Intracellular S-Glutathionyl Adducts in Murine Lung and Human

Department of Environmental and Occupational Health, Department of ... and Department of Pharmacology, University of Pittsburgh, 260 Kappa Drive,...
0 downloads 0 Views 68KB Size
Download PDF
Chem. Res. Toxicol. 1999, 12, 931-936

931

Intracellular S-Glutathionyl Adducts in Murine Lung and Human Bronchoepithelial Cells after Exposure to Diisocyanatotoluene Robert W. Lange,† Billy W. Day,†,‡ Ranulfo Lemus,† Vladimir A. Tyurin,† Valerian E. Kagan,†,§ and Meryl H. Karol*,† Department of Environmental and Occupational Health, Department of Pharmaceutical Sciences, and Department of Pharmacology, University of Pittsburgh, 260 Kappa Drive, Pittsburgh, Pennsylvania 15238 Received March 17, 1999

Diisocyanatotoluene (toluene diisocyanate, TDI), a 4:1 mixture of 2,4- and 2,6-isomers used in the preparation of polyurethanes, causes occupational asthma by an as yet unknown mechanism. We previously showed that it forms adducts with the apical surface of the bronchoepithelium in vivo, and with ciliary microtubules in cultured human bronchoepithelial (HBE) cells. These results suggested that TDI may not enter HBE cells. In vitro studies, however, showed that TDI avidly forms bis adducts with glutathione (GSH) and that these adducts transfer monoisocyanato-monoglutathionyl-TDI to a sulfhydryl-containing peptide. This study sought to elucidate intracellular reactions of TDI. Using an electron paramagnetic resonance spectrometric (EPR) method, we established that the level of thiol-dependent quenching of phenoxyl radicals of etoposide was decreased >40% in pulmonary tissue of mice that received TDI intrabronchially. Similarly, HBE cells exposed to 100 ppb TDI vapor experienced a >30% reduction in thiol levels as determined with a thiol-specific fluorescent probe (ThioGlo 1). HPLC/UV analysis of lysates from HBE cells exposed to 200 and 500 ppb TDI vapor suggested a dose-related formation of S-glutathionyl adducts. Data from the 500 ppb TDI-treated HBE cells verified the identity of the 2-monoglutathionyl-4-monoisocyanato adduct. The results provide firm evidence that TDI enters pulmonary cells and reacts with GSH. This rapid reaction leading to formation of S-glutathionyl adducts of TDI suggests the importance of cellular thiols in TDI-induced pulmonary disease. Diisocyanatotoluene (toluene diisocyanate, TDI)1 is a reactive chemical widely used in the paint and plastic industries. It is most frequently employed as an 80:20 mixture of the 2,4- and 2,6-isomers. TDI is the most prevalent cause of occupational asthma (1), yet the pathogenesis of the disease remains unknown. The initial step in the sensitization process is believed to involve covalent reaction of one or both isomers of this bifunctional chemical with cellular protein(s). In guinea pigs, TDI adducts have been identified predominantly at the apical surface of the respiratory epithelium following inhalation exposure (2). More recently, we established colocalization of TDI with ciliary tubulin (3). Other sites of TDI binding observed in guinea pigs after inhalation exposure are to soluble proteins in bronchoalveolar lavage fluid (4) and with hemoglobin within erythrocytes (5). * To whom correspondence should be addressed: Center for Environmental & Occupational Health & Toxicology, 260 Kappa Drive, Pittsburgh, PA 15238. Phone: (412) 967-6530. Fax: (412) 967-6611. E-mail: [email protected]. † Department of Environmental and Occupational Health. ‡ Department of Pharmaceutical Sciences. § Department of Pharmacology. 1 Abbreviations: ALI, air-liquid interface; DFO, deferoxamine mesylate; D-PBS, Dulbecco’s phosphate-buffered saline; EDTA, ethylenediaminetetraacetate; EPR, electron paramagnetic resonance; GPx, glutathione peroxidase; GR, glutathione reductase; GSH, glutathione; GST, glutathione S-transferase; HBE, human bronchoepithelial; KSFM, keratinocyte-serum free medium; MIC, methyl isocyanate; SDS, sodium dodecyl sulfate; SMC, S-(N-methylcarbamoyl)cysteine; SMG, S-(N-methylcarbamoyl)glutathione; TDI, toluene diisocyanate; VP-16, etoposide.

