Toluene Diisocyanate Protein Adducts in the Bronchoalveolar Lavage

To investigate the process of toluene diisocyanate (TDI) sensitization, studies were conducted to identify TDI-protein adducts in the bronchoalveolar ...
0 downloads 0 Views 2MB Size
906

Chem. Res. Toxicol. 1993,6, 906-912

Toluene Diisocyanate Protein Adducts in the Bronchoalveolar Lavage of Guinea Pigs Exposed to Vapors of the Chemical Ruzhi Jin,t Billy W. Day,tJ and Meryl H. Karol’9t Department of Environmental and Occupational Health and Department of Pharmaceutical Sciences, Center for Environmental and Occupational Health and Toxicology, University of Pittsburgh, 260 Kappa Drive, Pittsburgh, Pennsylvania 15238 Received June 7, 199P

To investigate the process of toluene diisocyanate (TDI) sensitization, studies were conducted to identify TDI-protein adducts in the bronchoalveolar lavage (BAL) fluid of guinea pigs exposed to a sensitizing atmosphere of the commercially used 4:l mixture of 2,4- and 2,6-TDI. Animals were exposed to 2 ppm TDI for 3 h. Immediately thereafter lungs were lavaged. TDI-modified proteins in the lavage fluid were identified by immunologic staining with a highly sensitive and specific rabbit antiserum raised to a TDI-keyhole limpet hemocyanin (TDI-KLH) conjugate. The sensitivity of the antiserum was demonstrated by its ability to identify TDI-guinea pig serum albumin (GSA) adducts with as few as 0.7 mol of TDI/mol of protein. The antiserum did not react with GSA nor with a GSA adducted with another aromatic diisocyanate, diphenylmethane 4,4’-diisocyanate. TDI-protein adducts in the BAL fluid were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by immunoblotting with use of the rabbit anti TDI-KLH antiserum. A t least 5 protein bands were recognized by the antiserum. Electrophoretic mobilities indicated molecular sizes equivalent to 10.5,38,45,66, and 148 kDa. Employing a murine anti-GSA antibody in immunoaffinity chromatography, one of the proteins in the 66-kDa band was identified as serum albumin. Attempts to purify the TDI-albumin adduct using a Cibacron-Sepharose column were unsuccessful. Studies with a model T D b . r GSA conjugate (which contained an average of 0.7 mol of TDI/mol of GSA) indicated that the TDI-albumin was not retained by the triazine dye column unless the adducted protein was first reduced by incubation with mercaptoethanol. These studies indicate that immunologic procedures can be used to identify covalently bound TDI-proteins in bronchoalveolar washings following in vivo exposure of guinea pigs to atmospheres of TDI.

Introduction Toluene diisocyanate (TDI)’is a highly reactive chemical used in the manufacture of polyurethane. It is a recognized chemical allergen and is the leading cause of occupational asthma in the Western world (1). Although cases of respiratory sensitization have been recognized for more than 30 years, the mechanism of TDI sensitivity is still uncertain since evidence of immunologic effects has not been demonstrated in the majority of cases. In 1980, we undertook development of an animal model for TDI hypersensitivity (2) to investigate the pulmonary and immunologic consequences of exposure to the chemical. We were able to develop an exposure paradigm which resulted in pulmonary hypersensitivity to the chemical. Sensitized animals demonstrated respiratory responsiveness and had TDI-specific hypersensitivity antibodies in their serum (3). Whereas hypersensitivity antibodies were detected in all animals sensitized to TDI, analogous

* Author to whom correspondence should be addressed.

