Diesel exhaust particles (DEP) induce a proinflammatory response in human bronchial epithelial cells (16HBE) characterized by the release of proinflammatory cytokines after activation of transduction pathways involving MAPK and the transcription factor NF-κB. Because cellular effects induced by DEP are prevented by antioxidants, they could be mediated by reactive oxygen species (ROS). Using fluorescent probes, we detected ROS production in bronchial and nasal epithelial cells exposed to native DEP, organic extracts of DEP (OE-DEP), or several polyaromatic hydrocarbons. Carbon black particles mimicking the inorganic part of DEP did not increase ROS production. DEP and OE-DEP also induced the expression of genes for phase I [cytochrome P-450 1A1 (CYP1A1)] and phase II [NADPH quinone oxidoreductase-1 (NQO-1)] xenobiotic metabolization enzymes, suggesting that DEP-adsorbed organic compounds become bioavailable, activate transcription, and are metabolized since the CYP1A1 enzymatic activity is increased. Because NQO-1 gene induction is reduced by antioxidants, it could be related to the ROS generated by DEP, most likely through the activation of the stress-sensitive Nrf2 transcription factor. Indeed, DEP induced the translocation of Nrf2 to the nucleus and increased protein nuclear binding to the antioxidant responsive element. In conclusion, we show that DEP-organic compounds generate an oxidative stress, activate the Nrf2 transcription factor, and increase the expression of genes for phase I and II metabolization enzymes.
- cytochrome P-450 1A1
- NADPH quinone oxidoreductase-1
- antioxidant responsive element
air pollution constitutes a major public health issue in urban countries today since epidemiological studies have shown that particulate atmospheric pollution is associated with an increase of respiratory and cardiovascular mortality and morbidity (17). Diesel exhaust particles (DEP) produced by diesel engines are a major component of particulate atmospheric pollution, especially in urban areas. DEP have a complex structure characterized by a carbonaceous core with adsorbed organic compounds like polyaromatic hydrocarbons (PAH) and quinones. Their small size (0.1-0.3 μm) allows them to penetrate deeply into the respiratory tract and reach the lung alveoles. The main effect observed in healthy human volunteers exposed to DEP is an inflammation characterized by an increase of inflammatory cells and chemokines and immunoglobulin E levels in nose lavages (10), which could account for the epidemiological association between chronic exposure to particulate matter (PM) and the increase of allergic diseases such as asthma and rhinitis (20). However, the cellular and molecular mechanisms underlying this inflammatory reaction, as well as the particle component involved, remain unclear.
In vitro toxicological studies have revealed that macrophages and epithelial cells are the main effectors of this inflammatory response, releasing proinflammatory cytokines involved in the recruitment of inflammatory cells (3, 4, 11). In particular, we have shown in previous studies that DEP induce the release of granulocyte macrophage-colony stimulating factor (GMCSF) in nasal and bronchial epithelial cells (5). This release results from the phagocytosis of DEP and the activation of transduction pathways (MAPK) and transcription factors like NF-κB, which control transcription of the GM-CSF gene (4, 5, 7). Comparison of the respective roles of the carbonaceous core and organic compounds in the molecular responses induced in respiratory epithelial cells has revealed that organic compounds are important for GM-CSF release and MAPK and NF-κB activation (6).
Two main families of organic compounds are adsorbed on DEP: PAH and quinones. PAH could be desorbed from DEP and become available to bind to the cytosolic aryl hydrocarbon receptor (AhR) and induce gene expression-like phase I metabolization enzymes, especially cytochrome P-450 1A1 (CYP1A1). The toxicity of PAH is known to be related to their bioactivation by CYP1A1. Indeed, PAH metabolism produces electrophilic and reactive metabolites, including reactive oxygen species (ROS). Some PAH metabolites can produce DNA adducts responsible for the genotoxicity of DEP if they are not detoxified by phase II metabolization enzymes.
Quinones are known to generate an oxidative stress by redox cycles. They are suspected to be responsible for the production of O2-· and ·OH radicals detected by electron paramagnetic resonance (EPR) in methanol extracts of DEP (23). The quinones could also be involved in indirect free radical production linked to the activity of enzymes like NADPH-cytochrome P-450 reductase (16). Indeed, this enzyme yields semiquinones that by autooxidation produce ROS. The detoxification of quinones occurs by two-electron reduction performed by NADPH-quinone oxidoreductase 1 (NQO-1).