To gain an understanding of the chemistry of TDI reactions with biomolecules in the primary site of entry, the lung, we previously studied the in vitro interaction of TDI with the most prevalent pulmonary nucleophile, glutathione (GSH). Under conditions of physiologic temperature, pH, and carbonate ion, reaction of TDI with 0.8 molar equiv of GSH leads to the bis(S-glutathionyl)thiocarbamoyl adduct as the favored reaction product (6). Less than 20% of the overall products are monoisocyanato adducts. Furthermore, incubation of the bis adduct of either TDI isomer with a single cysteine-containing MHC antigen-derived peptide results in rapid transcarbamoylation and adducted peptide. Likewise, the reaction of TDI with guinea pig hemoglobin results in the formation of carbamoylated protein both in vivo after inhalation and in vitro (5). In vivo and in vitro reactions of monoisocyanates with sulfhydryl-containing peptides and proteins have been reported (7, 8). Unfortunately, there is a paucity of information regarding protein specificity. Pearson’s report of primary (Cys-6) and secondary (Cys-1 and Tyr-2) sites of carbamoylation by S-linked glutathione conjugates of methyl isocyanate (MIC) on reduced oxytocin demonstrates that the preferred residue is cysteine. Furthermore, a significant reduction (g30%) of the free sulfhydryl content of the rat erythrocyte membrane occurs following in vitro or in vivo exposure to MIC (9). Cellular thiol modification may alter homeostatic regulatory mechanisms. Indeed, a 20-30% depletion in

10.1021/tx990045h CCC: $18.00 © 1999 American Chemical Society Published on Web 08/21/1999

932

Chem. Res. Toxicol., Vol. 12, No. 10, 1999

the level of cellular GSH by toxicants can result in impaired defense against xenobiotics and adverse health outcomes (10). In mice, a 25% reduction of GSH levels in antigen-presenting cells results in a Th2 phenotype which is characteristic of humoral immunity and allergy (11). These studies suggest that the thiol status of a cell may influence its susceptibility not only to TDI conjugation but also to sequelae of TDI exposure. In this study, we tested the hypothesis that TDI reacts with pulmonary thiols. Using a variety of methods, we determined thiol levels and reactivity in murine lung and human bronchoepithelial (HBE) cells after TDI exposure, and we identified some of the chemical adducts that were produced. The rapid formation of adducts suggests the importance of cellular thiols in the pathogenesis of TDIinduced pulmonary disease.

Experimental Procedures Caution: TDI is a skin and mucus membrane irritant, and may cause respiratory sensitization. It should be handled in a chemical vapor hood with appropriate protective equipment. Its hydrolysis products are aromatic amines, potential human carcinogens. Chemicals. Chemicals listed without a supplier were purchased from Sigma Chemical Co. (St. Louis, MO). Bayer, USA (Pittsburgh, PA), provided TDI (4:1 molar mixture of 2,4- and 2,6-isomers). Solvents with the highest available purity were purchased from Fisher Scientific (Pittsburgh, PA). Conjugates of TDI were synthesized as previously described (6). Exposure of Mice to TDI. Female, 60-day-old C57BL/6 mice (Hilltop Lab Animals, Inc., Scottdale, PA) were treated with 5 µg of TDI in 5 µL of corn oil intrabronchially as previously described (12). The control group received corn oil alone. The Institutional Animal Care and Use Committee, University of Pittsburgh, preapproved all procedures. Twenty-four hours following TDI exposure, animals were euthanized. The lungs were resected, perfused with cold D-PBS, and homogenized with five 5 s pulses using a Tissue-Tearer (BioSpec Products, Inc., Bartlesville, OK). The total protein concentration in lung homogenates was determined with a commercial Bradford assay (Bio-Rad, Richmond, CA). Human Bronchial Epithelial Cell Culture. Cell culture reagents, unless otherwise noted, were purchased from Gibco (Gaithersburg, MD). J. Pilewski (University of Pittsburgh Medical Center) provided primary cultures of HBE cells. The cells are noncancerous, surgical specimens from lung transplant recipients isolated as previously described (13). HBE cells were cultured submerged in keratinocyte-serum free medium (KSFM) supplemented with the following: epidermal growth factor (5 ng/mL), bovine pituitary extract (50 µg/mL), Fungizone (1% w/v), penicillin (100 units/mL)/streptomycin (100 µg/mL), and L-glutamine (2 mM). Cells cultured in this fashion represent undifferentiated populations. To generate differentiated HBE cells, passage 2 HBE cells were detached from culture flasks (0.05% trypsin with EDTA) and seeded at a density of 8 × 104 cells/cm2 on a 12 or 24 mm diameter, clear Transwell tissue culture inserts (Corning-Costar, Grand Island, NY) in K-SFM. Cells were fed with 1 and 1.5 mL of K-SFM in the apical and basal compartments, respectively. The apical medium was removed when HBE cells reached confluence. The basal medium was replaced with a 1:1 mixture of D-MEM and F-12 culture medium supplemented with 2% ULTROSER G (BioSepra S. A.), Fungizone, penicillin, streptomycin, and L-glutamine as above, henceforth termed the air-liquid interface (ALI) medium. ALI culture favors the development and retention of epithelial features, including bioelectric properties, mucin secretion, and ciliagenesis (14). HBE cells were incubated for 14-21 days at the ALI (15) prior to TDI exposure. The medium was changed on alternate days.