+ Department of

Environmental and Occupational Health. Department of Pharmaceutical Sciences. * Abstract published in Advance ACS Abstracts, November 15,1993. Abbreviations: APC, antigen-presentingcell;BAL,bronchoalveolar lavage; BLOTTO,buffer composed of 1% dry nonfat milk, 1% Tween 20, and 10mMTris-saline (pH 7.4);BSA, bovine serumalbumin;ELISA, enzyme-linked immunosorbentassay; GSA, guinea pig serum albumin; ip, intraperitoneal;KLH, keyhole limpet hemocyanin; MDI, diphenylmethane4,4’-diisocyanate; ME, mercaptoethanol;PAGE,polyacrylamide gel electrophoresis;SDS, sodium dodecyl sulfate;TDI,toluene diisocyanate.

antibodies have not been routinely identified in patients diagnosed with TDI sensitivity (4). We sought to gain a better understanding of the molecular reactions initiating sensitization. We had previously employed [l4C1TDIto investigate the uptake of inhaled TDI by sensitized and naive guinea pigs (5). Such studies indicated labeled proteins in the blood, with their amount dependent upon both the concentration and length of the TDI exposure ( 5 , 6 ) . In the current study, we employed the guinea pig animal model and an exposure paradigm which has consistently produced sensitization (7), to trace the molecular reactions of inhaled TDI in the respiratory tract. We elected to use immunostaining of TDI, rather than employing radiolabeled isocyanate, to accomplish our goals.

Experimental Procedures TDI was obtained in the form of TDm (Miles Chemical Corp., Pittsburgh, PA), a 4:l molar mixture of 2,4- and 2,6-isomers. Caution: TDI is a skin and mucus membrane irritant and should be handled inachemical vapor hood withappropriateprotective equipment. Its hydrolysisproducts are aromatic amines, which are potential human carcinogens. Exposure of Guinea Pigs to TDI. Male Hartley guinea pigs (HRP, Inc., Denver, PA; n = 4) weighing 450-600 g were placed in a 40-L glass chamber and exposed to TDI atmospheres for 3 h as previously described (3). Briefly, TDI was generated by bubbling metered amounts of dried air through an impinger which contained TDI. The TDI concentration in the chamber was

0893-228x/93/2706-0906$04.00/00 1993 American Chemical Society

TDI Protein Adducts in BAL Fluid adjusted by varying the amount of air entering the impinger. A vacuum pump was used to draw air through the exposure chamber at a rate of 40 L/min. Concentrations of TDI in the exposure chamber were determined by sampling and analysis by HPLC as previously described (6). Control animals (n = 41, placed in an identical chamber, were treated with saline rather than TDI. Lung Lavage. Immediately following the exposure, animals were anesthetized by intraperitoneal (ip) injection with 120 mg/ kg sodium nembutal (Abbot Laboratories, Chicago, IL). The trachea was isolated and instilled with 5 mL of Hank's balanced salt solution. Cells were pelleted from the bronchoalveolarlavage (BAL)fluid by centrifugation at 600g (4 "C). The BAL fluid was stored at -70 OC. Synthesis of TDI-Protein Conjugates. TDI (18 p L ) was added to 20 mL of a rapidly stirred 0.5% solution of keyhole limpet hemocyanin (KLH) in 0.05 M sodium borate-KC1 buffer (pH 9.4) at 37 OC. After 4 h, the reaction was stopped by addition of excess HOCHzCH2NHz. The conjugatewas isolated by dialysis, and then lyophilization. A series of TDI-guinea pig serum albumin (GSA) conjugates was prepared to achieve average hapten:protein molar ratios ranging from 0.7 to 26. Conjugates were synthesized by addition of increasing amounts of TDI 11%in (CH&COI to 0.5% GSA in 0.05 M sodium phosphate buffer (pH 7.4) at 37 OC for 4 h. Products were isolated as described for TDI-KLH. The number of moles of TDI bound per mole of GSA was calculated from UV spectra using tu2 = 15 680 (8). Antibody Production. Two New Zealand White female rabbits (Green Meadows Rabbitry, Inc., Murrysville, PA), each approximately 2 kg, were injected intradermally with 1.8 mg of the TDI-KLH conjugate in saline-Freund's adjuvant (2:3) emulsion distributed among 15 sites along both sides of the back. The rabbits were boosted subcutaneously 21 days later with the same dose of TDI-KLH in incomplete adjuvant distributed among 5 sites along the nape of the neck. Blood was drawn 7-10 days following the boost. Female Swiss Webster mice (Hilltop Lab Animals, Inc., Scottdale, PA; n = 15) weighing 18-20 g were injected ip with 1 mg of GSA in 0.25 mL of saline-complete Freund's adjuvant emulsion. Mice received booster injections which consisted of 1 mg of GSA in an emulsion of saline-incomplete Freund's adjuvant on days 14 and 35. Two weeks after the last injection, the ascites was drawn by tapping the peritoneal cavity with an 18-gaugeneedle. Approximately 10 mL of ascites was obtained per animal. The fluid was evaluated for the presence of GSAspecific antibodies by enzyme-linked immunosorbent assay (ELISA). ELISA. Microtiter plates (96-well PVC) were coated by addition of 50 pL of 5 pg/mL antigen in 0.1 M sodium carbonate buffer (pH 9.6) and incubation at 37 "C for 3 h. Increasing 2-fold dilutions of test antiserum or ascites fluid were added in 0.02 M PBS buffer (pH 7.4) containing 0.05 % Tween 20 (v/v) and 0.5% bovine serum albumin (BSA). Plates were incubated at 37 OC for 1h. Bound antibody was detected using either peroxidaselabeled rabbit anti-mouse IgG (1:1000 dilution) or peroxidasegoat anti-rabbit IgG (1:1000), both obtained from ICN Immunologicals. After 1 h 2,2'-azinobis(3-ethylbenzothiazoline-6sulfonic acid) and HzO2 (0.006%) were added. The enzymatic reaction was stopped after 3 min by addition of 0.5% sodium dodecyl sulfate (SDS) (50pL/well). Absorbancewas determined at 410 nm. ELISA Inhibition Assay. Increasing amounts of inhibitor antigens, dissolved in the PBS-Tween-BSA buffer, were added to the wells of the antigen-coated microtiter plates prior to the addition of antiserum. Final dilutions of antiserum were 1:4000 for TDI-KLH-coated plates and 1:1256 for plates coated with TDIzgGSA (GSA adducted with an average of 26 mol of TDI/ mol of GSA). Thereafter, the procedure described above for ELISA was followed. All assays were performed in triplicate.

