Chem. Res. Toxicol. 2006, 19, 255-261
255
Midpolarity and Nonpolar Wood Smoke Particulate Matter Fractions Deplete Glutathione in RAW 264.7 Macrophages Alena Kuba´tova´,*,†,§ Laura C. Dronen,† Matthew J. Picklo, Sr.,‡ and Steven B. Hawthorne† Energy & EnVironmental Research Center, P.O. Box 9018, UniVersity of North Dakota, Grand Forks, North Dakota 58202-9018, and Department of Pharmacology, Physiology & Therapeutics, School of Medicine and Health Sciences, UniVersity of North Dakota, P.O. Box 9037, Grand Forks, North Dakota 58203-9037 ReceiVed June 24, 2005
Wood smoke particulate matter (PM) is a complex mixture of components falling in a spectrum of highly polar to nonpolar species. Wood smoke PM is a likely factor in pulmonary disease and induces oxidative damage. Most toxicity studies focus upon nonpolar species such as polycyclic aromatic hydrocarbons (PAHs). However, the role of more polar PM constituents as toxicants is not clear. In this work, we evaluated the ability of multiple fractions of varying polarity to deplete glutathione (GSH) in RAW 264.7 macrophages and BEAS-2B bronchial epithelial cells. We utilized hot pressurized (subcritical) water to fractionate wood smoke PM into seven fractions of decreasing polarity. In contrast to polar fractions, midpolarity and nonpolar fractions exhibited greater GSH depletion (ED50 at PM concentrations of ∼50 µg/mL). GSH depletion caused by nonpolar fractions (extracted at 250-300 °C) was associated with the presence of PAHs. In midpolarity fractions (extracted at 100-150 °C), oxy-PAHs, syringylguaiacyls, disyringyls, and lower molecular weight PAHs were found. Direct comparison of GSH depletion by individual oxy-PAHs and PAHs suggests that oxy-PAHs are contributors to oxidative stress caused by wood smoke PM. However, other unidentified PM constituents contribute to GSH depletion as well. The results indicate the toxicological importance of oxygenated organics found in midpolarity PM fractions. Introduction Wood smoke particulate matter (PM)1 is a complex mixture of hundreds of organics contributing to indoor and outdoor air pollution (1). Wood smoke PM consists mainly of anhydrosugars (breakdown products of cellulose) and lignin pyrolysis products such as syringol (2,6-dimethoxyphenol) and guaiacol (2methoxyphenol) derivatives, oxy-polycyclic aromatic hydrocarbons (oxy-PAHs), and PAHs (2-4). The potential contribution of polar (e.g., oxygenated) organics to wood smoke PM toxicity has been recognized (5-7). However, the toxicity of wood smoke PM is usually attributed to nonpolar PAHs or iron (8, 9). The potential of oxygenated organics to cause adverse health effects is supported by toxicological data on pure compounds (10-12). Oxy-PAHs (mainly quinones) are known for their redox cycling activity and, therefore, are associated with oxidative stress (12). Phenolic compounds may function as antioxidants but may also generate secondary radicals (10, 11). Exposure to wood smoke PM can cause the generation of reactive oxygen species (ROS), lipid peroxidation, oxidative DNA damage, and inflammation (9). In vivo exposure of rats to concentrated ambient PM increases the steady state concentration of ROS (13), and instillation with residual oil fly ash in mice depletes glutathione (GSH) in bronchial lavage fluid (14). Alveolar macrophages exposed to PM in vitro undergo intra* To whom correspondence should be addressed. Tel: 701-777-0348. Fax: 701-777-2331. E-mail:
[email protected]. † Energy & Environmental Research Center. ‡ School of Medicine and Health Sciences. § Present contact address: Chemistry Department, University of North Dakota. 1 Abbreviations: DMSO, dimethyl sulfoxide; GSH, glutathione; HBSS, Hank’s balanced salt solution; LDH, lactate dehydrogenase; MCB, monochlorobimane; PM, particulate matter; PAHs, polycyclic aromatic hydrocarbons.
