Use of Proteomics to Demonstrate a Hierarchical Oxidative Stress Response to Diesel Exhaust Particle Chemicals in a Macrophage Cell Line*

Epidemiological studies demonstrate an association between short term exposure to ambient particulate matter (PM) and cardiorespiratory morbidity and mortality. Although the biological mechanisms of these adverse effects are unknown, emerging data suggest a key role for oxidative stress. Ambient PM and diesel exhaust particles (DEP) contain redox cycling organic chemicals that induce pro-oxidative and pro-inflammatory effects in the lung. These responses are suppressed by N-acetylcysteine (NAC), which directly complexes to electrophilic DEP chemicals and exert additional antioxidant effects at the cellular level. A proteomics approach was used to study DEP-induced responses in the macrophage cell line, RAW 264.7. We demonstrate that in the dose range 10–100 μg/ml, organic DEP extracts induce a progressive decline in the cellular GSH/GSSG ratio, in parallel with a linear increase in newly expressed proteins on the two-dimensional gel. Using matrix-assisted laser desorption ionization time-of-flight mass spectrometry and electrospray ionization-liquid chromatography/mass spectrometry/mass spectrometry analysis, 32 newly induced/NAC-suppressed proteins were identified. These include antioxidant enzymes (e.g. heme oxygenase-1 and catalase), pro-inflammatory components (e.g. p38MAPK and Rel A), and products of intermediary metabolism that are regulated by oxidative stress. Heme oxygenase-1 was induced at low extract dose and with minimal decline in the GSH/GSSG ratio, whereas MAP kinase activation required a higher chemical dose and incremental levels of oxidative stress. Moreover, at extract doses >50 μg/ml, there is a steep decline in cellular viability. These data suggest that DEP induce a hierarchical oxidative stress response in which some of these proteins may serve as markers for oxidative stress during PM exposures.

Epidemiological studies demonstrate an association between short term exposure to ambient particulate matter (PM) 1 and cardiorespiratory morbidity and mortality (1)(2)(3). Even though the relative risks are small, there is considerable public health concern because of the large number of exposed people and the existence of high risk groups. People suffering from asthma constitute a susceptible group, as exemplified by acute symptomatic flares after a sudden surge in ambient PM levels (3). This is likely the result of PM-induced airway inflammation and airway hyperreactivity (4 -10). In addition to these short term effects, animal and human studies conducted with diesel exhaust particles (DEP) as a model air pollutant showed that these particles can enhance allergen-specific IgE production and airway allergic inflammation in parallel with increased Th2 cytokine production (5,(11)(12)(13)(14). This raises the important question of the mechanism of these adverse health effects.
Although the biological hypotheses for the mechanisms of PM action are just beginning to develop (15), most of the limited mechanistic data generated to date suggest that oxidative stress is a key biological event in causing the adverse health effects of ambient PM (5, 16 -20). How does ambient PM induce oxidative stress? When exposed to intact DEP or organic extracts made from these particles, macrophages and epithelial cells respond by producing reactive oxygen species (ROS) (16,17). In this regard, it is known that DEP and ambient PM contain transition metals (21,22) as well as redox cycling organic components that elicit ROS production in various cellular locations (19,23,24). For instance, organic DEP extracts induce superoxide production in lung microsomes through the action of NADPH-dependent P450 reductase (20), as well as through damage to the mitochondrial inner membrane (16,17,25). DEP contain a large number of organic chemical compounds among which the polycyclic aromatic hydrocarbons (PAH), nitro-derivatives of PAH, oxygenated PAH derivatives (ketones, quinones, diones), heterocyclic organic compounds, aldehydes, and aliphatic hydrocarbons are the most abundant (23,24,26). We have shown that there is a good correlation between the induction of oxidative stress and PAH content of ambient PM (19,25). Another chemical group that needs to be considered is quinones (27,28). Chemical derivatization of quinones diminished the effect of organic DEP extracts on superoxide production in lung microsomal preparations (20). Moreover, we have shown that polar chemical groups, fractionated from DEP and enriched in quinones, act as potent inducers of oxidative stress in macrophages and epithelial cells (26,29). In addition to being produced by the fuel combustion process, quinones are also generated during enzymatic conversion of PAH in the lung, including their conversion by cytochrome P450 1A1 (28,30).