We postulate that the production of ROS by DEP components could be a key event triggering cellular activation leading to the inflammatory response. Indeed, we have shown that antioxidants reduced both the induction of GM-CSF release and the activation of NF-κB by DEP or organic extracts of DEP (OE-DEP) (6). Furthermore, Li et al. (18) have shown that, in macrophages, OE-DEP induced the expression of the antioxidant enzyme heme oxygenase I (HO-1) via the antioxidant responsive element (ARE). Finally, we have recently shown that DEP and OE-DEP induce CYP1A1 gene expression in bronchial epithelial cells; this could lead to the metabolization of these compounds and the release of ROS (6).
The aim of this study was to contribute to a better understanding of the mechanism of action of DEP involving ROS in airway epithelial cells using two cellular models: primary cultures of human nasal turbinates or polyps (nasal cells) and the 16HBE14o-human bronchial epithelial cell line (16HBE cells) (9). First, we assessed ROS production in epithelial cells using fluorescent probes directly detecting either peroxide production or thiol depletion. We investigated the fate of DEP studying 1) the expression of the phase I metabolization enzyme gene CYP1A1 and CYP1A1 activity and 2) the expression of the phase II metabolization enzyme gene NQO-1 and its modulation by antioxidants. Furthermore, activation of the ARE was studied, as this cis-acting regulatory element is known to be activated by ROS and electrophilic compounds and is located in the 5′-flanking region of genes encoding antioxidant enzymes (HO-1) or metabolization enzymes (NQO-1). We present evidence that DEP, used at noncytotoxic concentrations, induced both dose-dependent peroxide production and thiol depletion that are mainly due to their organic component. Organic compounds, especially PAH, are likely to become available, since DEP induce CYP1A1 mRNA expression and increase CYP1A1 enzymatic activity. In addition, NQO-1 mRNA expression was induced by DEP and reduced in the presence of antioxidants. We also demonstrate, for the first time, an increased binding to the ARE of nuclear proteins, including Nrf2, in DEP-treated cells.
MATERIALS AND METHODS
Chemicals and reagents. Diesel PM SRM1650 was purchased from the National Institute of Standards and Technology (Gaithersburg, MD). Carbon black particles (CB) 95 nm in diameter (FR103) were obtained from Degussa (Frankfurt, Germany). Stock solutions of particles were prepared by suspension in a solution of 0.04% dipalmitoyl lecithin (DPL; Sigma, St.-Quentin-Fallavier, France) in distilled water and sonicated two times for 5 min each at maximal power (100 W) (Vibracell; Bioblock Scientific, Illkirch, France). Particles were used at 10, 20, or 30 μg/cm2 equivalent to 50, 100, or 150 μg/ml. Concentrations are preferentially expressed in μg/cm2, since particles rapidly sediment onto the culture. DEP were extracted twice by dichloromethane in a soxhlet apparatus to ensure total extraction. Organic extract recovered from DEP represents 20% of DEP initial mass. It was further dried and redissolved in DMSO (Sigma). The OE-DEP were used at 10, 20, or 30 μg/ml. Six selected PAH present on DEP were used in a DMSO solution at the concentration they are on 10 μg/cm2 DEP: phenanthrene (20 nM), fluoranthene (10 nM), chrysene (3 nM), pyrene (10 nM), benzo(a)pyrene [B(a)p, 0.5 nM], and 1-nitropyrene (4 nM) (Sigma). B(a)p was used at 3 μM in the EMSA, CYP1A1, and immunofluorescence experiments as a positive control. Monochlorobimane (mBBr) and 2′,7′-dichlorofluorescin-diacetate (H2DCF-DA) were purchased from Molecular Probes (Eugene, OR). For H2DCF-DA experiments, a stock solution of tert-butyl hydroperoxide (t-BHP) used as positive control was freshly prepared at 10 mM in DMEM/F-12 (1:1; Invitrogen, Cergy-Pontoise, France). Two antioxidants were used: N-acetylcysteine (NAC) and mannitol. They were dissolved in water and used at 10 mM. The antioxidant enzyme catalase (from bovine liver, activity: 3,400 U/mg protein; Sigma), a scavenger of H2O2, was tested at 1,400 U/ml medium. Antioxidants and catalase were added to cell cultures 20 min before toxic treatment. For mBBr experiments, a stock solution of N-ethyl-maleimide (NEM) was made up as 10 mM in PBS. All others chemicals were purchased from Sigma, except when otherwise specified.
