each histogram represents the mean ± SD of at least three separate experiments. Asterisks in indicate statistically significant reduction in NF-κB-DNA binding activity in metal exposed cells pretreated with BMS-345541 compared with cells exposed to metals (Fe or Se) alone (** = p < 0.01; *** = p < 0.001).

Increase in AP-1 c-Jun-DNA binding activity is not inhibited by BMS-345541

To confirm that inhibitory effects of BMS-345541 were specific to NF-κB response, A549 cells were pretreated with or without BMS-345541, and exposed to Fe or Se, and nuclear extracts were prepared. After incubation of nuclear extracts with an oligonucleotide probe containing a TRE binding site (5′-TGAGTCA-3′), DNA-bound c-Jun was detected by staining with c-Jun antibody. As shown in Fig. 8A, the c-Jun-DNA binding activity in response to Fe or Se treatment was not significantly different from that in the BMS-345541-pretreated Fe- or Se-exposed cultures or cultures exposed to H₂O₂ (p > 0.05). While exerting a significant inhibitory effect on RelA nuclear translocation and DNA binding (as noticed in Fig. 5A and B), addition of BMS-345541 to cultures prior to and during metal treatment did not affect Fe- or Se-induced c-Jun DNA binding activity in A549 cells. The inducible binding activity was specific as DNA binding was abolished in the presence of wild type (WT) consensus oligonucleotide, which was used as competitor for DNA binding. Conversely, the NF-κB-DNA binding activity was not affected in the presence of the mutated (MUT) consensus oligonucleotide (Fig. 8B and C).

Fig. 8 Effects of BMS-345541 on Fe- or Se-induced increase in AP-1 c-Jun-DNA binding activity in A549 cells. A549 cells were pre-incubated with BMS-345541 (50 μM) for 1 h prior to metal exposure. Nuclear extract was incubated with an oligonucleotide containing a TRE binding site (5′-TGAGTCA-3′). The antibody (phospho-c-Jun) directed against AP-1 c-Jun dimer was added to confirm protein nature of DNA–protein complexes. H₂O₂ was used as a positive control. Each histogram represents the mean ± SD of at least three separate experiments. 8a and 8b illustrate the characterization of AP-1 complexes in the oligonucleotide competition assays. The wild type (WT) consensus oligonucleotide was used as a competitor for DNA binding and prevented AP-1 binding to the probe. The mutated (MUT) consensus oligonucleotide showed no effect on AP-1 binding. Asterisks indicate statistically significant AP-1 c-Jun-DNA binding activity over the controls (untreated cells) (* = p < 0.05; *** = p < 0.001).

Discussion

This study was conducted to assess the involvement of NF-κB pathway in the release of pro-inflammatory mediators in human lung epithelial cells exposed to Fe or Se. We examined the effects of Fe and Se at environmentally relevant concentrations comparable to those that elicit cellular effects in the lung in response to inhalation of ambient metals bound to ambient air particles. While mechanisms of toxicity of these metals applied at high concentrations have been extensively studied in multiple studies,^21,24–28 no information is available regarding whether an alternative mechanism would operate if lung epithelial cells are exposed to Fe and Se concentrations similar to those that would result from inhaling PM_2.5 containing Fe and Se at concentrations present in urban centers such as Baltimore City, MD. The concentrations of Fe and Se used in our study are lower than those that have been shown to induce cytotoxic effects and generation of ROS in response to Fe²⁸ and Se.^29,30 Additionally, we conducted cell viability assay over a 24 h exposure period on A549 cells to ensure that the chosen concentrations of Fe and Se are non-cytotoxic. We observed that Fe did not exert cytotoxic effects on A459 cells over the concentration range tested, whereas, exposure to Se caused a decrease in A549 cell viability in a dose-dependent manner. The selected concentrations for both Fe and Se (77 μM and 0.85 μM, respectively) fall within the non-cytotoxicity dose concentration range (cell viability over 90%). Our data indicate that the transition heavy metal, Fe, and the trace element, Se, at environmentally relevant levels, are non-cytotoxic, but, could affect the cellular oxidant concentrations by increasing generation of intracellular ROS.