Although much remains to be learned about the role of oxidative stress in PM-induced adverse health effects, we have demonstrated that organic DEP extracts induce a wide range of biological effects in epithelial cells and macrophages (16 -18, 21). This includes the induction of pro-inflammatory and cytotoxic effects, which can be suppressed by the thiol agent Nacetylcysteine (NAC) (16,18). These pro-inflammatory effects include the production of cytokines and chemokines (13,14), whereas the cytotoxicity depends on the perturbation of mitochondrial function (16,17,29). This includes disruption of the mitochondrial inner membrane potential, cytochrome c release, and caspase 9 activation (17). In addition to these harmful effects, organic DEP components have also been shown to induce cytoprotective responses, including the expression of an antioxidant enzyme, heme oxygenase 1 (HO-1) (26). Based on these diverse effects, we have postulated that DEP may induce a hierarchy of oxidative stress effects, which range from cytoprotective to injurious (19,26).
Proteomics offers a unique means for analyzing the expressed genome and has been successfully employed to look at the generation of oxidative stress at the cellular level (31)(32)(33)(34)(35)(36)(37)(38). In addition to displaying oxidative modification of proteins (31, 32, 34 -36, 38), this approach can also be used to look at newly expressed proteins (33,37,38). We used this approach to test the premise of an incremental oxidative stress response in RAW 264.7 cells during exposure to organic DEP extracts. Our data show that methanol DEP extracts induce a linear increase in newly expressed proteins, Ͼ50% of which are suppressed by NAC. We have subjected 32 of these proteins to mass spectrometry, and used select candidates to show that there is a difference in the dose-response kinetics of antioxidant versus proinflammatory proteins. These results support the existence of a hierarchical oxidative stress model.
Preparation of DEP Methanol Extracts-DEP were provided by Dr. Masura Sagai (Tsukuba, Japan). These particles were collected from the exhaust from a 4JB1-type LD, 2.74-liter, 4-cylinder Isuzu diesel engine under a load of 10 torque onto a cyclone impactor equipped with a dilution tunnel constant volume sampler (21). DEP was collected on high capacity glass fiber filters, from which the scraped particles were stored as a powder in a glass container under nitrogen gas. The particles consist of aggregates in which individual particles are Ͻ1 m in diameter. The chemical composition of these particles, including PAH and quinone analysis, has been described previously (26). Methanol extraction of DEP was performed as described previously (17). Briefly, 100 mg of DEP were suspended in 25 ml of methanol and sonicated for 2 min. The suspension was centrifuged at 425 ϫ g for 10 min at 4°C, and the supernatant transferred to a preweighed Eppendorf tube to determine the amount of extractable material. After drying under nitrogen gas, the dried material was completely dissolved in Me 2 SO and aliquots saved at Ϫ80°C in the dark until use.
Cellular Stimulation with DEP Extracts-RAW 264.7 cells were cultured in complete medium, which consisted of Dulbecco's modified Eagle's medium supplemented with 1% penicillin/streptomycin and 10% fetal bovine serum. For cellular stimulation, 2 ϫ 10 6 cells in 3 ml of culture medium were treated with the indicated amounts of the DEP extract in 6-well culture plates for 6 h at 37°C in a humidified CO 2 incubator. Control cultures received 0.1% of the Me 2 SO carrier. Some cultures received 20 mM NAC from a 1 M stock made in HEPES buffer immediately before use. NAC was added independently prior to, concomitant with or following the addition of the DEP extract as indicated.
To determine whether NAC interacts directly with electrophilic chemicals in the extract, we premixed 10 mg of NAC with 1 mg of the DEP extract in a small volume (50 l). This mixture was incubated at room temperature for 1 h before addition to the cell culture at a final extract concentration of 10 -50 g/ml, while limiting the NAC concentration in the medium to 61.5 M. The controls consisted of cells receiving DEP chemicals only, or 20 mM NAC added to the culture medium for 2 h prior to the addition of the DEP extract at the indicated concentrations. The cells were harvested 6 h later and used for HO-1 immunoblotting as described previously (26).
Determination of Cellular GSH/GSSG Ratios-Total glutathione (GSH plus GSSG) and GSSG were measured in a recycling assay that uses 5,5Ј-dithiobis(2-nitrobenzoic acid) and glutathione reductase (39). Briefly, cells were lysed and deproteinized in 3% 5-sulfosalicylic acid. Whole cell lysates were cleared at 4°C by centrifugation at 20,800 ϫ g in an Eppendorf centrifuge. The supernatant was used to measure total and oxidized glutathione. Total glutathione was read from a GSH standard curve, prepared in 5-sulfosalicylic acid. For the GSSG assay, 100 l of supernatant was incubated with 2 l of 2-vinylpyridine and 6 l of triethanolamine for 60 min on ice. GSSG standards were treated in the same way, and the GSSG content of the samples was calculated from a GSSG standard curve. Reduced GSH was calculated by subtracting GSSG from total glutathione.
