Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Hepatotoxic and pneumotoxic consequences are elicited by exposure to DCE, a volatile chemical used in the plastics manufacturing industry and a widespread water contaminant (Forkert et al., 1986; Forkert and Moussa, 1991). The cytotoxic effects are linked to metabolic activation of DCE to reactive intermediates that bind to tissue macromolecules (Forkert et al., 1986). Findings from in vitro studies have demonstrated that DCE metabolism in liver and lung microsomes is mediated by cytochrome P450, inasmuch as covalent binding of DCE to microsomal proteins was NADPH-dependent and was significantly inhibited in heat-inactivated microsomes or incubation in the absence of oxygen (Okine and Gram, 1986; Forkert et al., 1987). Previous studies have implicated the cytochrome P450 isozyme CYP2E1 in the metabolism of DCE (Siegers et al., 1979; Hewitt and Plaa, 1983; Kainz et al., 1993). More recent studies have shown that incubation of lung and liver microsomes with DCE in the presence of an NADPH-generating system caused significant reduction in p-nitrophenol hydroxylase activity, a catalytic marker selective for the CYP2E1 enzyme (Lee and Forkert, 1994; 1995). Inhibition with an anti-CYP2E1 antibody abrogated the DCE-induced reduction in lung and liver p-nitrophenol hydroxylation, a finding that correlated with levels of the immunodetectable CYP2E1 protein. Taken together, these data supported the premise that DCE metabolism is mediated by CYP2E1 in both liver and lung microsomes.

Primary metabolites formed from DCE in rat liver microsomal incubations have been identified as DCE-epoxide, 2,2-dichloroacetaldehyde and 2-chloroacetyl chloride (Liebler et al., 1985, 1988; Liebler and Guengerich, 1983; Costa and Ivanetich, 1984). All are electrophilic metabolites that undergo secondary reactions including further oxidation, conjugation with GSH and/or hydrolysis. We have recently shown that the major products formed in murine liver and lung microsomal incubations were the GSH conjugates GAG and GTA, which are both believed to be derived from the DCE-epoxide (Dowsley et al., 1995, 1996; fig. 1). It is also believed that reaction of 2 molecules of GSH with the DCE-epoxide leads to the formation of GAG and that GTA is derived, in part, from subsequent hydrolysis of GAG. S-(2,2-dichloro-1-hydroxy) ethyl glutathione, the GSH conjugate produced from reaction of GSH with 2,2-dichloroacetaldehyde (Liebler et al., 1985), was not detected in our experiments (Dowsley et al., 1995). The acetal as well as chloroacetic acid and S-(2-chloroacetyl)-glutathione, the hydrolysis and GSH-conjugated products of 2-chloroacetyl chloride, respectively, were detected, but the levels were low compared with those identified for GAG and GTA, conjugates derived from the DCE-epoxide. Formation of the DCE-epoxide GSH conjugates was enhanced 3-fold in mice treated chronically with acetone (Dowsley et al., 1995), a treatment regimen that induced a 3.7-fold increase in p-nitrophenol hydroxylation, compared with control levels (Forkert et al., 1994); immunoinhibition of the CYP2E1 enzyme resulted in significant reduction in the levels of GSH conjugates formed. These results indicated that the DCE-epoxide is the major metabolite generated from CYP2E1-dependent metabolism of DCE and that GTA may be an important product of this metabolic pathway.

View larger version (7K): [in this window] [in a new window]   Fig. 1.   Proposed pathway of DCE metabolism. Cytochrome P450-dependent oxidation of DCE leads to production of DCE-epoxide, which conjugates with GSH to form GAG and GTA. Subsequent hydrolysis of GAG results in the formation of GTA.

The DCE-epoxide appears to be an efficient scavenger of GSH in forming GSH conjugates, resulting in the depletion of GSH reported previously in liver and lung (Forkert and Moussa, 1991, 1993; Moussa and Forkert, 1992). The depletion correlated with increased covalent binding of DCE to cellular proteins and with severities of cellular damage in both liver and lung, which suggests that GSH represents an initial and important defense against DCE-mediated cytotoxicity. These findings further suggested that the cellular location of GSH conjugates reflects the sites of formation of the DCE-epoxide. We were interested in identifying the cellular sites of epoxide formation and in determining whether the locations coincided with the cellular targets of DCE-induced cytotoxicity. As an initial step to this end, we have undertaken herein to develop an immunological reagent for detection of GTA, a major conjugate derived from reaction of the DCE-epoxide with GSH. We have developed an antiserum against chemically synthesized GTA and characterized the specificity of the antibodies for recognition of the hapten and related compounds.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals and reagents. The following drugs and chemicals were used in this study: Na₂CO₃, NaHCO₃, Tween 20, trifluoroacetic acid, sodium borohyhride (BDH Inc., Toronto, Canada), sodium azide, sulfuric acid, hydrogen peroxide (30%, v/v), sodium acetate (Fisher Scientific Co., Toronto, Ontario, Canada), BSA, human serum albumin, OVB, GLUT (70%), O-phenylenediamine (Sigma Chemical Co., St. Louis, MO), Tris, glycine, EDC (ICN Biomedicals, Montreal, Quebec, Canada), citric acid (J. T. Baker Chemical Co., Phillipsburg, NJ), goat anti-rabbit IgG conjugated to horseradish peroxidase or alkaline phosphatase (Life Technologies, Burlington, Ontario, Canada), GSH (Aldrich Chemical Co., Montreal, Quebec), nitrocellulose membrane (0.45 µm), 5-bromo-4-chloro-3-indolylphosphate p-toluidine salt, p-nitroblue tetrazolium chloride, protein assay dye reagent concentrate (Bio-Rad Laboratories, Mississauga, Ontario, Canada), [³H]-GSH (specific activity 43.8 Ci/mmol) (Dupont Canada, NEN Ltd., Mississauga, Ontario, Canada), Spectrapor dialysis tubing (molecular weight cut-off 3500) (Spectrummedical Industries Inc., Los Angeles, CA), polystyrene 96-well flat-bottomed ELISA plates (Corning Costar Corp., Oneonta, NY). An IgG purification kit [ImmunoPure (A/G)] was purchased from Pierce Chemical Co. (Rockford, IL). All other chemicals were of reagent grade and were obtained from standard commercial suppliers.