Cell culture and toxic treatment conditions. Dr. D. C. Gruenert (9) (Colchester, VT) kindly provided the human bronchial epithelial cell subclone 16HBE. The cell line was cultured in DMEM/F-12 culture medium supplemented with penicillin (100 U/ml), streptomycin (100 μg/ml), glutamine (1%), fungizone (0.125 μg/ml, Invitrogen), and UltroserG (UG) (2%, Invitrogen). Cells were cultured on collagen (type I, 4 μg/cm2)-coated 25- or 75-cm2 flasks, 6- or 96-well plates (Costar, Cambridge, MA) at 20,000 cells/cm2. At the time of treatment, UG was not added to DMEM/F-12. Human nasal polyps or turbinates obtained from patients undergoing polypectomy or turbinectomy were cultured as previously described by Million et al. (21). Briefly, they were washed in DMEM/F-12 and incubated with 2 mg/ml of pronase (protease XIV) in DMEM/F-12 supplemented with 50 U-50 mg/ml of penicillin-streptomycin at 4°C for 16-20 h under rotary agitation (80 rpm). Ten percent fetal calf serum (FCS) was then added to neutralize the enzyme. After washing, the cell suspension was filtered on a 30-μm-diameter filter and centrifuged at 400 g for 5 min. The supernatant was eliminated, and cells were resuspended in 20 ml of DMEM/F-12. Aggregates were discarded, and dissociated cells were preplated for 2 h at 37°C on plastic dishes (Falcon Merck-Eurolab, Strasbourg, France) to eliminate most contaminating fibroblasts, and epithelial cells were counted. Cells were cultured on six-well plates at 500,000 cells/well for 6 days.
Cultures were incubated in humidified 95% air with 5% CO2 at 37°C. We made controls by using 0.04% DPL for DEP and CB or 0.1% DMSO for OE-DEP and PAH. The different antioxidants tested on the 16HBE cells were added in the culture medium 30 min before treatment with DEP or OE-DEP.
Analysis of intracellular ROS levels. Intracellular ROS levels were assessed with H2DCF-DA, an oxidation-sensitive fluorescent probe. Once inside the cell, this probe is deacetylated by intracellular esterases forming H2DCF, which in the presence of a variety of intracellular peroxides is oxidized to a highly fluorescent compound, 2′,7′-dichlorofluorescein (DCF). Stock H2DCF-DA solution was made at 20 mM in DMSO and stored at -20°C. Before the toxic treatment, cells were loaded for 20 min with 20 μM H2DCF-DA in Hanks' balanced salt solution (HBSS) either for microplate or cytometric fluorescent analysis. Positive control was obtained with t-BHP, and we verified that the toxins do not directly oxidize the probe. All results are expressed in relative units normalized to the control.
Analysis of intracellular thiol levels. Cellular thiol levels were analyzed with mBBr, which forms a fluorescent adduct with sulfhydryl groups. mBBr was made up as a 4 mM solution in 100% ethanol and stored at -20°C. After the toxic treatment, cells were labeled with 40 μM mBBr for 10 min at room temperature in HBSS (Invitrogen) in the dark according to Hedley and Chow (13). After labeling and rinsing, we quantified the fluorescence either with the fluorescent plate reader (Fluostar BMG Labtechnologies, Champigny-sur-Marne, France) using an excitation wavelength of 480 nm and an emission wavelength of 538 nm or with a flow cytometer. In the last case, cells were detached by trypsination, and trypsin action was stopped by addition of 10% FCS (Invitrogen). Cells in suspension were then kept on ice until flow cytometric analysis. Treatment with 0.1 mM NEM for 2 min before mBBr labeling was done as a positive control. All results are expressed in relative units normalized to the control.
DCF and mBBr fluorescence analysis by FACS. Before the fluorescence analysis performed with an EPICS-Elite-ESP flow cytometer (Coultronics-France), cells were labeled with 3 μg of propidium iodide (PI) per milliliter to ascertain viability. A 15-mW air-cooled argon-ion laser tuned at 488 nm was used for DCF fluorescence, and an Innova 90-5 argon-ion laser (Coherent-France) running at 100 mW of output in multilane 351-363 nm mode was used for mBBr fluorescence. DCF, mBBr, and PI fluorescence were collected, respectively, though a 525-, 470-, and 620-nm band-pass filter. Forward-angle light scatter and right-angle scatter was used to select cells.