Chem. Res. Toxicol., Vol. 6, No. 6,1993 907 Percent inhibition was calculated as Abs,,, with inhibitor % inhibition = 100 1- Abs,,, without inhibitor SDS-PAGE Electrophoresis. Twenty-microliter aliquots of protein samples (approximately 4 pg of protein) in 0.16 M Tris buffer (pH 8.8) containing 10% sucrose, 1% SDS, 0.001 % N d a , and 1% mercaptoethanol (ME) were heated for 3 min, placed in an ice bath, and then added to wells of a gradient gel (T = 4-22.5 % ). Electrophoresis was performed using 60-mA constant current. Proteins were stained with 0.1 % Coomassie Blue R-250 in a solution containing 25 % EtOH and 8 % HOAc. Immunoblot. Proteins were transferred from the SDSacrylamide gel to a nitrocellulose membrane (0.45 pm, Bio-Rad, Hercules, CA) in 25 mM Tris-glycine buffer (pH 8.3) at 4 OC using 50 mA per gel for 15-18 h. The nitrocellulose sheets were blocked by immersion in a solution containing 1% dry nonfat milk, 1% Tween 20, and 10 mM Tris-saline (pH 7.4) (BLOTTO) for 1.5 hatambienttemperature. Thefirstantibody(1:120rabbit anti-TDI-KLH or 1500 mouse anti-GSA) diluted in BLOTTO buffer was added and reaction was continued for 1.5h a t ambient temperature. The membrane was rinsed 5 times with Tris-saline buffer (pH 7.4). A peroxidase-labeled second antibody was then added at 1500 dilution, and incubation was continued for 1.5 h. Strips were developed by immersion in 0.05% aqueous 4-chloro1-naphthol containing 16.5% CHsOH and 0.04% H202. Dot Blot. For detection of TDI-protein adducts, 2 pLof 0.2% antigen solution was spotted onto a nitrocellulose sheet. Spots were visualized following the procedure described for the immunoblot. Affinity Chromatography. Columns of Affi-Gel Blue (Cibacron Blue F3GA on agarose, 50-100 mesh, 150-300 pm, BioRad, Richmond, CA) and Reactive Blue-2 Sepharose CL-6B (CibacronBlue on Sepharose CL-6B, Sigma, St. Louis, MO) were used to remove albumin from the lung lavage fluid. Columns were typically 0.7 cm i.d. X 10 cm length operated at flow rates of 1.5-1.7 mL/min for the Affi-Gel Blue column, and 0.3-0.5 mL/min for the ReactiveBlue-2 column. BAL or in vitro prepared TDI-GSA adducts were applied to the Reactive Blue columns in 0.05 M Tris-O.05 M NaCl (pH 8.0) and eluted with 1.5 M KC1. For Affi-Gel Blue chromatography, the starting buffer was 0.02 M sodium phosphate (pH 7.4). Bound materials were eluted with the same buffer containing 1.4 M NaC1. Eluted fractions were concentrated and desalted using a Concentricon 10 concentrator filter (cutoff 10 kDa, Amicon, Beverly, MA) to a final volume of 0.5 mL. Immunoaffinity Chromatography. GSA was isolated from BAL fluid utilizing an immunoaffinity chromatography column prepared with Affinica protein A agarose (Affinica antibody orientation kit, Schleicher & Schuell, Keene, NH). Two mL of mouse anti-GSA ascites fluid, having an antibody titer of 1:16 400 by ELISA, was diluted with 4 mL of "binding buffer" (pH 8.9) and incubated with 1mL of the protein A gel. The antibody was cross-linked to the gel with dimethyl suberimidate. The column had a flow rate 0.5 mL/min. BAL was added to the column in 0.05 M sodium phosphate-O.15 M NaCl (pH 7.2) and the bound GSA eluted with 0.2 M glycine (pH 2.3).