Lange et al. Exposure of HBE Cells to TDI and Collection. Submerged, adherent undifferentiated HBE cell monolayers in glass 28 cm Petri dishes were prepared for TDI exposure by washing cultures and replenishing with 1 mL of fresh medium. The Transwells with the HBE cell monolayer were placed onto a stainless steel sieve held in a 28 cm glass Petri dish which contained 1 mL of medium. ALI cells remained hydrated during the exposure by wicking of the medium through the sieve. The dishes containing undifferentiated or differentiated HBE cells were placed on a platform in a 70 L glass chamber and exposed to 100-500 ppb TDI for 40 min. Samples of differentiated HBE cells dedicated to LC/MS analyses were exposed for 1 h. TDI vapor was generated by passing dried air through a glass impinger containing 4 mL of TDI as previously described (4). The chamber air was exhausted at a rate of 60 L/min. Realtime chamber atmosphere monitoring was performed using an Autostep continuous toxic gas analyzer (Bacharach, Inc., Pittsburgh, PA) with its probe placed ca. 5 cm above the culture dishes. Control, air-exposed cells were placed in a similar tank with water in the impinger. Following exposure, HBE cells were wetted with D-PBS (0.5 mL) and scraped from the filters using a rubber policeman. Cells and subsequent washes (approximately 1 mL) were centrifuged and washed, and then stored at -70 °C. Electron Paramagnetic Resonance (EPR) Assay of Radical Scavenging by Thiols in Tissue and Cells. Reactivity of thiols (GSH and protein thiols) toward radicals was assayed by determining the course and amplitude of the oxidative/reductive conversions of the exogenous VP-16 phenoxyl radical in TDI-treated mouse lung homogenates or homogenates of undifferentiated HBE cells (16). In the experimental system, an EPR-detectable steady-state level of VP-16 phenoxyl radical is generated and a metered aliquot of the sample to be probed is added. Reductants in the sample cause the disappearance of the VP-16 phenoxyl radical signal. The time course for the reappearance of this signal (the lag period) is proportional to the concentration of low-molecular mass (primarily ascorbate and GSH) reductants in the added sample. Intact tissue contains ascorbate, which is the first reductant consumed during the lag period and can be detected in EPR spectra as its one-electron oxidized form, the semidehydroascorbyl radical. However, the contribution that ascorbate makes to the duration of the lag period is small compared to that contributed by endogenous GSH. Ascorbate is not present in standard cell culture systems. Hence, in the presence of cell homogenates, the time course for the reappearance of VP-16 radical in an EPR spectrum is mainly dependent on the intracellular concentration of GSH and protein thiols (17). In the absence of cell or tissue homogenates, this system permits persistent detection of the characteristic VP-16 phenoxyl radical EPR signal for 50-60 min. For measurement of VP-16 phenoxyl radical formation, VP16 (0.7 mM) and tyrosinase (8 units) were incubated in 50 mM Na2PO4 buffer (pretreated with Chelex-100 to remove possible transition metal ions) containing 0.1 M NaCl and 100 mM deferoxamine mesylate (DFO) (pH 7.4) at 25 °C in the presence or absence of homogenates. Tyrosinase was predissolved in distilled H2O, while VP-16 was predissolved in DMSO. Lung or undifferentiated HBE cell homogenates were added (time zero). The lag period was then identified as the period of time prior to the reappearance of the VP-16 phenoxyl radical signal (arbitrarily chosen as 2 times the background noise) (16). EPR measurements were performed on an RE1X spectrometer (JEOL USA Inc., Peabody, MA) at 25 °C in gas-permeable Teflon tubing (0.8 mm i.d., 0.013 mm o.d., 8 cm length) from Alpha Wire Corp. (Elizabeth, NJ). The tube was filled with 60 µL of mixed sample, folded into quarters, and placed in an open, 3 mm i.d., EPR quartz tube such that all of the sample was within the effective microwave irradiation area. Spectra of VP-16 were recorded at a center field of 335.5 mT, a power of 20 mW, a field modulation of 0.04 mT, a sweep width of 5 mT, a receiver gain of 500, and a time constant of 0.03 s.