RNA isolation and Northern blot analysis. Total cellular RNA was isolated from subconfluent cells (10 × 106) cultured in 75-cm2 flasks, by using the Tri reagent according to the manufacturer's instructions. The amount of RNA in aqueous solution was determined by absorbance at 260 nm. Equal amounts (30 μg) of total cellular RNA were separated by size by 0.65 M formaldehyde-agarose (0.8%) gel electrophoresis and transferred onto Hybond-Plus membranes (Amersham Biosciences, Orsay, France) using 10× SSC buffer (1×: 150 mM NaCl, 15 mM sodium citrate). The blot was then baked for 2 h at 80°C and hybridized to specific probes. The blots were prehybridized at 65°C in rapid Hybrid solution (Amersham) and hybridized overnight at 65°C with 1-5 × 107 cpm/ml of 32P-labeled cDNAs. Probes were synthesized from cDNAs with the Megaprime DNA labeling kit (Amersham) according to the manufacturer's instructions. After hybridization, the blots were washed for 30 min at 65°C with 2× SSC, then 1× SSC, then 0.5× SSC, in the presence of 0.1% SDS. The membranes were exposed, and the radioactivity was quantified with PhosphoImager and ImageQuant software (Molecular Dynamics).
Dr. I. De Waziers and Dr. R. Barouki provided the cDNA probes of the CYP1A1 and NQO-1 genes, respectively.
Semiquantitative reverse transcriptase-polymerase chain reaction. We reverse-transcribed 1 μg of total RNA into cDNA using Oligo-dT Primer and Superscript II RNase H- Reverse Transcriptase (Invitrogen).
For semiquantitative experiments, an aliquot of cDNA libraries was amplified by 23 cycles of PCR with NQO-1 oligonucleotides or with GAPDH oligonucleotides in a PTC-100 thermocycler (MJ Research, Watertown, MA). Amplified products were then separated by agarose gel electrophoresis and visualized by ethidium bromide staining.
Evaluation of CYP1A1 activity by ethoxyresorfin-O-deethylase test. Subconfluent primary nasal cells were treated with 10 μg/cm2 DEP or 3 μM B(a)p for 24 h. After being washed with phosphate-buffered saline (PBS, Invitrogen), cells were incubated with DMEM/F-12 containing 5 μM ethoxyresorufin and 2 mM salicylamide. Ethoxyresorufin is metabolized by CYP1A1 in resorufin, which is a fluorescent compound. Kinetic fluorescence measurements were made with a microspectrofluorimeter (Fluostar Galaxy, BMG Labtechnologies) with an excitation wavelength of 530 nm and an emission wavelength of 590 nm for 40 min.
EMSA. After treatment, nuclear extracts were isolated from subconfluent 16HBE cells (3 × 106) cultured in 25-cm2 flasks as described by Staal et al. (25). Cells were washed and removed by scraping in Tris-buffered saline (25 mM Tris · HCl, 136 mM NaCl, 2.7 mM KCl, pH 7.4) and pelleted. The pellets were resuspended in 400 μl of ice-cold hypoosmotic buffer (10 mM HEPES, 10 mM KCl, 2 mM MgCl2, and 0.1 mM EDTA, pH 7.8) supplemented with 0.5 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), 1 μg/ml antipain, 0.3 μg/ml leupeptin, and 0.5 μg/ml pepstatin. Nuclei were spun down at 16,000 g for 30 s after addition of a 10% Nonidet P-40 solution and then resuspended in 40 μl of a hyperosmotic buffer (50 mM HEPES, 50 mM KCl, 300 mM NaCl, 0.1 mM EDTA, and 10% glycerol, pH 7.8) supplemented with 0.5 mM DTT, 0.5 mM PMSF, 1 μg/ml antipain, 0.3 μg/ml leupeptin, and 0.5 μg/ml pepstatin. Nuclear proteins were extracted by incubating the nuclei for 30 min at 4°C, with a slow rotation followed by a 16,000-g centrifugation for 10 min. The supernatants containing the nuclear extracts were stored until use at -80°C. They were complexed with different radiolabeled double-stranded oligonucleotides: 1) human consensus ARE containing the human (h) ARE core sequence of glutathione-S-transferase (GST) or NQO-1 shown in bold letters (Fig. 6) (synthesized by Invitrogen), 2) different mutant hARE with mutations in the hARE core sequence of NQO-1 (hARE mutant 1, hARE mutant 2, hARE mutant 3, Invitrogen; Fig. 7) and activator protein (AP)-1 consensus (Invitrogen) oligonucleotides. For cold competition, 100-fold excess unlabeled probe (hARE, AP-1) was incubated for 15 min before addition of the labeled probe. Shifted complexes were electrophoresed on 5% polyacrylamide gel as described by Baeza-Squiban et al. (2).