Results C h a r a c t e r i z a t i o n of the A n t i - T D I A n t i s e r u m . The sensitivity and specificity of the rabbit anti-TDI antiserum was evaluated. T h e antiserum had a titer of 1:20 480 with TDI-KLH, and 1:5120 with TDIzs-GSA, as determined by ELISA. Inhibition studies were performed t o characterize t h e epitopes recognized b y the antibodies. TDI-KLH, and to a more limited degree, K L H , were able t o inhibit the reaction of antibodies with t h e homologous TDI-KLH

908 Chem. Res. Toxicol., Vol. 6, No.6, 1993

:

0

Jin et al.

60 8

20

0

40!0

.o 1

.1

1

100

10

ug Inhibitor 100

b 80

20

0

p ' i

cS A 'DI-KLH

TDI,GSA

20

400

-

TD17GSA, 40

200

u mol TDI added Figure 2. Synthesis of TDI-GSA conjugates with increasing haptenic density. Increasing amounts of TDI (in acetone) were added to 6-57 mL of 0.5% GSA in 0.05 M sodium phosphate buffer (pH 7.4) and allowed to react at 37 O C for 4 h. The amount adducted to GSA was determined using ~ , 2= 15 680. TDI,GSA

60

40

:LH

TDlo.7GSA

I

.01

.1

1

10

100

ug Inhibitor

Figure 1. ELISA inhibition assays using rabbit antiserum to TDI-KLH. Increasing amounts of inhibitors were added to the wells of plates just prior to addition of antiserum. Points indicate the mean of triplicate values. (a) Microtiter plates were coated with 5 pg/mL TDI-KLH; antiserum dilution was 1:4000. From the regression lines, the amount of inhibitor required to achieve 50% inhibition of reaction was 1.2 pg of TDI-KLH and 7.5 mg of KLH. None of the following proteins gave significant inhibition: TDIB-GSA, MDI21-GSA, phenyl isocyanate-GSA, or GSA. (b) Microtiter plates were coated with 5 pglmL TDIwGSA; antiserum dilution was 1:1250. TDI-KLH and TDIzgGSA had equal potency as inhibitors, with 0.28 pg yielding 50% inhibition. No inhibition was obtained with up to 50 pg of GSA, TDI2-GSA, MDI-GSA, or phenyl isocyanate-GSA. Each of the latter four test inhibitors is indicated by X.