S-Glutathionyl Adducts following TDI Exposure Quantitation of Cellular GSH and Protein Thiol Levels. Cellular GSH and protein thiols in differentiated HBE cells were titrated using the thiol-specific fluorophore ThioGlo 1 (Covalent Associates, Inc., Woburn, MA) (18, 19). Briefly, HBE pellets were thawed on ice, resuspended in 0.5 mL of 50 mM Na2PO4 buffer (pH 7.4) containing 0.1 M NaCl, and sonicated at 0 °C with two 15 s pulses using an Ultrasonic homogenizer 4710 tip sonicator (Cole-Parmer Instrument Co., Chicago, IL). The total amount of protein was determined using the Bradford assay. For analysis of thiols in each sample, a quartz cuvette containing 2.5 mL of 50 mM Na2PO4 buffer (pH 7.4) was positioned in a spectrofluorometer (model RF-5301 PC, Shimadzu Scientific Instruments, Columbia, MD) equipped with a continuous mixer. ThioGlo 1 was added (2.5 µL of a 10 mM stock solution). The solution was mixed with a plastic Pasteur pipet, and the fluorescence at 500 nm was recorded. Following stabilization of baseline fluorescence, a 30 µg (total protein) aliquot of sample was added to the cuvette. Instantaneous fluorescence represented emission from ThioGlo 1-adducted GSH. When the instantaneous emission stabilized, 100 µL of 1 mM sodium dodecyl sulfate (SDS) was added to denature proteins. After output stabilization (∼10 min), the resulting fluorescence represented emission of ThioGlo 1-adducted total cellular thiols. Subtraction of the instantaneous fluorescence from the maximum fluorescence gave the cellular protein thiol levels (nanomoles per milligram). A standard curve of GS-adducted ThioGlo 1 fluorescence was prepared for determination of thiol content in the samples. Validation of the specificity of the ThioGlo 1 assay for GSH and protein sulfhydryls was performed with cumene hydroperoxide/GSH peroxidase. The magnitude of the fluorescent response to GSH was reduced by >95%, whereas the magnitude of the response evoked by SDS-induced protein unfolding was unaffected. HPLC and Mass Spectrometry. Analytical HPLC was performed on a Hewlett-Packard 1090 Series II liquid chromatograph equipped with a Hewlett-Packard 1040 UV-vis diode array detector (DAD), both controlled by a HewlettPackard Series 300 Chemstation. Separations were performed with a 3 µm particle size, 100 mm × 2.0 mm, Luna C18 column (Phenomenex, Torrance, CA) using an isocratic mobile phase of 21:4 H2O/CH3CN containing 0.1% CF3CO2H flowing at a rate of 0.1 mL/min. The column effluent was monitored at 250 nm, and spectra over the range of 200-400 nm were recorded every 1.28 s. A Perkin-Elmer/Sciex APII mass spectrometer with an atmospheric pressure ionization source and an articulated IonSpray interface linked in tandem with glass capillary tubing to the HPLC was employed to determine the molecular masses of analytes. The pelleted HBE cell lysates were redissolved in 50 µL of the mobile phase and sonicated twice for 15 s on ice with the tip sonicator, and then centrifuged at 10000g for 20 min at 4 °C to remove debris. The supernatant was transferred to a new tube and again centrifuged. Twenty-five microliters of the supernatant was injected onto the column. Column effluent traveled through the DAD and was introduced without splitting into the ionization source of the mass spectrometer and nebulized with highly pure air at an operating pressure of 40 psi. Highly pure N2 heated to 55 °C and flowing at a rate of 0.6 L/min was used as the curtain gas. The IonSpray interface was maintained at a voltage of 5 kV and the orifice at 70 V. The quadrupole was scanned over the range of m/z 100-900 at a rate of 10 s/scan at a resolution of m/z 0.1. Raw m/z data were collected and analyzed with the Sciex MacSpec program.

Results EPR Analysis of Thiol Reactivity toward VP-16 Phenoxyl Radicals. To assess TDI-induced disruption of epithelial redox status, homogenates from TDI-treated mouse lungs and HBE cells were analyzed for their ability to scavenge the tyrosinase-generated VP-16 phenoxyl radical. The typical spectrum of the ascorbate

Chem. Res. Toxicol., Vol. 12, No. 10, 1999 933

Figure 1. Representative EPR spectra and the time courses for the appearance of the semidehydroascorbyl and VP-16 radicals in the presence and absence of mouse lung. (A) Typical EPR spectrum of semidehydroascorbyl (left) and VP-16 (right) radicals. (B) Time course of the appearance of VP-16 radical in control mouse lung homogenates. Disappearance of semidehydroascorbyl radical (0) (3.5 min) occurs prior to the detection of the VP-16 radical (O) (8.5 min). (C) Time course of the appearance of semidehydroascorbyl radical (0) and VP-16 radical (O) in mouse lung homogenate 24 h after receiving 5 µg of TDI in vivo. The appearance of the VP-16 radical signal (O) at 5.0 min in samples exposed to TDI, compared with its appearance at 8.5 min in control samples (panel B), indicates a >40% reduction in the level of intracellular thiols.

radical is shown in Figure 1A (left) and that of the VP16 phenoxyl radical in Figure 1A (right). When control lung homogenate was added to the VP-16/tyrosinase system (Figure 1B), a typical doublet signal of ascorbate radical could immediately be discerned in the spectrum. The signal was transient and disappeared following incubation for approximately 3.5 min. No signals were observed in the EPR spectra during the subsequent ∼5 min. A typical signal of VP-16 phenoxyl radical became apparent in the spectrum over the next 1-1.5 min, and its magnitude increased very slowly during the following 5-7 min. The presence of ascorbate radical was due to one-electron oxidation of ascorbate by VP-16/tyrosinasederived phenoxyl radicals (17). The time interval during which no EPR signals could be detected (∼5 min) corresponds to thiol-dependent reduction of the phenoxyl radicals. After depletion of both endogenous ascorbate and thiols, the VP-16 phenoxyl radical EPR signal became evident. In the presence of homogenates from TDI-exposed lungs, the ascorbate-dependent reduction of VP-16 phenoxyl radicals and persistence of the ascorbate radical EPR signal (3.5 min) did not differ when compared with those of homogenates from control lungs (Figure 1C). By contrast, the thiol-dependent quenching of the VP-16 radicals was dramatically shorter, and the period of time between the disappearance of the ascorbate radical signal and the appearance of the VP-16 radical signal constituted only ∼1.5 min (Figure 1C). On average, the total lag period for the appearance of VP-16 phenoxyl radical produced by control mouse lung homogenates (400 µg of total protein) was 8.5 ( 0.5 min (Figure 1B), 3.75 ( 0.5 min of which could be ascribed to ascorbate (detected as the semidehydroascorbyl radical, Figure 1A). In the samples obtained from TDI-treated animals, the total lag period was reduced 40% to 5.0 ( 0.5 min (Figure 1C). This effect was entirely due to a decreased capacity of thiols to interact with VP-16 phenoxyl radicals, as the

934

Chem. Res. Toxicol., Vol. 12, No. 10, 1999

Lange et al.