Immunocytochemistry. 16HBE cells were grown on slides in DMEM/F-12 medium with or without 10 μg/cm2 DEP, 10 μg/ml OE-DEP or 3 μM B(a)p for 4 h. The cells were then washed with PBS (Invitrogen) supplemented with 0.1% Tween 20 and 3% BSA. The cells were fixed in paraformaldehyde at room temperature for 20 min. After being washed with PBS/Tween 20/BSA, the cells were incubated with rabbit anti-human Nrf2 antibody (100-fold dilution; Santa Cruz Biotechnology) in PBS/Tween 20/BSA overnight at 4°C, and after washing were incubated with FITC-conjugated goat anti-rabbit immunoglobulin antibody (200-fold dilution, Zymed) for 1 h at room temperature. After a last washing with PBS/Tween 20, a 4,6-diamidino-2-phenylindole solution at 0.5 μg/ml in PBS/Tween 20 was placed on the slides for 2 min. The cells were then examined by confocal fluorescence microscopy.
Statistical analysis. All data were expressed as the means ± SE of three cultures (for six-well plates) or of eight cultures (for 96-well plates) from a representative experiment. Means were compared by analysis of variance. The equal variance test is significant with α = 0.05 (P < 0.001). All pairwise multiple comparisons were made with the Student-Newman-Keuls method (t-test, P < 0.05).
DEP and OE-DEP induce ROS production in airway epithelial cells. In both 16-HBE cells (Fig. 1A) and primary cultures of nasal cells (Fig. 1B) treated with DEP at 10, 20, or 30 μg/cm2 for 4 h, an increase of DCF fluorescence intensity was observed, revealing a dose-dependent increase of intracellular ROS levels. Indeed, we verified that there was an increase of fluorescence per cell and not an increase in the number of cells with fluorescence above baseline levels. By contrast, 95-nm-diameter CB, used as a surrogate of the carbonaceous core of DEP and known to be endocytosed by 16HBE cells to the same extent as DEP (4), did not induce an increase of DCF fluorescence intensity at 10 μg/cm2 (Fig. 1C). OE-DEP at 10 μg/ml (the concentration of organic compounds on 10 μg/cm2 DEP) produces an increase of DCF fluorescence intensity similar to the one triggered by native DEP at 10 μg/cm2 (Fig. 1C), suggesting that the organic fraction of DEP is responsible for ROS generation by DEP.
Because PAH are the main organic compounds of DEP, several of them were tested at their relevant concentration (i.e., corresponding to their amount in DEP): phenanthrene (20 nM), fluoranthene (10 nM), chrysene (3 nM), pyrene (10 nM), B(a)p (0.5 nM), and 1-nitropyrene (4 nM). Only phenanthrene and 1-nitropyrene exhibited a statistically significant increase of DCF fluorescence in 16HBE cells (Table 1). The OEDEP-induced increase of DCF fluorescence appears less important when evaluated by microspectrofluorimetry (Table 1) in comparison with flow cytometry (Fig. 1). Indeed, DCF fluorescence was evaluated in the whole culture well with viable and dead cells by microspectrofluorimetry, whereas flow cytometry was used to measure DCF fluorescence cell-by-cell in viable cells (determined by PI incorporation).
The antioxidants NAC (10 mM) and mannitol (10 mM) have no significant effect on the basal DCF fluorescence intensity of 16HBE cells, but they reduce the DCF fluorescence intensity induced by 10 μg/ml of OE-DEP, in particular NAC (Fig. 2A). The same results were obtained with the antioxidant enzyme catalase at 1,400 U/ml (Fig. 2B).
The mBBr probe becomes fluorescent when alkylated by thiols and allowed us to determine the intracellular thiol levels. In 16HBE cells, both DEP and OE-DEP provoke a dose-dependent depletion of intracellular thiols that is smaller than that induced by NEM (0.1 mM), which is known to alkylate all intracellular thiols (Table 2).