We tested the hypothesis that chemokine induction in response to ambient Fe or Se occurs predominantly via the classical NF-κB signaling pathway, and that oxidative stress is involved in the NF-κB-induced release of IL-8 and MCP-1. Data presented in this study demonstrate that stimulation of A549 cells with either Fe or Se resulted in the release of IL-8 and MCP-1 in a time-dependent manner. Expression of these mediators was preceded by the nuclear translocation of NF-κB with subsequent increase in NF-κB-DNA binding activity. The following observations corroborate the role of oxidative stress and implicate the NF-κB signaling pathway in the Fe- or Se-induced IL-8 and MCP-1 release in A549 epithelial cells. First, results from the DCFH-DA assay demonstrated increase DCF fluorescence in response to either Fe or Se at the tested concentration; this is indicative of increased oxidative stress due to generation of ROS in A549 cells. Second, ELISA experiments showed that either Fe or Se promotes increased release of both IL-8 and MCP-1. Third, in the immunoblot experiments, we have shown increased translocation of RelA from the cytosol to the nucleus in cultures stimulated with Fe or Se compared with medium-treated controls. Fourth, pretreatment of Fe- or Se-stimulated cells with the NF-κB-inhibitor, BMS-345541, completely blocked the release of IL-8 and MCP-1. Finally, blockage of NF-κB activation by BMS-345541 resulted in the inhibition of the NF-κB-DNA binding activity.

As detected in the DCFH-DA oxidative stress assay, increase in the dye fluorescence occurred as early as 30 min following exposure to both metals. Although production of ROS in response to both Fe and Se followed a steady ascending trend, the magnitude of fluorescence-increase was observed to be higher for Fe than that for Se at all the time points tested. Similarly as noticed from the immunoblot images the peak increase in nuclear accumulation of NF-κB (RelA) was evident at different time periods of exposure for the two different treatments. Nuclear RelA peaked at 2 h in cultures challenged with Fe, whereas, a biphasic peaking pattern in nuclear RelA accumulation (1 h and 6 h) was evident in response to Se. These observations reflect the apparent differences between the mechanisms of action of each metal to induce generation of free radicals. Fe facilitates the generation of ROS directly via Fenton reaction and hence could induce a strong oxidative burst early on, which possibly explains the peak in nuclear RelA by 2 h. Se, on the other hand, renders a shift in the pro-oxidant/antioxidant balance towards the pro-oxidant side by way of depleting endogenous antioxidant reserves, causing a possible delay in promoting intracellular oxidative stress. The difference in observations could also be attributed to the type of free radical species generated in response to each metal.

Numerous studies indicate that ROS may serve as potential intracellular signaling messengers that can mediate the transmission of signaling cascades,^31–33 ultimately reaching the MAPK and IKK pathways.³⁴ Some of the mechanisms proposed to substantiate the involvement of ROS in NF-κB activation include, H₂O₂-regulated oxidation of cysteine residues of the DNA-binding domain of NF-κB subunits,³⁵ direct activation of the IKK kinase activity by ROS,³⁶ and oxidative stress-induced elevation of intracellular calcium-dependent proteolytic processes,³⁷ which can lead to NF-κB activation causing gene up-regulation and expression of pro-inflammatory cytokines.³⁸ Thus transcriptionally controlled expression of gene products like cytokines and chemokines are modulated by ROS-induced oxidative stress. Moreover, on the basis of inhibitory studies using various antioxidants, previous reports have shown the involvement of free radicals in the activation of the transcription factor.^39–41

To support our hypothesis, we examined the effect of BMS-345541 on the activity of NF-κB in Fe- or Se-exposed cells. BMS-345541 binds to I kappa B kinase (IKK),^42,43 thereby blocking phosphorylation and proteasomal degradation of IκB, and inhibits nuclear translocation of NF-κB and down-regulation of NF-κB-driven genes, including those encoding for MCP-1 and IL-8. Pretreatment with this inhibitor in the metal challenged cultures blocked the nuclear translocation of RelA, causing its retention in the cytoplasmic fraction with an accompanying loss of its appearance in the nucleus. This effect was not demonstrated in cultures challenged with either Fe or Se without the presence of BMS-345541. Additionally, to confirm that inhibitory effect of BMS-345541 was specific to NF-κB responses, we assessed the impact of BMS-345541 on the DNA binding activity of the transcription factor activator protein-1 (AP-1) in cells challenged with metals. AP-1 is a transcription factor implicated in the transcriptional regulation of a wide range of genes including those involved in inflammation and innate immune response.⁴⁴ Since ROS can regulate AP-1-DNA binding activity,⁴⁵ and because AP-1 has been identified as an important component of signal transduction pathways leading to inflammation, AP-1 has been recognized to be involved in the innate immune response to toxicants. We therefore wanted to confirm that the inhibition observed with BMS-345541 specifically targets Fe- and Se-mediated NF-κB activation without affecting induction of AP-1. Our results demonstrate that addition of BMS-345541 to cultures prior to and during metal treatment did not affect Fe- or Se-induced c-Jun DNA binding activity in A549 cells. These observations validate the specificity of BMS-345541 to NF-κB alone and support the argument that pro-inflammatory response to ambient Fe or Se is, in part, due to the involvement of transcription factor NF-κB.