cellular oxidation, become apoptotic, have depressed respiratory burst responses, and secrete proinflammatory cytokines (14-18). Similarly, exposure of bronchial epithelial cell lines to PM causes upregulation of cytokines and cytotoxicity (16, 19-21). The extent of oxidative stress depends on the redox balance of the cells governed by a system of enzymes with an ongoing number of cyclic reactions such as generation of reactive oxygen and nitrogen species, oxidation of GSH, and reduction of GSH disulfide (9, 16, 20, 22). The cellular redox balance can be disturbed by the depletion of GSH. GSH participates in detoxification at several different levels; it reduces peroxides or may scavenge free radicals or conjugates with electrophilic compounds. Fractionation allows for risk assessment of complex PM mixtures (23-27). Besides our previous work, there is only one toxicological study employing fractionation of wood smoke PM (7). However, Pimenta et al. (7) characterized different polarity fractionations of wood smoke PM by different toxicological end points. Thus, the potential of different polarity fractions to affect the same toxicological end point was not evaluated. We have previously demonstrated the use of hot pressurized water for PM extraction/fractionation into wide-range polarity fractions (28). The extraction recoveries from wood smoke PM using hot pressurized water were higher (54 wt %) than those obtained with traditionally employed organic solvents such as methanol (34 wt %), methylene chloride (16 wt %), or hexane (5 wt %). We have shown that water from 25 to 300 °C (under sufficient pressure to maintain the liquid phase) is an excellent extraction solvent for polar as well as nonpolar organics (28, 29). The practical effect of being able to control water’s polarity with temperature is the ability to control the solubility of various organic compounds (29). Therefore, compounds of different polarities, such as phenols, oxy-PAHs, and PAHs, are sequen-
10.1021/tx050172f CCC: $33.50 © 2006 American Chemical Society Published on Web 12/13/2005
256 Chem. Res. Toxicol., Vol. 19, No. 2, 2006
tially extracted from wood smoke PM using hot pressurized water (28). Thus, in this work, we have employed hot pressurized water fractionation to test the hypothesis that polar constituents of wood smoke PM deplete cellular GSH levels. The impact of different polarity fractions of wood smoke PM on GSH levels was tested using murine macrophages RAW 264.7 and BEAS2B bronchial epithelial cells.
Experimental Procedures Sample Material. Bulk wood smoke PM was collected from a chimney that vented an airtight wood stove burning a mix of hardwoods (2). Fifty milligram portions were used for all extraction experiments. Hot Pressurized Water Extraction/Fractionation. Hot pressurized water fractionation was performed using an extraction apparatus previously described in detail (28). The extractions were performed at a flow rate of 0.5 mL/min with HPLC grade water (Fisher Scientific, Pittsburgh, PA), which was purged with nitrogen to remove dissolved oxygen prior to the extraction. Extractions were carried out in a 3.47 mL supercritical fluid extraction cell (9.4 mm i.d., 50 mm long, Keystone Scientific, Bellefonte, PA) equipped with 0.5 µm frits. The extraction procedure started by pressurizing the system with water at a flow of 1 mL/min to ca. 50 bar (3 min). At this time, the water flow rate was set to 0.5 mL/min, the outlet valve was opened, and collection of the eluent began (time ) 0). To obtain different polarity fractions, the extraction was performed sequentially at temperatures of 25, 50, 100, 150, 200, 250, and 300 °C. Pressure does not have a significant effect on extraction with pressurized water (29). For easier operation and safety reasons, the pressure was held 50-100 bar above the saturation pressure at each temperature. The pressure was held at 50 bar until 150 °C, and then at 200 °C, it was increased to 100 bar to maintain water in the liquid state and to 150 bar at 250 and 300 °C. Each temperature was held constant for 30 min, at which time the collection vial was replaced and the system was heated to the next higher temperature (requiring 30-90 s). The polarity of extracted organics decreased with the water temperature; therefore, fractions extracted at 25-50 °C, 100-150 °C, and 200-300 °C were labeled polar, midpolarity, and nonpolar, respectively (28). The residue after the extraction was air-dried and sonicated overnight (18 h) with methylene chloride to ensure complete extraction of the least polar organics. Between each extraction, the system was washed with pressurized water heated from 25 to 325 °C (100 bar) and with steam at 325 °C (5 bar). The system was also washed with a sequence of acetone, methylene chloride, and acetone (5 mL each). No PM was extracted in the blank extraction performed after the wash of the system. Toxicity Assays. Murine macrophages (RAW 264.7) and human bronchial epithelial cells (BEAS-2B) were purchased from American Type Culture Collection (Manassas, VA). The RAW 264.7 cells were grown in tissue culture flasks (75 mL) in Dulbecco’s modified Eagle media (Fisher Scientific) with 4 mM glutamine, 3.7 g of NaHCO3, 4.5 g/L glucose, 10% (v/v) fetal calf serum, 1% (v/v) penicillin, and 1% (v/v) streptomycin in a 37 °C incubator with a 95% O2, 5% CO2 atmosphere. The BEAS-2B cells were grown in Collagen I precoated tissue culture flasks (ca. 5 µg per 1 cm2) in Lechner and LaVeck medium-9 (LHC-9) with supplements and growth factors (BPE, hydrocortisone, hEGF, epinephrine, insulin, triiodothyronine, transferrin, gentamicin/amphotericin-B, and retinoic acid) (Clonetics, Walkersville, MD). To transfer cells onto microplates, the cells were harvested with trypsin (0.05%) and 53 mM EDTA in Hank’s balanced salt solution (HBSS; Fisher Scientific) and centrifuged at low speed (400g for 5 min). To prevent any interference from media and serum, the cells were resuspended in fresh medium without serum and phenol red. GSH depletion and upregulation were evaluated by exposing murine macrophages and human bronchial epithelial cells to PM fractions or purified compounds for 12 h. Assays with RAW 264.7
Kuba´ toVa´ et al. cells were performed on three or four independent days (each time in quadruplicate in five 2-fold dilutions) in clear-bottom, white polystyrene 96 well microplates. The one-day quadruplicate experiment (single passage of cells) with BEAS-2B was performed in collagen I precoated microplates. On day one, culture cells were plated to each well, 180 µL of 106/mL of RAW 264.7 cells or 180 µL of 105/mL of BEAS-2B. After 24 h (48 h for BEAS-2B), 20 µL of sample was added. Intracellular GSH determinations were performed with a monochlorobimane (MCB) assay in which MCB is conjugated with GSH to form a fluorescent MCB-GSH adduct (30). DL-Buthionine-(S,R)-sulfoximine was used as a control for complete GSH depletion. The MCB-GSH was analyzed using a fluorescence microplate reader with an excitation wavelength of 385 nm and an emission wavelength of 485 nm (Molecular Devices Gemini XS, Molecular Devices, Sunnyvale, CA). Parallel to GSH depletion, cytotoxicity was monitored using lactate dehydrogenase (LDH), which is a marker for cell breakdown (leakage of cytosol). The LDH activity was determined by