DEP and OE-DEP induce CYP1A1 and NQO-1 mRNA expression. The expression of metabolization enzymes was studied by Northern blot experiments of total RNA from 16HBE cells treated with control buffer or DEP (10 μg/cm2), OE-DEP (15 μg/ml), CB (10 μg/cm2), or B(a)p (3 μM), a PAH known to induce the expression of CYP1A1 and NQO-1 genes. A typical time-course study shown in Fig. 3A revealed that CYP1A1 is not expressed in control cultures. CYP1A1 expression appears following 2 h of treatment with DEP or OE-DEP (Fig. 3A). Its expression was clearly increased and maximal at 6 h, then it decreased at 24 h and returned to basal levels after 48 h. This time course is similar to that obtained when B(a)p was used as a positive control. In contrast, CB did not induce CYP1A1 mRNA expression. A similar study on primary culture of human nasal cells (Fig. 3B) treated for 6 h with DEP or OE-DEP confirmed the clear induction of CYP1A1 mRNA by these compounds.
The Northern blots were reprobed with a 32P-labeled cDNA for NQO-1 mRNA (Fig. 3, C and D). Two transcripts were observed that were already present in control cultures. However, their expression increased following 6 h of treatment with DEP, OE-DEP, or B(a)p; remained relatively high at 24 h; and returned nearly to basal levels at 48 h. Similar to CYP1A1, CB had no effect on NQO-1 mRNA expression. The induction of NQO-1 mRNA after 6 h of exposure was also observed in primary cultures of human nasal cells treated with DEP or OE-DEP (Fig. 3D).
By RT-PCR, the increase of NQO-1 mRNA expression induced by DEP was also observed after 24 h of treatment. When cells were cotreated with DEP and the antioxidants (NAC, catalase) (Fig. 4), NQO-1 mRNA expression was reduced and similar to the level of expression in their respective controls.
DEP increases CYP1A1 enzymatic activity. We have shown that DEP induces CYP1A1 mRNA expression. We therefore confirmed CYP1A1 enzymatic activity using the ethoxyresorufin-O-deethylase (EROD) assay. In control nasal cells, no increase of fluorescence was detected for 40 min, but when cells were treated with DEP (10 μg/cm2) or B(a)p (3 μM), a time-dependent linear increase in fluorescence was observed (Fig. 5).
DEP and OE-DEP increase protein binding to ARE. Because DEP and OE-DEP increase ROS production, the binding of proteins to the ARE was evaluated by EMSA. The ARE sequence of the human GST promoter hARE (GST) was first used; it is known to bind predominantly Nrf2, a transcription factor sensitive to oxidative stress. As shown in Fig. 6A, two protein-ARE complexes (C1 and C2) were detected with nuclear protein extracts of 16HBE cells treated with DEP at 30 μg/cm2. These two complexes are specific, since they are displaced by an excess of unlabeled hARE (GST) oligonucleotide. The most shifted complex, C2, is displaced by an excess of AP-1 oligonucleotides, suggesting that Jun or Fos could be involved in this complex. In contrast, the C1 complex is not modified by the AP-1 competitor oligonucleotide and is displaced by unlabeled hARE (GST).
Because NQO-1 expression is induced by DEP treatment and the NQO-1 promoter contains an ARE-like element, gel shift experiments using the ARE of the NQO-1 promoter, hARE (NQO-1), were performed. We also detected several complexes (C1′ and C2′) with the hARE (NQO1) probe (Fig. 6B). The C1′ complex was further characterized using different mutated hARE (NQO-1) probes (Fig. 6). Compared with the wild-type hARE (NQO-1) sequence, mutant 1 displays a modified TGAC sequence, which is required both for a functional ARE and for binding of AP-1. In the case of mutant 2, the AP-1-specific TCA sequence is altered, whereas mutant 3 displays an altered ARE-specific GC site. The binding profiles of the B(a)p-treated 16HBE nuclear extracts for the wild-type labeled hARE (NQO-1) and mutant labeled hARE (NQO-1) oligonucleotides are shown in Fig. 6B. Striking differences in the intensity of the complexes are observed between the different oligonucleotides. The migration and abundance of the C1′ complex were similar in the case of the wild-type and the mutant 2 sequence, whereas the abundance of complex C1′ was lower in the case of mutant 3. Interestingly, no complexes were formed using the mutant 1 probe. Together, these data suggest that although the other complexes are related to the AP-1 complex, the C1′ complex is more specific for the ARE sequence. We studied the effect of DEP and OE-DEP on the abundance of this complex. As shown in Fig. 7, both treatments lead to an increase in C1 and C1′ complex formation with either the GST or the NQO-1 probes. C1 and C1′ are probably similar to the ARE-specific complexes and different from AP-1 complexes. Both are increased in DEP-treated cells.