Determining Cell Viability by Propidium Iodide (PI) Staining-Cells (3 ϫ 10 6 ) were plated into 3.5-cm plates in 3 ml of medium and rested for 4 h. Some cultures were preincubated with 20 mM NAC for 2 h. Varying concentrations of DEP were added to these cultures for 18 h. Cells were collected, washed twice in PBS, and resuspended in 500 l of PBS containing 0.5 g/ml PI for 5 min. Cells were analyzed in a FACScan (Becton Dickinson, Mountain View, CA) equipped with a single 488-nm argon laser. Dead cell fragments were gated out by forward and side scatter, and PI analysis was performed at excitation and emission settings of 488 and 575 nm, respectively.
Protein Extraction and Sample Preparation-Aliquots of 2 ϫ 10 7 RAW 264.7 cells were washed twice with ice-cold PBS containing protease inhibitors and sonicated in ice-cold radioimmune precipitation assay buffer containing 10 mM NaPO 4 , pH 7.2, 0.3 M NaCl, 0.1% SDS, 1% Triton X-100, 1% sodium deoxycholate, 2 mM EDTA, protease inhibitor mixture set III (100 mM AEBSF, 80 M aprotinin, 5 mM bestatin, 1.5 mM E-64, 2 mM leupeptin, 1 mM pepstatin), and phosphatase inhibitor mixture set II (200 mM imidazole, 100 mM sodium fluoride, 115 mM sodium molybdate, 100 mM sodium orthovanadate, 400 mM sodium tartrate dihydrate) (Calbiochem, La Jolla, CA) for 10 s. Lysates were centrifuged at 1000 ϫ g for 5 min. To remove the salt from the lysates, the supernatant proteins were precipitated with trichloroacetic acid (10% w/v), 20 mM DTT for 30 min on ice. The precipitate was collected at 20,800 ϫ g for 10 min at 4°C and washed three times with 10% trichloroacetic acid, 20 mM DTT. Trichloroacetic acid in the precipitate was removed through the extraction with diethyl ether or acetone plus 10 mM DTT. After drying, the pellet was resuspended by sonication in a buffer containing 7 M urea, 2 M thiourea, 4% w/v CHAPS, 100 mM DTT, 0.2% v/v Bio-Lyte pH 3/10:4/6:5/8 (1:0.5:0.5), 5% glycerol, and protease/phosphatase inhibitors (mixture sets II and III). After stand-ing for 1 h at room temperature, the sample was centrifuged at 23,800 ϫ g for 10 min at 15°C, and the supernatants stored at Ϫ80°C until use for two-dimensional PAGE. Protein concentration in these samples was estimated by using a commercial Bradford kit (DC reagent kit, Bio-Rad), and bovine serum albumin as standard.
Two-dimensional Polyacrylamide Gel Electrophoresis (PAGE)-Twodimensional gel electrophoresis was performed with the Bio-Rad system as described by Jungblut and Thiede (40). 350 g of whole cell lysate was added to each immobilized pH gradient (IPG) strip, which was rehydrated in 8 M urea, 2% CHAPS, 50 mM, 0.2% Bio-Lyte 3/10 ampholyte, 0.001% bromphenol blue. The pre-isoelectric focusing and isoelectric focusing (IEF) were performed using pre-made 17-cm length IPG strips (pH 3-10 NL) on the Protean IEF cell. The pre-isoelectric focusing was performed linearly up to 500 V for 1 h, held at 500 V for 1.5 h. Formal IEF was then performed with a linear increase up to 10,000 V over 2 h and then held at 10,000 V for 7 h a total of 90 KV-h. For the second dimension, the IPG strips were equilibrated in a buffer containing 37.5 mM Tris-HCl, pH 8.8, 20% glycerol, 2% SDS, and 6 M urea with 2% dithiothreitol (Sigma), followed by 8 -16% SDS-PAGE on a Protean Plus Dodeca Cell (Bio-Rad). Gels were stained with Sypro-Ruby (Molecular Probes, Eugene, OR) and visualized under ultraviolet light with a Molecular Imager FX Pro Plus (Bio-Rad). To check the reproducibility of the data, three independent two-dimensional analyses were performed on each cellular lysate.
Protein Identification-Protein spots were selected based on staining intensity of the Sypro Ruby as determined by the PDQuest software (Bio-Rad). This sensitivity of the software is set to detect a 2-fold increase in staining intensity as a criterion for a significant increase in protein expression. For the purpose of this study, we increased the stringency to 8-fold. Spots were excised by a spot-excision robot (Proteome Works, Bio-Rad) and deposited into 96-well plates. Gel spots were washed and digested with sequencing-grade trypsin (Promega, Madison, WI), and the resulting tryptic peptides were extracted using standard protocols (41). Trypsin digestion and extraction, and peptide spotting onto a matrix-assisted laser desorption ionization (MALDI) targets, were accomplished by a robotic liquid handling work station (MassPrep, Micromass-Waters, Beverly, MA). MALDI peptide fingerprint mass spectra were acquired with a MALDI time-of-flight instrument (M@LDI-R, Micromass-Waters), using ␣-cyano-4-hydroxycinnamic acid (Sigma) as the matrix. Peptide sequencing was accomplished with nanoflow high performance liquid chromatography with electronic flow control (1100 Series nanoflow liquid chromatography system, Agilent Technologies, Palo Alto, CA) interfaced to an ion trap mass spectrometer (LC-MSD Trap SL, Agilent Technologies). A reverse phase column (75 m ϫ 150 mm, C18 Zorbax StableBond) was used as the analytical column. A Zorbax 300SB enrichment pre-column (0.3 ϫ 5 mm) was used to concentrate and desalt the peptide mixtures.