the CytoTox 96 NonRadioactive Cytotoxicity Assay purchased from Promega (Madison, WI) employing absorption at 490 nm, which was measured on a microplate reader (Spectromax plus 384, Molecular Devices). Toxicity assays were performed on nitrogen-dried extracts (to determine mass extracted) redissolved in dimethyl sulfoxide (DMSO, Fisher Scientific) and then HBSS. DMSO was used to transfer PM into aqueous suspension. The final DMSO concentration was 0.25 wt %. GSH depletion was expressed relative to the control consisting of 0.25% DMSO in HBSS. To evaluate the effect of DMSO, a separate quadruplicate experiment was performed in parallel without DMSO. Because the solubility in aqueous solution is limited, the samples were manipulated in a suspension. Before the toxicity assays, the pH of each extract was adjusted with 0.1 M NaOH to ∼7.4. Analyses. To perform GC analysis, aqueous fractions of wood smoke PM were extracted with methylene chloride employing U.S. EPA Method 625 for base/neutrals and acids extraction. Prior to the extraction with methylene chloride, internal standard (1chloronaphthalene) was added to each water fraction. To extract bases and neutrals, the pH of the water fractions was adjusted to >11 with 1 M NaOH. The water fractions were extracted by three 2 min liquid-liquid extractions with 10 mL aliquots of methylene chloride. To extract acids, the aqueous fractions were acidified to pH < 2 using ∼9 M sulfuric acid and extracted with methylene chloride. Methylene chloride extracts (from base/neutrals and acids extractions) of each fraction were combined and concentrated under a stream of nitrogen to ∼0.2 mL. GC/MS analyses were performed using a Hewlett-Packard model 5890 GC with a Hewlett-Packard model 5972 MS in the full-scan mode (45-500 m/z) with electron impact ionization. Chromatographic separations were accomplished with a 30 m DB-5 column with a 0.25 mm i.d. and a 0.25 µm film thickness (J&W Scientific, Rancho Cordova, CA) with injections in the splitless mode. The oven temperature was held at 40 °C for 0.2 min followed by a 10 °C/min gradient to 320 °C and then held for 20 min. The individual components in aerosol extracts were identified using computer matches to standard reference mass spectra of the National Institute of Standards and Technology (NIST) library and literature data (24). The identification was confirmed by reference standards of PAHs, oxy-PAHs, and dimethoxy- and methoxyphenols, etc. Tentative quantifications were performed using GC with flame ionization detection (FID) on a Hewlett-Packard model 5890 Series II GC equipped with an autosampler. Chromatographic separations were accomplished with a 50 m DB-5 column with a 0.25 mm i.d. and a 0.25 µm film thickness (J&W Scientific) with injections in the splitless mode. The oven temperature was held at 40 °C for 0.2 min followed by a 10 °C/min gradient to 320 °C and then held for 20 min. Statistical Analyses. Statistical comparisons were performed using one-way or two-way analysis of variance (ANOVA) for p < 0.05 using Prism software (GraphPad). Results are expressed as the means with one standard deviation.
Glutathione Depletion Caused by Wood Smoke
Chem. Res. Toxicol., Vol. 19, No. 2, 2006 257
Figure 1. Dose-response curves of GSH depletion expressed as % control MCB fluorescence (9) and cytotoxicity expressed as % LDH in cytosol (O) caused by combined water extract from wood smoke PM in RAW 264.7. The data presented are means ( one standard deviation of three quadruplicates performed on three different days. The symbols (*) and (§) mark statistically significant GSH depletion and/or LDH increase vs controls, respectively.