OE-DEP induces nuclear translocation of the transcription factor Nrf2. To determine whether Nrf2 is regulated by DEP, we studied nuclear translocation of Nrf2. The transcription factor Nrf2 was immunodetected in 16HBE cells treated with either control buffer or 10 μg/cm2 DEP, 10 μg/ml OE-DEP, or 3 μM B(a)p for 4 h. Contrast-phase microscopy of the same cells verified the localization and the integrity of the nuclei. In control cultures, Nrf2 was predominantly found in the cytoplasm (Fig. 8). When the cells were treated with B(a)p (positive control), a clear nuclear translocation of Nrf2 was observed (Fig. 8). Cells treated with DEP or OE-DEP exhibited an intermediate situation: Nrf2 was translocated in the nuclei but to a lesser extent than with B(a)p.
It is now well established that DEP can elicit an inflammatory response in airway epithelial cells (3, 4, 26). The DEP-induced cytokine release requires signaling pathways modulating cytokines gene expression (2, 6). We showed that DEP organic compounds play a critical role in this inflammatory response (6, 22), probably via the production of ROS in bronchial epithelial cells. Hiura et al. (14) reported similar results in macrophages. Furthermore, antioxidants have been shown to reduce GM-CSF release and NF-κB activation induced by DEP and their extracts (6, 7), suggesting that ROS could be a “byproduct” of DEP responsible for the DEP-induced cellular activation. The present study was undertaken to characterize the metabolic pathways induced by the various components of DEP and to analyze the contribution of ROS to these cellular effects.
ROS production was evaluated in human airway epithelial cells exposed to native DEP, OE-DEP, or CB, a model of carbonaceous core, to determine which component is mainly responsible for this production. It clearly appears that only DEP and their organic extracts induced a dose-dependent modification of the redox state of human airway epithelial cells. Previous studies in macrophages show that high concentrations of OE-DEP (100 μg/ml) triggered peroxide production (14, 22). In contrast, the carbonaceous core of 95 nm diameter mimicked by CB did not induce peroxide production, even at high concentrations.
Several additional observations suggest that the organic compounds are most likely the main contributors of ROS production: 1) OE-DEP used at 10 μg/ml produced a similar amount of peroxides as DEP at 10 μg/cm2; 2) some PAHs, known to be present in DEP and used at their real concentration in DEP, can trigger peroxide production that was only significant for phenanthrene and 1-nitropyrene; and 3) cotreatment with antioxidants reduced OE-DEP-induced peroxide production. Different antioxidants were used to determine which ROS were produced and where they were produced. A partial reduction occurs with the extracellular antioxidant mannitol, suggesting that ROS could be generated outside the cells. This is consistent with the demonstration of O2-· and ·OH radical production by a methanol extract of DEP measured by EPR (16, 23) involving quinones. Quinoid redox cycling is also proposed as a mechanism for sustained free radical generation by airborne particulate matter with an aerodynamical diameter <2.5 μm (PM2.5) (24). The proposal is further strengthened by the results obtained with the antioxidant enzyme catalase, which may scavenge extracellular, DEP-produced H2O2. The antioxidant NAC, known to enter cells, inhibits OEDEP-induced peroxide production, suggesting that ROS can enter cells or could be produced in cells. An integrated question is the mechanism by which DEP provoke an oxidative stress within the cells.
We hypothesized that intracellular ROS production could be partially due to PAH metabolism, which requires the CYP1A1 enzyme. Indeed the CYP1A1 catalytic activity is postulated to generate ROS (unpublished data) as well as reactive metabolites such as PAH-O-quinones, which could produce ROS by redox cycles. Evidence for PAH metabolization was indirect and relied on investigating CYP1A1 gene expression, which is specifically induced by PAH. We observed in previous preliminary experiments a clear CYP1A1 mRNA induction in DEP- or OE-DEP-treated 16-HBE cells (6). We have now confirmed this observation on primary cultures of human nasal cells, and we have characterized its time course. Induction starts at 2 h of treatment but is transient and is almost back to basal levels at 24 h. The kinetics are nearly the same for DEP as for OE-DEP and B(a)p, suggesting that organic compounds could be quickly desorbed from DEP. Indeed this desorption is a prerequisite for PAH to interact with the AhR and activate the CYP1A1 gene, leading to an increase in CYP1A1 enzymatic activity. In addition to CYP1A1 mRNA induction, both DEP- and OE-DEP-treated cells exhibited increased expression of NQO-1 mRNA, occurring later than that of CYP1A1. The delayed induction of NQO-1 argues in favor of an indirect mechanism. Indeed, antioxidants partially reduce DEP-induced NQO-1 expression, suggesting that NQO-1 is induced by ROS, which could result either from the induction of CYP1A1 by DEP organic compounds or from the redox cycling of quinones. Thus DEP organic compounds could influence gene expression sequentially, first through AhR activation and then through generation of ROS by phase I metabolizing enzymes.