The MS data from both tandem mass spectra from the LC-MS/MS experiments and the MALDI-MS peptide fingerprint mass spectra were searched against a subset of rodent proteins in the SWISS-PROT protein sequence data base, using the Mascot search program (Matrix Science, London, United Kingdom) (www.matrixscience.com). Positive protein identification was based on standard Mascot criteria for statistical analysis of the MALDI peptide fingerprint mass spectra and the LC-MS/MS data. A Ϫ10log(P) score, where P is the probability that the observed match is a random event, of Ͼ72 was regarded as significant.
Data Analysis-GSH/GSSG ratio, cell viability, and newly induced protein data are expressed as the mean Ϯ S.E. One-way analysis of variance was used to determine differences between groups with post hoc comparisons by the method of Fisher. Significance was assumed at p Ͻ 0.05. important role in ROS scavenging and maintenance of cellular redox equilibrium (42). A decline in the ratio of reduced to oxidized glutathione (GSSG) is a sensitive parameter for cellular oxidative stress (42). When exposed to incremental amounts of a methanol DEP extract, RAW 264.7 cells show a progressive and statistically significant decline in the GSH/GSSG ratio at doses Ͼ10 g/ml (Fig. 1). This effect was diminished by pretreating the cells with NAC (Fig. 1).

Organic DEP Extracts
Cells respond to oxidative stress in a variety of ways, including the activation of intracellular signaling pathways that exert pro-inflammatory effects in the lung (5). One example is the activation of the JNK and p38 MAPK cascades by the organic DEP extract in RAW 264.7 cells (Fig. 2). This effect is demonstrated by the increased phosphorylation of p38 MAPK ( Fig. 2A) and the 46-and 54-kDa JNK isoforms on allosteric sites that lead to their activation (Fig. 2B). Although an extract dose of 10 g/ml failed to induce JNK activation, doses of Ն50 g/ml did induce kinase activation as determined by anti-phosphopeptide immunoblotting (Fig. 2B). Prominent p38 MAPK activation also required an extract dose of Ն50 g/ml, while registering a smaller effect at 10 g/ml ( Fig. 2A). The increase in site-specific phosphorylation was not the result of changes in kinase abundance, as demonstrated by parallel immunoblotting for kinase protein (Fig. 2, bottom panels). Prior treatment with NAC interfered in these phosphorylation events, confirming that these MAP kinase cascades are activated under conditions of oxidative stress (Fig. 2, A and B).
In addition to activating these signaling cascades, extract doses Ն10 g/ml induced cellular toxicity as shown by increased PI uptake (Fig. 3). This increase in cell death was more noticeable at doses Ն50 g/ml (Fig. 3). We have previously shown that this cytotoxic effect is a programmed cell death event that involves mitochondrial perturbation and release of cytochrome c (16,17). The involvement of oxidative stress is confirmed by the ability of NAC to interfere in cytotoxicity (Fig.  3). Taken together with the data in Fig. 2, these findings demonstrate that at doses Ն10 g/ml, organic DEP extracts induce a progressive increase in injurious cellular responses. However, elucidation of cellular responses at 10 g/ml is important because not all oxidative effects are injurious in nature (19,26). This necessitated the use of a discovery tool that is more appropriate for revealing an extensive dose-response relationship.