Results GSH Depletion by Wood Smoke PM. Wood smoke PM was extracted with hot pressurized water to obtain a wide polarity range of fractions. Initial experiments on a “crude” sample (combined fractions extracted from 25 to 300 °C) have shown significant GSH depletion at a concentration of 100 µg/ mL of PM (Figure 1). In contrast to the GSH depletion, the cytotoxicity (LDH) increased only at 2-fold higher PM concentrations of 200 µg/mL. The contribution of different polarity species was evaluated using GSH depletion and cytotoxicity assays in individual wood smoke PM fractions sequentially extracted with hot pressurized water at different temperatures (Figure 2). On the basis of the PM concentration used in the GSH assay, midpolarity and nonpolar fractions (extracted from 150 to 300 °C) caused the most pronounced GSH depletion at concentrations as low as 50 µg/mL in a dose-dependent manner. In contrast to GSH depletion, the cytotoxicity increased only for the highest concentration (200 µg/mL). To evaluate the potential of different polarity fractions to cause GSH depletion, the same dilutions (aliquots) of each fraction were compared based on the estimation of a dilution factor necessary for 50% GSH depletion. As for concentrationbased data (Figure 2), 50% GSH depletion of midpolarity and nonpolar fractions occurred at higher dilution factors than of polar fractions extracted at lower temperatures (Table 1). The production of ROS and, therefore, GSH depletion can be affected by hydroxyl scavengers such as DMSO (16, 31). Therefore, we have compared GSH depletion by different wood smoke PM fractions with and without DMSO (Figure 3). No statistically significant effect of DMSO (0.25% v/v) was observed. Murine macrophage RAW 264.7 is a commonly used model in PM research (9, 15, 16, 20). However, we confirmed the significance of our results using RAW 264.7 by the GSH depletion in human bronchial cells BEAS-2B. Depletion of GSH in BEAS-2B was comparable to that in RAW 264.7 (Figure 4). In contrast to RAW 264.7, none of the fractions extracted from wood smoke PM was cytotoxic to BEAS-2B cells (data not shown). Composition of Midpolarity Fraction. We have previously shown that polar and nonpolar species are extracted with hot pressurized water at lower and higher temperatures, respectively (28). A significant GSH depletion was observed in midpolarity
Figure 2. GSH depletion expressed as % control MCB fluorescence (9) and cytotoxicity expressed as % control LDH (O) in RAW 264.7 cells after 12 h of exposure to wood smoke PM fractions extracted with hot pressurized water from 25 to 300 °C. The data presented are means ( one standard deviation of four quadruplicates performed on four different days. The symbols (*) and (§) mark statistically significant GSH depletion and/or LDH increase vs controls, respectively. Table 1. Dilution Factor of Each Fraction (v/v Dilution of Each Extract) Necessary to Achieve 50% GSH Depletiona extraction temperature (°C) dilution factor
25
50
100
150
200
250
300
150
70
125
330
500
500
500
a
GSH depletion at lower concentrations (greater dilutions) demonstrates a higher potential to cause oxidative stress of midpolarity and nonpolar fractions (extracted from 150 to 300 °C).
and nonpolar fractions extracted from 150 to 300 °C. The toxicity of nonpolar PAHs is well-characterized (8). Thus, we have focused our further work on the identification of compounds in the midpolarity fraction extracted at 150 °C. Table 2 and Figure 5 compare the composition of different polarity fractions. On the basis of chromatograms (Figure 5), syringols were observed in both midpolarity and nonpolar fractions. However, the majority of syringols (6600 µg/g of wood smoke PM) were extracted at 25 °C fraction in contrast to 1600 µg/g at 150 °C and 500 µg/g in 250 °C (Table 2). Table 2 shows higher concentrations of oxy-PAHs, disyringyls, syringylguaiacyls, and PAHs in the midpolarity fraction than in any other fraction. Major identified oxy-PAHs were 9-fluorenone, 1-phenalenone, 9,10-anthraquinone, and hydroxycadalene. Lower concentrations of benzanthracenones and xanthones were observed as well. PAHs found in the midpolarity fraction were mainly represented by phenanthrene (27 wt % of all PAHs found in the fraction), fluoranthene (16 wt %), and pyrene (16 wt %).
258 Chem. Res. Toxicol., Vol. 19, No. 2, 2006
Kuba´ toVa´ et al.
Figure 3. GSH depletion (expressed as % control MCB fluorescence) dose-response curves in RAW 264.7 macrophages by wood smoke PM fractions extracted by hot pressurized water dissolved in HBSS (0) and dissolved 0.25% DMSO in HBSS (9). The data are presented as means + one standard deviation of head to head quadruplicate experiments performed on the same day.