antigen (Figure la). No inhibition was obtained using TDI2&SA, phenyl isocyanates&SA, MDI~I-GSA,or GSA. Recognition of TDI epitopes was apparent from the inhibition studiesperformed using microtiter plates coated with TDIzgGSA (Figure lb). In this case, TDI-KLH and TDI2e-GSA were equally potent as inhibitors of the antibody reaction, whereas phenyl isocyanate-GSA,T D I r GSA, MDI~I-GSA, and GSA gave no reactivity when included at concentrations as high as 50 pg of inhibitor per well. To assess the affinity of the antiserum to proteins containing few covalently bound TDI molecules, a series of TDI-GSA conjugateswas synthesized,with TDI content ranging from an average of 0.7 to 26 mol of TDI/mol of GSA. Conditionsemployed for synthesisof the conjugates

h,ADI,,GSA

Rb Anti TDI- KLH 1: 200 Figure 3. Dot blot of reactionsof the anti-TDI-KLH antiserum with adducted and native proteins. Two-microliter aliquots of each protein (4 pg) were spotted onto a nitrocellulose sheet. AntiTDI-KLH antiserum (1:200) was added, followed by goat antirabbit IgGperoxidase (1:5OO) and substrate. The antiserum reacted with each of the antigens which contained either TDI or KLH. No reactions were noted with GSA or MDI.

are shown in Figure 2. A linear relationship was noted between the amount of TDI added and that covalently bound to GSA using low molar ratios of TDI to GSA. Using higher TDI concentrations, such as a 50:l molar ratio of TDI to GSA, less than 68 % of the added isocyanatereacted to form stable adducts with the protein. The reaction of the rabbit antiserum with TDI-protein adducts was assessed using the dot blot assay. As shown in Figure 3, a strong reaction was noted with TDI-KLH and with KLH (1:200 dilution of antiserum). Reaction was also apparent with each of the TDI-GSA conjugates, whereas no response was obtained with GSA. These results indicate that the antiserum has affinity for protein which is covalently modified with as few as 1mol of TDI/mol of protein. The antiserum was tested for possible cross-reactivity with another aromatic diisocyanate. No cross-reactivity of the rabbit antiserum was observed with GSA extensively modified with diphenylmethane4,4'-diisocyanate (MDI21GSA). Analysis of BAL Fluid. Bronchoalveolar lavage was performed immediatelyfollowing a 3-h exposureof guinea pigs to 2 ppm TDI. The protein constituents in the BAL

Chem. Res. Toxicol., Vol. 6, No.6, 1993 909

TDI Protein Adducts in BAL Fluid

200

-

+SH

-S H

N1 N2 N3 E l E2 E3

E3 E2 E l N3 N2 N1

BAL

r

1 2 1''

*

f

148 50

-

10

-

Figure 4. SDS-PAGE of BAL fluid from 3 guinea pigs exposed to TDI vapor (El-3), and 3 control animals (Nl-3). Lanes on the left contained samples which were denatured in the presence of those on the right were denatured in ita mercaptoethanol (+SH); absence (-SH). Center lanes are molecular mass standards (2006.5 kDa). Numerous proteins were detected in the BAL fluid. Arrow and unnumbered lines indicate positions of the bands. No difference was apparent between animals exposed to TDI and control animals.