Two-dimensional Gel Electrophoresis and Mass Spectrometry Reveal a Hierarchical Response to Organic DEP Extracts-
Changes in the proteome of RAW 264.7 macrophage cells were examined in cell populations exposed to incremental amounts of the DEP extract. Protein expression was displayed by twodimensional PAGE and yielded Ͼ1200 individual polypeptides in unstimulated cells. The addition of DEP extracts induced new protein expression, which was defined as Ͼ8-fold increase (p Ͻ 0.01) in the staining intensity of each individual Sypro Ruby-stained polypeptide (Fig. 5). The number of newly expressed proteins increased linearly as the extract dose increased, and yielded 10, 65, and 100 new proteins at DEP extract concentrations of 10, 50, and 100 g/ml, respectively (Fig. 4A). Linear regression analysis showed an excellent correlation (r 2 ϭ 0.982) between extract dose and the number of newly induced proteins (Fig. 4B). There was some overlap as well as unique expression profiles for each extract dose (Fig.  4C). Thus, six new proteins were expressed at all extract doses and are listed in Fig. 4D. These include proteins that play a role in antioxidant defense (HO-1, catalase, and metallothionein), a signaling pathway component (the ␣1 subunit of p38 MAPK ), a transcription factor (Rel A), and a component of the Emden Meyerhoff pathway (GAPDH). An additional 45 newly expressed proteins were shared in cell populations treated with 50 and 100 g/ml extract, whereas the respective cell populations treated with doses of 10, 50, and 100 g/ml showed 4, 14, and 51 uniquely expressed proteins (Fig. 4D).
NAC addition diminished protein expression by ϳ50%, confirming the possible relationship to oxidative stress (Fig. 4B). NAC suppression or subtraction was defined as a 50% decrease in staining intensity of an inducible protein. This suppression by NAC was used as a criterion to select response markers for further analysis by protein mass spectrometry (Fig. 5). GAPDH is an example of an oxidative stress protein that was induced in RAW 264.7 cells during exposure to 50 g/ml DEP extract (Fig.  5A). Compared with untreated cells, GAPDH expression increased Ͼ8-fold, whereas the inclusion of NAC decreased that response by 70% (Fig. 5B). Use of this approach led to the identification of an additional 31 proteins by mass spectrometry (Table I). These proteins were distributed into two zones according to pI values (Fig. 5A). The first zone falls in the pI range 4.5-5.5, and includes a signaling component, p38 MAPK ␣1, the tyrosine kinase, ErbB-2, as well as the detoxification RAW 264.7 cells were exposed to DEP extracts at indicated concentrations, in the absence or the presence of 20 mM NAC, for 6 h before cellular extraction and analysis of the soluble proteins by two-dimensional electrophoresis. These data were reproduced three times, during which the variability in protein expression was Ͻ10%. enzyme, alcohol dehydrogenase (Table I). The second zone, spanning pI 5.8 -9.0, contains several proteins involved in intermediary metabolism, ATP production, and oxidative stress (e.g. GAPDH), a transcription factor (e.g. Rel A), and antioxidant defense proteins (e.g. HO-1, catalase, and metallothionein) (Fig. 5). To increase the protein resolution in this zone of the gel, cellular extracts were further analyzed on two-dimensional gels, which utilized a narrower focusing range (pH 5.5-6.7) (Fig. 6). These zoom gels helped to confirm the induction of HO-1 and catalase expression by the DEP extract, as well as the ability of NAC to suppress their expression (Fig. 6, A-C).
To examine the fidelity of these newly induced proteins and to confirm the two-dimensional PAGE analysis, Western blotting was performed. GAPDH immunoblotting confirmed its expression at all DEP extract doses tested (Fig. 5C). Interestingly, GAPDH expression was fully induced at the lowest DEP extract dose, and showed increased sensitivity to NAC suppression at higher extract doses (Fig. 5C). Similar subtractive protein expression, making use of two-dimensional PAGE and Western blotting, was demonstrated for catalase and HO-1 ( Fig. 6D), as well as Rel A (p65) and metallothionein (Fig. 6E).

The Suppressive Effect of NAC Is Dependent on Cellular Antioxidant Effects as Well as Direct Electrophilic
Interactions with DEP Chemicals-NAC is the N-acetyl derivative of the naturally occurring amino acid, L-cysteine, and functions as a radical scavenger as well as a precursor for glutathione synthesis (43). In addition to these cellular antioxidant effects, NAC also utilizes its SH group to directly complex to electrophilic DEP chemicals. This interaction could take place in the tissue culture medium as well as intracellularly. To discern between these different modes of action, we demonstrated that NAC addition 2 h after the introduction of the DEP extract could suppress HO-1 expression, provided that the stimulus was removed before the addition of NAC (Fig. 7A, lane 2). However, if not removed from the culture medium, the stimulating effects of the DEP chemicals were unopposed (lane 3). These data suggest that NAC interfere in the pro-oxidative effects of DEP chemicals at a cellular level. In the same experiment, it could also be demonstrated that NAC addition prior to the delivery of the stimulus can prevent HO-1 induction (Fig.   FIG. 5. Two-dimensional gel electrophoresis profile in the presence of 50 g/ml organic DEP extract. A, proteins that were induced Ͼ8-fold and subtracted in the presence of NAC were selected as oxidative stress markers that were identified by MS. Those proteins are numbered and their identities disclosed in Table I. B, excerpt of the two-dimensional profile to show how above criteria led to the identification of GAPDH as an oxidative stress marker. The top panel shows background expression in untreated cells, the middle panel shows increased expression by the extract, and the bottom panel shows the subtracted response in the presence of NAC. C, GAPDH immunoblotting shows the subtractive expression of this protein in crude cell lysates. See "Materials and Methods" for experimental details. These data were reproduced three times, during which the variability in protein expression was Ͻ10%. 7A, lanes 5-7); the effect was more prominent in unwashed cell cultures (lane 7) compared with cells where the thiol was added for 2 h and then washed away (lane 6). This raises the possi-bility that NAC may also interfere in the effects of the inducing chemicals by direct chemical interactions, some of which may occur in the culture medium. This possibility was further ex- a DEP dose: 10 only ϭ 10 g/ml; Ն10 ϭ 10, 50, and 100 g/ml; 50 only ϭ 50 g/ml; Ն50 ϭ 50 and 100 g/ml; 100 only ϭ 100 g/ml. b NAC suppressibility: ϩ ϭ 25% decrease in intensity; ϩϩ ϭ 50%; ϩϩϩ ϭ 75%; ϩϩϩϩ ϭ 100%.