Figure 4. Comparison of GSH depletion (% control MCB fluorescence) by different polarity fractions of wood smoke PM in human bronchial epithelial cells BEAS-2B (right) and murine macrophages RAW 264.7 (left). The fractions were obtained with hot pressurized water at temperatures 25 (9), 50 (0), 150 (2), and 250 °C (O). The data presented are means + one standard deviation of head to head experiments. Table 2. Distribution of Compounds (µg/g of Total Sample) Identified in the Midpolarity Fraction in Selected Temperature Fractions (28) fraction/extraction temperature (°C) polar 25 disyringyls and syringylguaiacyls oxy-PAHs PAHs syringols
midpolarity
nonpolar
100
150
250
30
340
580
30
120 5 6590
560 60 3420
1220 400 1630
150 180 510
Figure 5 and Table 2 demonstrate that a high concentration of PAHs in midpolarity fractions is mainly due to phenanthrene (eluting at 28 min). In a nonpolar fraction (extracted at 250 °C), the concentrations of oxygenated aromatics declined, while concentrations of higher molecular weight PAHs (eluting after 40 min) increased.
Figure 5. GC/MS (total ion current) chromatograms of wood smoke PM fractions extracted with hot pressurized water at 150 (top) and 250 °C (bottom), respectively. The peak labels are attributed as follows: IS, internal standard; S, syringols; O, oxy-PAHs; D, disyringyls and syringylguaiacyls; and P, PAHs; unlabeled peaks were either not evaluated for toxicity or not identified.
Besides previously identified disyringyls and syringylguaiacyls (2-4), new syringylguaiacyls and disyringyls were tentatively identified. These compounds were observed using the reconstructed ion chromatogram of m/z ) 194 (Supporting Information Figure 1a), the mass that is also the molecular weight and the mass spectral base peak of 4-allylsyringol. The tentative identifications were based on the lower mass range of the spectrum (50-194), which was comparable to the spectrum of 4-allylsyringol (Supporting Information Figure 2a) and molecular ions of 302, 316, 330, 332, 346, 360, and 386. The peaks were tentatively identified as derivatives of 1,2-disyringylethene and 1-guaiacyl-2-syringylethene (for proposed structures and mass spectra, see Supporting Information Figures 1b and 2a,b). The occurrence of peaks representing syringylguaiacyls and disyringyls was confirmed in the methylene chloride extracts of wood smoke PM (18 h of sonication). GSH Levels after Exposure to Oxy-PAHs and PAHs. On the basis of the identification above, GSH depletion caused by the midpolarity fraction (extracted with pressurized water at 150 °C) may be due to the presence of oxy-PAHs, PAHs, syringylguaiacyls, and/or disyringyls. To evaluate whether oxygenated polyaromatics or PAHs contribute most to GSH depletion in the midpolarity fraction, pure standards were tested for GSH depletion in RAW 264.7. Because syringylguaiacyls are not commercially available, the GSH assay was performed on selected oxy-PAHs and PAHs (Figure 6). Quinonic PAH derivatives are highly redox active molecules (12), and as expected, 1,8-dihydroxy-9,10-anthraquinone and 9,10-phenanthraquinone depleted GSH. In contrast to GSH depletion by wood smoke PM, several compounds elevated GSH content, among them 9,10-anthraquinone, anthrone, 1-hydroxypyrene, and heavier molecular weight PAHs (e.g., pyrene and
Glutathione Depletion Caused by Wood Smoke
Figure 6. Upregulation/depletion of GSH in RAW 264.7 cells by oxyPAHs (left) and PAHs (right). Oxy-PAHs: 9,10-anthraquinone (1), xanthone (9), 9-fluorenone (-), anthrone ([), 1,8-dihydroxyanthraquinone (b), 1-hydroxypyrene (2), and 9,10-phenathraquinone (f). PAHs: phenanthrene (4), benzo[e]pyrene (]), benzo[a]pyrene (3), 1-methylpyrene (0), and pyrene (O). The relative standard deviations from three independent days were within ∼20% for the highest concentrations and