Table 1. EPR Determination of the Effect of TDI on Thiols Capable of Reducing the VP-16 Phenoxyl Radical in Undifferentiated HBE Cells TDI (ppb)

lag perioda (min)

0 100

7.5 ( 0.5 3.5 ( 0.5

thiols oxidized by VP-16 phenoxyl radical nmol/1 × 106 cells

nmol/mg of protein

14.6 ( 1.2 6.9 ( 1.1

83.4 ( 3.9b 41.5 ( 6.6

a Duration of the lag period indicates the amount of thiols present. b Values represent the mean ( SD of triplicate determinations.

lifetime of semidehydroascorbyl radicals (3.5 ( 0.5 min) was not changed by TDI exposure. The undifferentiated HBE cells exposed to 100 ppb TDI vapor exhibited a >50% reduction in the duration of the lag period compared to control HBE cells (Table 1). Standard cell culture media do not contain ascorbate (vitamin C); hence, no ascorbate-dependent quenching of VP-16 phenoxyl radicals occurred in the presence of HBE cell homogenates. As expected, no lag period was observed by the addition of the buffer alone in the absence of either homogenate preparation. The viability of cells following TDI exposure exceeded 85% (data not shown). Fluorescent Determination of the Levels of Thiol Groups in HBE Cell Homogenates. A maleimidecontaining, thiol-labeling, fluorescent reagent, ThioGlo 1, was employed to determine the effect of TDI on thiol levels in differentiated HBE cells and to compare this effect with that of EPR-elucidated thiol-dependent radical scavenging. The reagent also allowed us to distinguish GSH from protein thiols as the nucleophiles involved in TDI binding. ThioGlo 1 emits low fluorescence prior to thiol addition and very high fluorescence after reaction with active SH groups (ca. 50-fold increase). Both lowmolecular mass and protein-associated thiolates (R-S-) are thought to form fluorescent adducts with ThioGlo 1 (18). A typical graph of the fluorescence emission resulting from the reaction of ThioGlo 1 with GSH and protein thiols in HBE cells is displayed in Figure 2. The reaction of ThioGlo 1 with GSH and other small thiols is complete within seconds. Addition of SDS (unfolding of proteins) yields an additional fluorescence response that increases with time (Figure 2). This second fluorescence response was protein-thiol-dependent. The amplitude of emission 10 min after the addition of SDS was chosen as a measure of total thiol content (GSH and protein sulfhydryls). HBE homogenates exposed to 100 ppb TDI exhibited a 32% reduction in the level of ThioGlo 1-reactive low molecular mass thiols (Table 2). Adduction of highmolecular mass reactive cellular protein thiols by ThioGlo 1 occurred after the reaction with the low-molecular mass thiols. To ensure reaction of ThioGlo 1 with remote or buried cysteines, SDS was added to the cuvette after the first phase of the reaction had plateaued. TDI did not reduce cellular protein thiols in HBE cells (Table 2). The reduction in the first, but not in the second, phase of the ThioGlo 1 reaction indicates that GSH is the primary thiol target of TDI in differentiated HBE cells. The viability of the cells following TDI exposure exceeded 85% (data not shown). HPLC/MS Analysis of (GS)TDI Conjugates. Glutathionyl (GS) adducts of TDI were synthesized as analytical standards using previously described methods (6). Lyophilized standards and differentiated HBE cell

Figure 2. Representative fluorescence responses of ThioGlo 1 upon addition of HBE cell homogenates and following addition of SDS. Emission plateaus (from time zero) represent baseline fluorescence (following addition of ThioGlo 1), GSH and lowmolecular mass thiols (following addition of cell homogenates), and high-molecular mass thiols (following SDS-induced denaturation). Cell homogenates were added following the stabilization of ThioGlo 1 baseline fluorescence. Emission was monitored for 2 h. Since no significant change in the profile was obtained after 10 min, final measurements were recorded at 10 min. A standard curve of GS-adducted ThioGlo 1 fluorescence was prepared for quantification of the thiol content in the samples. Table 2. Fluorometric (ThioGlo 1, a maleimide-containing fluorescent indicator) Determination of the Effect of TDI on GSH and Protein Thiols in Differentiated HBE Cells

b

TDI (ppb)

maleimide-reactive glutathionyl species (nmol/mg of cell protein)

maleimide-reactive protein thiols (nmol/mg of cell protein)