plored by premixing a weight excess of NAC with the DEP extract in a small volume before adding the mix to the cell culture medium (Fig. 7B). In this experiment, in which the final NAC concentration in the culture medium was Ͻ100 M, the interference in HO-1 expression (lanes 4 -6) was equivalent to the effect of 20 mM NAC introduced with the stimulus (lanes  7-9). This is compatible with the data in Fig. 4. Although the extent to which a direct electrophilic interactions versus cellular antioxidant effects contribute to the NAC effect is difficult to quantify, the net effect is to prevent a decline in the cellular GSH/GSSG ratio as well as new protein expression (Fig. 1). Taken together, the data depicted in Figs. 4 -7 demonstrate that organic DEP extracts induce the expression of a range of proteins, ϳ50% of which are suppressed by NAC. Although prominent MAPK activation and the induction of cellular toxicity require DEP extract doses Ͼ10 g/ml, the antioxidant enzymes (HO-1, catalase, and metallothionein) and GAPDH were induced at lower extract doses. This suggests a segrega-FIG. 6. Narrow pH range focusing gels improve differentiation of oxidative stress-related proteins. HO-1 and catalase are shown in the two-dimensional gel with the pH range of 5.5-6.7 (shown as dashed lines in Fig. 5A). A, control sample (RAW 264.7 cells exposed to the Me 2 SO carrier); B, cells exposed to 50 g/ml dose of DEP, showing induction of HO-1 and catalase. C, cells exposed to the same dose of the extract in the presence of NAC to demonstrate the suppression of HO-1 and catalase expression. D, HO-1 and catalase immunoblotting to confirm the expression of these proteins in crude cell lysates. E, Rel A p65 and metallothioneins immunoblotting to confirm the expression of these proteins in crude cell lysates. *, MTT1 isoform identified by proteomics; Tesmin, metallothionein-like protein also identified in our immunoblot. tion of protective versus injurious cellular effects at different extract doses and at different levels of oxidative stress. DISCUSSION We demonstrate that organic DEP extracts induce a dosedependent decrease in the GSH/GSSG ratio in RAW 264.7 cells, in parallel with a linear increase in the number of newly expressed proteins. More than half of these proteins were suppressed in the presence of NAC. Using mass spectrometry analysis, 32 newly induced/NAC-suppressed proteins were identified. These include antioxidant enzymes, e.g. HO-1 and catalase, as well as proteins that play a role in pulmonary inflammation, namely p38 MAPK and Rel A. HO-1 was induced at a low extract dose and with minimal decline in the GSH/GSSG ratio, whereas prominent Jun and p38 MAPK activation required higher extract amounts and incremental levels of oxidative stress. Moreover, at extract doses Ͼ50 g/ml, there is an increase in the rate of cytotoxicity. These data suggest that organic DEP chemicals induce a hierarchical oxidative stress response, which is reflected by the types of proteins being expressed.
Mass spectrometry and proteome analysis have been useful in identifying oxidative stress markers under a variety of disease conditions, including cellular hypoxemia, T-cell dysfunction in setting of AIDS, Alzheimer's disease, and tissue inflammation (31)(32)(33)(34)(35)(36)(37)(38). Typically, proteome analysis of oxidative stress markers requires the identification of protein S-nitrosation, tyrosine nitration, glutathionylation, or methionine oxidation (31,32,34,35,38). Although these post-translational modifications are helpful as a qualitative display of oxidative stress, this approach is not helpful in quantifying the cellular response to oxidative stress. We therefore used an alternative proteomics approach, which looks at the total number of newly expressed proteins as well as their NAC subtraction, to quantify the oxidative stress response. This showed that increasing amounts of the organic DEP extract induce a progressive decline in the cellular GSH/GSSG ratio, in parallel with a linear increase in the number of newly expressed proteins (Figs. 1 and  4). The decline in the GSH/GSSG ratio is a representative cellular marker for oxidative stress (Fig. 1) and is directly involved in eliciting cellular responses, including antioxidant defense and protection of the mitochondrial PT pore (42,44,45). The inhibitory effect of NAC is particularly relevant to the induction of oxidative stress by organic DEP chemicals (Figs. 1-7). Among a wide range of antioxidants tested, thiol antioxidants were the most specific in interfering in the pro-oxidative effects of organic DEP chemicals in vitro and in vivo (18).