fluid were evaluated by SDS-PAGE under reducing (+SH) conditions (Figure 4). Numerous proteins were detected in BAL fluid from both TDI-exposed (El-3) and nonexposed (Nl-3) guinea pigs. Components had apparent molecular masses which ranged from 10to 200 kDa. There was no obvious difference between the bands from the exposed and those from the nonexposed animals. The greatest intensity of bands was located in the 60-70-kDa region. To assist in identification of the bands, samples were analyzed by SDS-PAGE following heating both with and without addition of ME. As shown on the right portion of the gel in Figure 4 which contained proteins denatured under nonreducing conditions (-SH), fewer bands were detected in the 30-50-kDa region. In addition, the major bands (60-70 kDa) migrated farther into the gel. TDI-protein adducts were identified by immunoblot assay employing the rabbit anti-TDI-KLH antiserum. As can be seen in Figure 5,BAL sample 1had TDI-containing constituents migrating with molecular masses corresponding to 10.3,38,45,66, and 148 kDa. Sample 2 had TDI adducts at 66 and 148kDa. In these analyses, samples were heated in the presence of ME prior to electrophoresis. After denaturing the samplesby heatingwithout ME, none of the resulting electrophoretic bands reacted with the anti-TDI antiserum. BAL from the control animals did not react with the anti-TDI antiserum when analysis was performed with or without addition of ME. Identification of the TDI-Adducted Proteins in BAL. The heavy concentration of TDI-adducted bands with molecular mass of 65-70 kDa in the BAL prompted investigation of whether serum albumin (66 kDa) was one of the modified proteins. Studies were undertaken to test the ability of Sepharose-Reactive Blue affinity medium to bind guinea pig albumin,as well as GSA which contained a small number of covalently bound TDI moieties. The results shown in Figure 6 indicate that the column was able to bind GSA (left portion: lane 1contained material not retained by the column; lane 2 contained protein eluted from column). Very little of the TDI0.rGSA conjugate,

45

-

38 -

10.3Figure 5. Immunoblot of BAL samples from two guinea pigs (1 and 2) exposed to TDI. Samples were heat denatured in the presence of ME. TDI-protein adducts were identified by reaction with rabbit anti-TDI-KLH antiserum (1:125). Samples from animal 1had bands with molecular masses of 10.3,38,45,66, and 148 kDa. Bands from animal 2 had molecular masses of 66 and 148 kDa. No adduct bands were observed in BAL from control animals (not shown).

LGSAJTDI,.,

-GSA

Figure 6. Reactive Blue 2-Sepharose affmity media bound GSA but not TDb.&SA. One milligram each of GSA and T D b . r GSA was applied to the column. Lanes 1and 3 indicate GSA and TDb-TGSA,respectively, which were not retained by the column. Lanes 2 and 4 contain proteins eluted from the column with 1.5 M KCl. All lanes received 4 pg of protein.

however, was bound by the column (right lanes). Lane 3 contained adducted protein which was not retained, and lane 4 contained an apparently small amount of conjugate eluted from the column. During these studies it was noted that the rabbit antiserum reacted with these TDI-GSA conjugates only after reduction with ME. Immunoblot analysis after reducing and nonreducing conditions is shown in Figure 7. Bands were detected when reduced TDIo.&SA (lane 2) and T D I A S A (lane 3) were analyzed. No bands were detected when these conjugates were analyzed after heating without ME (lanes 1 and 4).

910 Chem. Res. ToxicoZ., VoZ. 6, No. 6, 1993

If1

2

+ 148

Jin et at.

SM F1

-

66 -

1 2 3 4 Figure 7. Immunoblot of TDb.&SA and TDI&SA after heat denaturation with ME (+SH) and without ME (-SH). Rabbit anti-TDI-KLH antiserum (1:125) reacted only with the +SH samples as evidenced by the 66-kDa bands and minor bands a t 148 kDa.

TDI,- GSA zSY+.SH_

Figure 9. SDS-PAGE of nonimmune rabbit serum. Rabbit serum (100pL) was applied to the anti-GSA affinity column, and the nonretained material was collected. Samples were denatured in the presence of ME and then placed in the lanes. Lane SM indicates starting material; lane F1 is the nonretained fraction. The dark band in each lane is albumin. Patterns were identical except for partial removal of a 56-kDa band from lane F1.