0 100

15.7 ( 0.04a 10.6 ( 0.86b

43.1 ( 3.7 48.4 ( 5.0

a Values represent the mean ( SD of triplicate determinations. p < 0.05 vs 0 ppb control (Student’s t test).

homogenates were resuspended in the mobile phase and analyzed by HPLC with tandem UV (250 nm) and pneumatically assisted electrospray MS detection. The retention times for the 2,6- and 2,4-(GS)TDI adduct standards were ca. 8 and 15 min, respectively. For each isomer, the mono (GS)TDI adduct elutes just prior to the bis adduct, hence the broad peaks. HPLC/UV analysis of lysates from HBE cells exposed to 200 and 500 ppb TDI suggested the dose-dependent presence of the 2,4-S-glutathionyl adducts (Figure 3). The presence of adducts in these cells was further indicated by the near identity of their UV spectra with those from the synthetic standards (inset of Figure 3). The HPLC eluates from the (GS)TDI standards and HBE cell samples were further analyzed by on-line pneumatically assisted electrospray mass spectrometry. Although levels of the bis adducts in the HBE cell samples were too low for MS detection, spectra from the 500 ppb TDI-treated HBE cells verified the presence of the 2-monoGS-4-monoNCO form of the adduct (Figure 4), which elutes just prior to the bis adduct. We have previously noted a much higher molar detector response of the mono adducts versus that of the bis adducts in the MS system that was used.

Discussion Our previous studies have found adducts on external surfaces of bronchial cells from animals exposed to TDI

S-Glutathionyl Adducts following TDI Exposure

Figure 3. HPLC chromatograms of 2,6- and 2,4-bis(GS)TDI adducts from HBE cell homogenates exposed to 200 ppb TDI (s), 500 ppb TDI (- - -), or the standard mix (- - -). Each chromatogram is representative of the product from ca. 1.2 × 107 homogenized cells suspended in 50 µL of the mobile phase (25 µL injected). (Top) UV spectrum of the peak appearing at a retention time of ca. 15 min, indicating its similarity to 2,4-bis(GS)TDI.

Figure 4. Electrospray mass spectrum of the HPLC-fractionated, 2-monoGS-4-monoisocyanato adduct (inset), with a retention time of ca. 15 min, from HBE cell homogenates exposed to 500 ppb TDI vapor for 1 h.

vapor (2). More recent results (3) implicated adduct formation with tubulin in the cilia of bronchial cells, suggesting that TDI may form adducts with structural/ skeletal and insoluble cellular targets. We also found TDI adducts on intracellular hemoglobin, indicating that an isocyanato or a masked form of the isocyanato group(s)

Chem. Res. Toxicol., Vol. 12, No. 10, 1999 935

of TDI survives passage from the lung into erythrocytes (5). Furthermore, in vitro and in vivo studies indicated that glutathione may act as a carrier for TDI (6) and monoisocyanates (20). Thus, we sought to investigate the intracellular consequences of cell exposure to TDI. Since GSH is the most important intracellular reducing agent, and accounts for 90% of the intracellular nonprotein thiols (22), we examined modification of GSH and other cellular thiols within HBE cells. Two kinetic analytical methods were used to quantify TDI-induced modification of intracellular thiols: (1) the EPR tracking of thiol-dependent VP-16 phenoxyl radical quenching and (2) the fluorometric determination of thiols with ThioGlo 1. Results show that TDI significantly decreased the endogenous thiol capacity of pulmonary tissue and HBE cells to scavenge phenoxyl radicals. To compare TDI-induced thiol depletion by the two methods, one has to consider that the culture conditions of the HBE cells evaluated by EPR differed from those evaluated by the ThioGlo 1 method. Differentiated control cells (ThioGlo 1 method) contain approximately 30% of the total thiols observed in undifferentiated control cells (EPR method). In addition, cells analyzed by EPR were normalized by absolute number and by protein content, whereas the ThioGlo 1 samples were normalized by total cellular protein content since much less homogenate was used for analysis. The variation in the extent of TDIinduced GSH depletion in differentiated versus undifferentiated HBE cells (33 vs 50%) is likely the collective effect of the aforementioned factors. In light of these differences, results obtained with the two methods are remarkably similar. By analyzing murine lung as well as human airway epithelial cells, we have evaluated the interaction of TDI with various lung tissues. We found comparable cellular thiol changes in these lung tissues, i.e., 40-50% reduction in the level of thiols (Figure 1 and Table 1). TDI did not react with ascorbate, nor did it cause depletion of ascorbate by radical-induced mechanisms, but specifically interacted with thiols. To our knowledge, the mass spectrometric data shown here are the first definitive identification of a TDI adduct in human tissue following an acute exposure to TDI vapor. Identification of specific molecular targets of TDI in the airways is crucial to understanding the mechanism of TDI-induced pulmonary disease. In vitro, TDI reacts readily with GSH without the need for enzymatic assistance (6). Further, incubation of approximately equimolar quantities of TDI with GSH leads almost exclusively to the bis(GS) adduct. It is not known if such adducts form in humans following TDI exposure. Results reported here provide evidence that thiol-containing molecules of human pulmonary epithelial cells form adducts with TDI and that carbamoylation of glutathione is the primary thiol reaction. TDI, Thiols, and Lung Disease. One of the frequent findings in TDI asthma is a disrupted or denuded airway epithelium (1). In guinea pigs, TDI reacts with, and is retained by, the airway epithelium (2). We have confirmed colocalization of TDI with ciliary tubulin (3). The ca. 20 free sulfhydryls on both the free and microtubular R/β tubulin heterodimer make it an appropriate target for TDI. Binding of TDI with the many available sulfhydryls of ciliary tubulin may be responsible for cell damage and sloughing observed in TDI-induced lung diseases.