In conclusion, we have demonstrated for the first time the effects of Fe or Se at ambient level concentrations leading to expression of the pro-inflammatory mediators such as IL-8 and MCP-1 in A549 epithelial cells. These data highlight the importance of ambient metals that are implicated as plausible causative constituents in eliciting PM-induced oxidative stress and airway inflammation. Since the fine and ultrafine particles (diameter ≤2.5 μm) represent a significant fraction of airborne PM, and because they have a relatively large surface area, exposure to such particles can lead to high concentrations of transition metals at their sites of deposition. Animal inhalation studies and in vivo observations have shown that transition metals derived from the particle surface play an important role in generating free radicals and mediating the overall lung inflammatory response to exposure to ambient PM.^46–48

At non-cytotoxic concentrations, the effects of Fe and Se may be due to their ability to induce formation of intracellular ROS and activate the classical NF-κB signaling pathway, suggesting that both elements pose the potential for inducing lung inflammation via an oxidative stress pathway. Moreover, our results have demonstrated that Se, like other metals such as Fe, is working through NF-κB pathway to stimulate an inflammatory response, even though its concentrations in PM_2.5 are low compared to other metals. The reliance on NF-κB for the production of chemokine is highlighted based on the observation that the NF-κB-DNA binding activity and the inflammatory response to Fe or Se is remarkably inhibited following addition of BMS-345541 to cell cultures. Our data provide evidence suggesting that NF-κB is involved, at least in part, in the ROS-induced chemokine release following exposure to Fe or Se in lung alveolar epithelial cell line (A549).

Acknowledgements

The authors would like to thank graduate school University of Maryland at Baltimore (UMAB), School of Medicine Program in Toxicology for providing funding support.