BAL

1 2 3 m

148 -

- 148 120 -

66-1

55 -

-66

45 38 -

Figure 8. SDS-PAGE of TDIfiSA. Heat denaturation of the conjugate in the presence of mercaptoethanol (+SH) yielded a major band a t 66 kDa (minor band a t 148 kDa). Denaturation without mercaptoethanol (-SH) yielded a major band with molecular mass equivalent to 55 kDa and a minor band a t 120 kDa.

TDI-GSA conjugates which were heat denatured in the presence of mercaptoethanol were found to have reduced electrophoretic migration when compared with those denatured without ME. As shown in Figure 8,TDI2-GSA gave a band at 55 kDa (with a minor band at 120 kDa) when heat denatured without ME, whereas inclusion of ME yielded a major band at 66 kDa with a minor band at 148 kDa. Anti-GSA Immunoaffinity Column. An anti-GSA immunoaffinity column was next used to determine if GSA was one of the proteins covalently modified by TDI in the BAL fluid. To prepare the column, murine ascites fluid, which contained anti-GSA, was linked to immobilized protein A. We tested the column for possible nonspecific protein binding by applying 100 pL of nonimmune rabbit serum. The nonretained proteins were then compared with the starting material using SDS-PAGE. The results in Figure 9 illustrate that the protein pattern of the nonimmune rabbit serum before chromatography (lane SM) was nearly identical with that obtained after passage

10.3Figure 10. Immunoblot of BAL fluid from guinea pig 1after inhalation of 2 ppm TDI for 3 h, and chromatography using an anti-GSA immunoaffinity column. TDI-protein adducts were detected using rabbit anti-TDI-KLH antiserum (1:125). Lane marked "BAL" contains material prior to column chromatography. Lane 1indicates protein not retained by the column. Lanes 2 and 3 show TDI adducts sequentially eluted with 0.2 M glycine, pH 2.3. The weak band a t 66 kDa in lane 1and strong band a t this location in lanes 2 and 3 indicate most of the 66-kDa TDIprotein adduct reacted with the anti-GSA column.

through the anti-GSA column (lane Fl). Partial removal of a 56-kDa band was noted. BAL fluid was applied to the column. Eluates were analyzed by immunoblot using the anti-TDI-KLH antiserum. Results are shown in Figure 10. BAL fluid prior to chromatography was found to contain numerous TDIadducted proteins (BAL lane). Lane 1illustrates material not retained by the anti-GSA column. Comparison of this lane with that containing the BAL demonstrates column retention only of protein in the 66-kDa region. The eluate also demonstrated a major band with approximate molecular mass of 55 kDa. Lanes 2 and 3 illustrate bound conjugates which were eluted with 0.2 M glycine (pH 2.3).

TDI Protein Adducts in BAL Fluid

These proteins were confirmed to be GSA conjugates of TDI by separate staining of bands with anti-GSA antibody (not shown).

Discussion The process whereby inhalation of a reactive chemical leads to development of pulmonary sensitization is not clearly understood. The first stage in the process is believed to involve a strong association (either covalent or ionic) of the chemical with a biomacromolecule or a cell surface constituent. If the adducted material is not cellular, it is believed to next associate with an antigen-presenting cell (APC) expressing major histocompatibility antigens. Recognition of the resulting complex by appropriate T cells initiates induction of a specific immunologicresponse. The macromolecules with which TDI reacts in the airways are not known. Initial studies utilized [14C]TDI to investigate the uptake and fate of inhaled TDI by sensitized and naive guinea pigs (5). Sensitized animals were challenged by inhalation of 0.1 ppm [l4C1TDIfor 3 h. Examination of blood drawn immediately following the challenge indicated that the label was associated with a plasma component, and in particular, only with a 70kDa protein in plasma (9). Subsequent studies utilized naive guinea pigs and found that airway tissue was the most heavily labeled tissue following inhalation of [l4C1TDI (6). The radiolabel was detected in regions down to the level of the terminal bronchiole, but not in the alveolar region. The radioactivity was localized mainly to the subepithelial region of the airways (10). Lung homogenates were examined by SDS-PAGE, under reducing conditions, and revealed three labeled proteins with molecular masses of >200,70, and