936

Chem. Res. Toxicol., Vol. 12, No. 10, 1999

GSH deficiency in the epithelial lining fluid of the lung has been correlated with significant membrane injury and exacerbation of chronic pulmonary disease (23, 24). In this study, TDI-induced cellular redox and thiol changes have been confirmed. Recent evidence has implicated a lowered cellular GSH content in promoting the Th2 lymphocyte phenotype that is characteristic of allergy (11). GSH-dependent enzymes of airway epithelium include glutathione reductase (GR), glutathione S-transferase, and GSH peroxidase. The close association of GSH with these enzymes suggests that they may be targets of (GS)TDI adducts via transfer reactions. The propensity of TDI to carbamoylate nucleophilic sulfur and the resulting thiocarbamate’s instability at physiological pH suggest that transcarbamoylation may occur (6). Indeed, the carbamoylating metabolites of methyl isocyanate, S-(Nmethylcarbamoyl)glutathione (SMG) and S-(N-methylcarbamoyl)cysteine (SMC), have been shown to significantly inhibit GR in vitro (25). Further, N,N′-bis(2chloroethyl)-N-nitrosourea (BCNU) has been reported to carbamoylate GR (26). If GSH-dependent enzymes are subject to modification by TDI, the result may be altered cellular redox status and increased cellular susceptibility to injury and disease. In summary, significant TDI-induced thiol modification in HBE cells suggests that TDI asthma may be GSHdependent. Thiol depletion would impair pulmonary free radical scavenging and increase the sensitivity of the lung to TDI and other toxicants. Evidence is presented here that TDI-GSH adduct formation occurs at TDI concentrations reported in occupational spills. Whether TDI sensitization involves formation of an immunogenic TDIadducted antigen or modulation of enzyme activity, the transcarbamoylation reactions of S-glutathionyl adducts of 2,4- and 2,6-diisocyanatotoluene provide a mechanism for protein modification. Future studies will examine the specificity of TDI modification of GSH-dependent enzymes in mammalian airway epithelium.

Acknowledgment. This work was supported by NIH grants (ES05651 to M.H.K. and ES/HL09371 to B.W.D.) and the NCI Oncology Research Faculty Development Program (V.A.T.). We are grateful to Joseph D. LaToche for his technical expertise in isolating and propagating the HBE cells.

References (1) Mapp, C. E., Saetta, M., Maestrelli, P., Di Stefano, A., Chitano, P., Boschetto, P., Ciaccia, A., and Fabbri, L. M. (1994) Mechanisms and pathology of occupational asthma. Eur. Respir. J. 7, 544-554. (2) Karol, M. H., Jin, R., and Lantz, R. C. (1997) Immunohistochemical detection of toluene diisocyanate (TDI) adducts in pulmonary tissue of guinea pigs following inhalation exposure. Inhal. Toxicol. 9, 63-83. (3) Lange, R. W., Lantz, R. C., Beer-Stolz, D., Watkins, S. C., Sundareshan, P., Lemus, R., and Karol, M. H. (1999) Toluene diisocyanate colocalizes with tubulin on cilia of differentiated human airway epithelial cells. Toxicol. Sci. 50, 64-71. (4) Jin, R., Day, B. W., and Karol, M. H. (1993) Toluene diisocyanate (TDI) protein adducts in the bronchoalveolar lavage of guinea pigs exposed to vapors of the chemical. Chem. Res. Toxicol. 6, 906912. (5) Day, B. W., Jin, R., and Karol, M. H. (1996) Reactions of toluene diisocyanate isomers with guinea pig hemoglobin in vitro and in vivo. Chem. Res. Toxicol. 9, 568-573. (6) Day, B. W., Jin, R., Basalyga, D. M., Kramarik, J. A., and Karol, M. H. (1997) Formation, solvolysis, and transcarbamoylation reactions of bis(S-glutathionyl) adducts of 2,4- and 2,6-diisocyanatotoluene. Chem. Res. Toxicol. 10, 424-431.