What conclusions can be drawn from the proteome analysis of DEP-treated RAW 264.7 cells? The linear increase in new protein expression with increasing extract doses suggests an escalating cellular response to oxidative stress (Fig. 4). This notion is supported by the fact that HO-1, catalase, and metallothionein (46 -48) are induced at the lower (10 g/ml) extract dose (Figs. 6 and 7; Table I), whereas prominent MAP kinase activation (Fig. 2) and induction of cellular toxicity require extract doses Ͼ10 g/ml (Fig. 3). HO-1 and catalase are antioxidant enzymes (49), suggesting that cytoprotective pathways are induced at the lowest levels of oxidative stress (Fig.  8). This may constitute the first tier of a hierarchical oxidative stress response (Fig. 8). HO-1 expression is also a very sensitive marker for oxidative stress in bronchial epithelial cells (29,50), another key cellular target for PM. The induction of HO-1 expression by redox cycling chemicals, including cadmium and organic DEP compounds, is dependent on the anti-oxidant response element in the promoter of that gene (Fig. 8) (26,50). This genetic response element is transcriptionally activated by a basic leucine zipper transcription factor, Nrf2 (Fig. 8) (51). It is interesting that oxidative DNA damage and the accumulation of 8-hydroxydeoxyguanosine in the lungs of Nrf2 knockout mice is exaggerated during exposure to diesel exhaust fumes (52).
Although increased expression of the p38 MAPK ␣1 isoform can be seen to occur at 10 g/ml extract (Fig. 4D), prominent activation of the p38 MAPK and Jun kinase cascades required Ͼ10 g/ml amounts of the same material (Fig. 2). These stressactivated protein kinases play a role in the expression and transcriptional activation of several AP-1 proteins (53), and are often linked to pro-inflammatory and injurious cellular responses (Fig. 8). This includes the transcriptional activation of cytokine and chemokine genes (Fig. 8). We propose that these pro-inflammatory effects constitute a second tier or a superimposed level of oxidative stress, and that proteins that are induced or activated in this zone play a role in the pro-inflammatory and adjuvant effects of DEP in the lung (5, 10 -13). This notion is strengthened by increased expression or oxidative In the same experiment, we also tested the effect of prior NAC addition before adding 50 g/ml DEP extract for 6 h (see legend for lanes 5-7). HO-1 immunoblotting was conducted as described in Fig. 6. B, HO-1 immunoblotting was used to demonstrate that premixing of NAC and the DEP extract is effective in suppressing the pro-oxidative effects of electrophilic DEP chemicals. 10 mg of NAC was premixed with 1 mg of the DEP extract in a small volume (50 l). This mixture was added to the cell culture for 6 h to give a final DEP extract concentration of 10 -50 g/ml, while limiting the NAC concentration to 61.5 M (lanes 4 -6). The controls consisted of cells receiving DEP chemicals and either no NAC (lanes 1-3) or 20 mM NAC added to the culture medium 2 h before the addition of the DEP extract (lanes 9 -12). The cellular extracts were used for HO-1 immunoblotting as described under "Materials and Methods" for experimental details. modification of proteins that play a role in the regulation of inflammation, e.g. Rel A (54), granulocyte/macrophage colonystimulating factor precursor (55), tumor necrosis factor receptor 2 (56), glucocorticoid receptor (54), EGR-4 (57), and acetyl-CoA carboxylase 2 (58) ( Table I). Although increased expression of the glucocorticoid receptor may be important for the treatment of allergic disease, it is interesting that steroid administration does not reverse the pro-inflammatory effects of DEP in the nasal mucosa (59). This could be related to the fact that this receptor is a zinc finger transcription factor and can be oxidatively inactivated by cross-linking of critical cysteine groups (54). Among the pharmacologic agents tested to curb the proinflammatory effects of DEP, only NAC was fully effective in suppressing the adjuvant effects of DEP in a murine allergen challenge model (18).