Together, these data suggest that DEP organic compounds are bioavailable and are probably metabolized, since CYP1A1 enzymatic activity is increased, as we first showed by an EROD test. They support the in vivo observation of increased CYP1A1 activity and protein levels in pulmonary microsomes from rats (19) or mice (1) exposed to DEP. In addition, they shed new insight on experiments on dogs exposed to aerosol boluses of B(a)p adsorbed on denuded diesel soot, showing that 20% of the released PAH are deposited on the tracheobronchial epithelium and are adsorbed and metabolized, whereas 80% are deposited in the alveolar region, where they rapidly passed into the blood without such metabolism (12).
As mentioned above, the expression of NQO-1, one of several phase II metabolization and antioxidant enzymes, is under the control of the ARE. This genetic response element is known to be activated by ROS and electrophilic compounds. The demonstration of its activation provides additional evidence for the biological impact of ROS production and DEP organic compound metabolization in DEP-treated 16HBE cells. By mobility shift assay, we demonstrate for the first time that in human bronchial epithelial cells, DEP and their extracts dose dependently increased the formation of a complex between an ARE consensus sequence and nuclear proteins, suggesting that ARE activation could occur and be responsible for the induction of NQO-1 gene expression. Such activation has already been shown in macrophages exposed to 3,6-benzo(a)pyrene quinone, an oxidation metabolite of B(a)p, which could explain the induction of the antioxidant enzyme HO-1 by the polar fraction of OE-DEP (18).
Recent evidence suggests that Nrf2 is a critical transcription factor for the regulation of ARE-containing genes that encode antioxidant proteins. These genes include phase II genes and other genes such as the HO-1 gene (15). Thus the critical role of Nrf2 has been supported by knockout mice. Indeed nrf2(-/-) mice exhibit significant reduction of constitutive and/or inducible phase II enzymes (8). Recently, it was shown that both oxidative and PAH-derived DNA adducts formation are accelerated in the lungs of nrf2(-/-) mice compared with nrf2(+/-) mice upon exposure to DEP, whereas the CYP1A1 mRNA levels are similar in both types of mice (1). Our in vitro investigations indicate that in DEP-treated 16HBE cells, Nrf2 is likely involved in the ARE binding proteins, since nuclear translocation of Nrf2 was observed in treated cells.
In conclusion, we have shown that in airway epithelial cells, DEP, via their organic components, modify the cellular redox state. We provide evidence that organic compounds are bioavailable as they induce phase I (CYP1A1) and phase II (NQO-1) gene expression and can be metabolized, as CYP1A1 enzymatic activity is increased. Resulting ROS and reactive metabolites likely contribute to the increased binding of proteins such as Nrf2 to an hARE consensus sequence involved in the expression of the NQO-1 gene. Finally DEP and their extracts induce, in human bronchial and nasal epithelial cells, the expression of numerous genes implicated in detoxification that are activated via xenobiotic responsive element and ARE as well as in the secretion of proinflammatory cytokines via NF-κB-responsive element (6).
This work was supported by Renault Grant DIMAT 235, Ademe Grants BOU 9808 and BOU 0138, and Primequal Grant 97034.
This work was also supported by Caisse d'Assurance Maladie des Professions Libérales de Province, Paris, France.
We acknowledge Dr. D. C. Gruenert for the human bronchial cell line and Dr. A. Coste (Professor Peneygre's department, Hôpital Intercommunal de Créteil, France) and Professor P. Herman (Centre hospitalier universitaire Lariboisière, Paris, France) for the nasal turbinates. We thank Marie-Claude Gendron and Arulraj Nadaradjane for excellent technical help in flow cytometry and Professor D. Ojcius for English corrections.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked ”advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 the American Physiological Society