Lange et al. (7) Pearson, P. G., Slatter, J. G., Rashed, M. S., Han, D. H., and Baillie, T. A. (1991) Carbamoylation of peptides and proteins in vitro by S-(N-methylcarbamoyl)glutathione and S-(N-methylcarbamoyl)cysteine, two electrophilic S-linked conjugates of methyl isocyanate. Chem. Res. Toxicol. 4, 436-444. (8) Kassahun, K., Jochheim, C. M., and Baillie, T. A. (1994) Effect of carbamate thioester derivatives of methyl- and 2-chloroethyl isocyanate on glutathione levels and glutathione reductase activity in isolated rat hepatocytes. Biochem. Pharmacol. 48, 587594. (9) Bhattacharya, B. K., Sharma, S. K., and Jaiswal, D. K. (1996) Binding of [1-14C]methyl isocyanate to erythrocyte membrane proteins. J. Appl. Toxicol. 16, 137-138. (10) Reed, D. J., and Fariss, M. W. (1984) Glutathione depletion and susceptibility. Pharmacol. Rev. 36, 25s-33s. (11) Peterson, J. D., Herzenberg, L. A., Vasquez, K., and Waltenbaugh, C. (1998) Glutathione levels in antigen-presenting cells modulate Th1 versus Th2 response patterns. Proc. Natl. Acad. Sci. U.S.A. 95, 3071-3076. (12) Ebino, K., Kramarik, J., Lemus, R., and Karol, M. H. (1998) A mouse model for study of localized toluene diisocyanate adducts following intrabronchial administration of the chemical: inflammation and antibody production. Inhal. Toxicol. 10, 503-529. (13) Mette, S. A., Pilewski, J., Clayton, C. A., and Albelda, S. M. (1993) Distribution of integrin cell adhesion receptors on normal bronchial epithelial cells and lung cancer cells in vitro and in vivo. Am. J. Respir. Cell Mol. Biol. 8, 562-572. (14) Gray, T. E., Guzman, K., Davis, C. W., Abdullah, L. H., and Nettesheim, P. (1996) Mucociliary differentiation of serially passaged normal human tracheobronchial epithelial cells. Am. J. Respir. Cell Mol. Biol. 14, 104-112. (15) Whitcutt, M. J., Adler, K. B., and Wu, R. (1988) A biphasic chamber system for maintaining polarity of differentiation of cultured respiratory tract epithelial cells. In Vitro Cell Dev. Biol. 24, 420-428. (16) Gantchev, T. G., van Lier, J. E., Stoyanovsky, D. A., Yalowich, J. C., and Kagan, V. E. (1994) Interaction of phenoxyl radical of antitumor drug, etoposide, with reductants in solution and in cell and nuclear homogenates: electron spin resonance and high performance liquid chromatography. Methods Enzymol. 234, 631642. (17) Kagan, V. E., Yalowich, J. C., Day, B. W., Goldman, R., and Stoyanovsky, D. A. (1994) Ascorbate is the primary reductant of the phenoxyl radical of etoposide (VP-16) in the presence of thiols both in cell homogenates and in model systems. Biochemistry 33, 9651-9660. (18) Langmuir, M. E., Yang, J.-R., Le Compte, K. A., and Durand, R. E. (1996) New thiol active fluorosphores for intracellular thiols and glutathione measurement. In Fluorescence Microscopy and Fluorescent Probes (Slavik, J., Ed.) pp 229-233, Plenum Press, New York. (19) Storey, B. T., Alvarez, J. G., and Thompson, K. A. (1998) Human sperm glutathione reductase activity in situ reveals limitation in the glutathione antioxidant defense system due to supply of NADPH. Mol. Reprod. Dev. 49, 400-407. (20) Baillie, T. A., and Slatter, J. G. (1991) Glutathione: a vehicle for the transport of chemically reactive metabolites in vivo. Acc. Chem. Res. 24, 264-270. (21) Huffman, L. J., Judy, D. J., Frazer, D., Shapiro, R. E., Castronova, V., Billie, M., and Dedhia, H. V. (1997) Inhalation of toluene diisocyanate is associated with increased production of nitric oxide by rat bronchoalveolar lavage cells. Toxicol. Appl. Pharmacol. 145, 61-67. (22) Deneke, S. M., and Fanburg, B. L. (1989) Regulation of cellular glutathione. Am. J. Physiol. 257, L163-L173. (23) Thanislass, J., Raveendran, M., Sivasithamparam, N., and Devaraj, H. (1996) Effect of chronic glutathione deficiency on rat lung mitochondrial function. Pulm. Pharmacol. 9, 119-122. (24) Cantin, A. M., Hubbard, R. C., and Crystal, R. G. (1989) Glutathione deficiency in the epithelial lining fluid of the lower respiratory tract in idiopathic pulmonary fibrosis. Am. Rev. Respir. Dis. 139, 370-372. (25) Jochheim, C. M., and Baillie, T. A. (1994) Selective and irreversible inhibition of glutathione reductase in vitro by carbamate thioester conjugates of methyl isocyanate. Biochem. Pharmacol. 47, 1197-1206. (26) Vanhoefer, U., Yin, M.-B., Harstrick, A., Seeber, S., and Rustum, Y. M. (1997) Carbamoylation of glutathione reductase by N,Nbis(2-chloroethyl)-N-nitrosourea associated with inhibition of multidrug resistance protein (MRP) function. Biochem. Pharmacol. 53, 801-809.

TX990045H