The final proposed tier or superimposed level of oxidative stress is cytotoxicity, including the initiation of programmed cell death (Fig. 8) (19,29). We have previously demonstrated that this effect is dependent on mitochondrial perturbation, including effects on the mitochondrial membrane potential and cytochrome c release (16,17,25). This notion is strengthened by the induced expression of proteins that regulate mitochondrial function and apoptosis, including mitochondrial fumarate hydratase (60), voltage-dependent anion-selective channel protein 1 (VDAC-1) (61, 62), mitogen-activated protein kinase/ extracellular signal-regulated kinase kinase kinase 1 (63), and diacylglycerol kinase (64) ( Table I). It is interesting that the DEP extracts also induced Fas-associating death domain-containing protein (FADD) expression, which may play a role in receptor-induced apoptosis, as well as the expression of proteins that play a role in intermediary metabolism and are linked to regulation of oxidative stress, e.g. phosphoenolpyruvate carboxykinase (65), ␣-phosphoenolase (66), and glyceraldehyde-3-phosphate dehydrogenase (66 -68) (Table I). We are in the process of analyzing these pathways in more detail.
The pro-oxidative effects of organic DEP extracts likely reflect the presence of redox cycling chemicals (10,24,26). In this regard, we have recently demonstrated that the use of increasing polar elutants to fractionate organic DEP extracts by silica gel chromatography yielded aromatic and polar chemical groups, which mimic the effect of the crude extract in cellular toxicity assays (26,29). Chemical analysis has shown that the aromatic fraction is enriched for PAH, whereas the polar fractions are enriched for quinines (26). We are currently investi-gating the hypothesis that these chemical groups are responsible for the pro-oxidative and pro-inflammatory effects of PM. It may be relevant that analysis of ambient PM with aerodynamic diameter Ͻ0.15 m (ultrafine particles), collected by particle concentrators in the Los Angeles basin, demonstrated an excellent correlation between PAH content and their capacity to generate ROS in the presence of DTT (25). Both parameters were linearly correlated with the HO-1 expression in RAW 264.7 cells (25).
The inhibitory effects of NAC on protein expression is interesting from a number of different perspectives. The ROS scavenging effects of NAC is explained by its SH-group, which has the potential to directly interact with oxidants such as H 2 O 2 , leading to the formation of H 2 O and O 2 . Deacetylation of NAC also leads to the formation of cysteine, which is a precursor for glutathione synthesis (43). In addition to its radical scavenging effects, GSH directly conjugates to some of the quinone species presenting DEP, including benzo-and naphthoquinones (69). In addition, NAC itself can participate in electrophilic interactions, thereby establishing multiple mechanisms by which this thiol agent can interfere in the oxidative stress effects of DEP chemicals. Whatever the exact contribution of direct electrophilic interactions versus effects on GSH synthesis and radical scavenging may be, the net effect of NAC is to prevent a drop in the cellular GSH/GSSG ratio (Fig. 1) as well as to interfere in ROS generation (16,17). In fact, the specificity of the NAC antioxidant effects (18) may prove useful in identifying the major chemical groups in DEP that is responsible for ROS generation.
How does exposure to 1-100 g/ml of the DEP extract compare with in vivo PM exposures in humans? Although it is difficult to directly extrapolate from the in vitro to the in vivo exposure amounts, it is possible to demonstrate using human dosimetry models that the dose of PM2.5 (particulate matter with aerodynamic diameter Յ 2.5 m) deposition at airway bifurcation points is comparable with the in vitro tissue culture concentrations recalculated as extract dose/cm 2 (70). Thus, we have shown that 1-100 g/ml DEP extract is equivalent to 0.14 -14 g/cm 2 in a tissue culture dish, whereas an asthmatic person with airway stasis, breathing polluted air in Rubidoux, California, can deposit 2.3 g/cm 2 PM2.5 at tracheobronchial bifurcation sites (70). It is possible, therefore, that at these so-called hot spots of deposition (airway bifurcations), the bronchial mucosa may be exposed to DEP chemical doses that are FIG. 8. Schematic to explain the hierarchical oxidative stress model in response to redox cycling DEP components. Activation of antioxidant enzymes HO-1 and catalase reflects the first tier oxidative stress response, activation of the p38 MAPK and Jun kinase cascades constitutes the second tier of oxidative stress responses, whereas the final tier of oxidative stress response, mediated by mitochondrial perturbation, leads to cytotoxic effects. Please note that the suggested tiers are not rigidly demarcated, but represent an escalating trend, in which cytoprotective yield to pro-inflammatory and cytotoxic responses. The data in Table I indicate some overlap and intermingling of protective versus injurious effects at the interface of these zones. toxicologically relevant from an oxidative stress perspective (70).
In summary, we have shown that proteomics analysis can be used to study the linear increase in new protein expression in parallel with increased levels of DEP-induced oxidative stress.