References

1. K. Donaldson and W. MacNee, Int. J. Hyg. Environ. Health, 2001, 203, 411–415.
2. L. A. Jimenez, E. M. Drost, P. S. Gilmour, I. Rahman, F. Antonicelli, H. Ritchie, W. MacNee and K. Donaldson, Am. J. Physiol. Lung Cell Mol. Physiol., 2002, 282, L237–L248.
3. G. Oberdorster, R. M. Gelein, J. Ferin and B. Weiss, Inhal. Toxicol., 1995, 7, 111–124.
4. C. J. Obot, M. T. Morandi, R. F. Hamilton and A. Holian, Inhal. Toxicol., 2004, 16, 69–76.
5. J. Schwartz, Environ. Health Perspect., 2001, 109(Suppl 3), 405–409.
6. S. Qu, E. N. Liberda, Q. Qu and L. C. Chen, J. Toxicol. Environ. Health A, 2010, 73, 1113–1121.
7. A. J. Ghio and R. B. Devlin, Am. J. Respir. Crit. Care Med., 2001, 164, 704–708.
8. A. R. Osornio-Vargas, J. C. Bonner, E. Alfaro-Moreno, L. Martinez, C. Garcia-Cuellar, R. S. Ponce-de-Leon, J. Miranda and I. Rosas, Environ. Health Perspect., 2003, 111, 1289–1293.
9. R. P. Schins, J. H. Lightbody, P. J. Borm, T. Shi, K. Donaldson and V. Stone, Toxicol. Appl. Pharmacol., 2004, 195, 1–11.
10. L. D. Martin, T. M. Krunkosky, J. A. Voynow and K. B. Adler, Environ. Health Perspect., 1998, 106(Suppl 5), 1197–1203.
11. C. Monn, R. Naef and T. Koller, Environ. Res., 2003, 91, 35–44.
12. L. Fialkow, Y. Wang and G. P. Downey, Free Radicals Biol. Med., 2007, 42, 153–164.
13. S. G. Menon and P. C. Goswami, Oncogene, 2007, 26, 1101–1109.
14. Q. Li and I. M. Verma, Nat. Rev. Immunol., 2002, 2, 725–734.
15. I. Rahman and W. MacNee, Thorax, 1998, 53, 601–612.
16. S. Ghosh, M. J. May and E. B. Kopp, Annu. Rev. Immunol., 1998, 16, 225–260.
17. M. S. Hayden and S. Ghosh, Cell, 2008, 132, 344–362.
18. E. Kim, P. K. Hopke, D. M. Kenski and M. Koerber, Environ. Sci. Technol., 2005, 39, 4953–4960.
19. J. L. Turi, F. Yang, M. D. Garrick, C. A. Piantadosi and A. J. Ghio, Free Radicals Biol. Med., 2004, 36, 850–857.
20. N. Ercal, H. Gurer-Orhan and N. Aykin-Burns, Curr. Top. Med. Chem., 2001, 1, 529–539.
21. R. R. Ramoutar and J. L. Brumaghim, J. Inorg. Biochem., 2007, 101, 1028–1035.
22. M. M. Tarpey and I. Fridovich, Circ. Res., 2001, 89, 224–236.
23. A. H. Boulares, F. J. Contreras, L. A. Espinoza and M. E. Smulson, Toxicol. Appl. Pharmacol., 2002, 180, 92–99.
24. V. Castranova, L. Bowman, J. R. Wright, H. Colby and P. R. Miles, J. Toxicol. Environ. Health, 1984, 13, 845–856.
25. S. H. Kim, V. J. Johnson, T. Y. Shin and R. P. Sharma, Exp. Biol. Med., 2004, 229, 203–213.
26. L. Letavayova, V. Vlckova and J. Brozmanova, Toxicology, 2006, 227, 1–14.
27. S. J. Stohs and D. Bagchi, Free Radicals Biol. Med., 1995, 18, 321–336.
28. M. Valko, H. Morris and M. T. Cronin, Curr. Med. Chem., 2005, 12, 1161–1208.
29. G. X. Li, H. Hu, C. Jiang, T. Schuster and J. Lu, Int. J. Cancer, 2007, 120, 2034–2043.
30. V. Valdiglesias, E. Pasaro, J. Mendez and B. Laffon, Arch. Toxicol., 2010, 84, 337–351.
31. T. Finkel, IUBMB Life, 2001, 52, 3–6.
32. J. J. Haddad, Crit. Care, 2003, 7, 47–54.
33. H. Nakamura, K. Nakamura and J. Yodoi, Annu. Rev. Immunol., 1997, 15, 351–369.
34. C. Bubici, S. Papa, K. Dean and G. Franzoso, Oncogene, 2006, 25, 6731–6748.
35. W. MacNee, Eur. J. Pharmacol., 2001, 429, 195–207.
36. A. Bowie and L. A. O'Neill, Biochem. Pharmacol., 2000, 59, 13–23.
37. C. K. Sen and L. Packer, FASEB J., 1996, 10, 709–720.
38. D. M. Brown, L. Hutchison, K. Donaldson, S. J. MacKenzie, C. A. Dick and V. Stone, Toxicol. Lett., 2007, 168, 1–6.
39. N. S. Chandel, W. C. Trzyna, D. S. McClintock and P. T. Schumacker, J. Immunol., 2000, 165, 1013–1021.
40. M. Meyer, R. Schreck and P. A. Baeuerle, EMBO J., 1993, 12, 2005–2015.
41. S. Oka, H. Kamata, K. Kamata, H. Yagisawa and H. Hirata, FEBS Lett., 2000, 472, 196–202.
42. J. R. Burke, M. A. Pattoli, K. R. Gregor, P. J. Brassil, J. F. MacMaster, K. W. McIntyre, X. Yang, V. S. Iotzova, W. Clarke, J. Strnad, Y. Qiu and F. C. Zusi, J. Biol. Chem., 2003, 278, 1450–1456.
43. Y. Peng, M. R. Power, B. Li and T. J. Lin, J. Leukocyte Biol., 2005, 77, 975–983.
44. S. Kang, S. Cho, J. H. Chung, C. Hammerberg, G. J. Fisher and J. J. Voorhees, Am. J. Pathol., 2005, 166, 1691–1699.
45. A. R. Murray, E. R. Kisin, C. Kommineni, V. Vallyathan, V. Castranova and A. A. Shvedova, Carcinogenesis, 2007, 28, 1582–1588.
46. F. Schaumann, P. J. A. Borm, A. Herbrich, J. Knoch, M. Pitz, R. P. F. Schins, B. Luettig, J. M. Hohlfeld, J. Heinrich and N. Krug, Am. J. Respir. Crit. Care Med., 2004, 170, 898–903.
47. R. M. Rice, W. C. Robert, J. J. Godleski, E. Al-Mautairi, N. Jian and R. Hauser, Toxicol. Appl. Pharmacol., 2001, 177, 46–53.
48. D. Cherdwongcharoensuk, S. Upathan, A. S. Pereira and A. P. Aguas, Inhal. Toxicol., 2004, 16, 901–909.

---

Footnotes

† Current address: Division of Biology, Center for Devices and Radiological Health (CDRH), Food and Drug Administration, Silver Spring, MD, USA.

‡ Current address: Center for Biologics Evaluation and Research (CBER), Food and Drug Administration, Rockville, MD, USA.

§ Current address: Shire Human Genetic Therapies, 300 Shire Way, Lexington, MA, USA.

---