Synthesis and purification of GTA. GTA was synthesized by combining 307 mg of GSH with 94 mg of chloroacetic acid in 10 ml of 100 mM potassium phosphate buffer, pH 7.4, to yield a final equimolar concentration of 100 mM. The solution was heated at 50°C for 2 h and was subsequently left at room temperature for 16 h before storing at 20°C. Samples (25 µl) were analyzed by HPLC with a reverse-phase C-18 column (5 µm, 4.6 × 250 mm, Phenomenex, Torrance, CA). The mobile phase was 0.2% phosphoric acid in H₂O, pH 3.15, and the flow rate was 1 ml per min. The presence of the peak corresponding to GTA was identified by an elution time of 7.6 min, and this was similar to the elution time determined for GTA in previous studies (Dowsley et al., 1995). The identity of GTA has also been confirmed and characterized in previous studies (Liebler et al., 1985; Dowsley et al., 1995). The solution was then concentrated in vacuo and subjected to semipreparative HPLC analysis by using an Ultrasphere ODS column (5 µm, 10 × 250 mm, Beckman, Palo Alto, CA) with 0.1% trifluoroacetic acid in H₂O as the mobile phase and a flow rate of 5.0 ml per min. The eluent corresponding to the GTA peak was identified and collected. Approximately 40 HPLC injections were performed, and the GTA peak eluent was collected, pooled, concentrated in vacuo and stored at 20°C. The sample was analyzed by ¹H-NMR and electrospray mass spectroscopy.

Synthesis of immunogen. GTA was conjugated to BSA by using a single-step coupling procedure employing GLUT as a cross-linking reagent. The coupling reaction was performed at room temperature by applying procedures detailed by Harlow and Lane (1988). BSA (55 mg) was added to a solution containing GTA (2 mg) in 1 ml of PBS. An equal volume of 0.2% GLUT (20 µmol) in PBS was then slowly added to the hapten-protein solution over 0.5 h with constant stirring. In addition, a tracer quantity of [³H]-GSH (10⁶ dpm) was included in the reaction mixture to evaluate coupling efficiency. [³H]-GSH was selected as a monitor for coupling efficiency because of its similarity in chemical characteristics to the hapten, including the presence of identical functional groups for the coupling chemistry (primary amine), and the availability of the radioactive derivative. GLUT is highly reactive to primary amines, and coupling via sulfhydryls is expected to be insignificant. The reaction was allowed to proceed for 1 h at room temperature, and the free aldehyde coupling groups and Schiff base intermediates were reduced by the addition of NaBH₄ (10 mg/ml). At the conclusion of the reaction, the mixture was transferred to dialysis tubing, and noncovalently coupled hapten was removed by dialyzing extensively against PBS and finally water. A small amount of the final product was sampled for determination of coupling efficiency. Under these conditions, a coupling efficiency of 83% was obtained, which corresponded to a hapten coupling ratio of 0.15 µmol of hapten per mg of BSA (10 mol hapten per mol BSA).

Immunization procedures. All procedures for immunization were performed using guidelines published by the Canadian Council for Animal Care and approved by the Bioquant Laboratories Limited Review Committee. Female New Zealand white rabbits (2-3 kg) were immunized s.c. at multiple sites using a "biweekly" injection/bleeding protocol. Initially, 300 µg of the immunogen (GTA-GLUT-BSA) in 1 ml of Freund's complete adjuvant was injected s.c. at multiple sites, followed by boosts of 150 µg of GTA-GLUT-BSA in Freund's incomplete adjuvant at monthly intervals. Bleeds were performed at two weeks following each boost by collection from the ear vein. For all rabbits tested, high titre antiserum was obtained by the second bleed and remained high for the duration of the collection (total of 3 months).

Affinity purification of antisera. A protein A/G column was equilibrated with 10 ml of binding buffer provided in the antibody purification kit. A sample (1 ml) of the antiserum was diluted with an equal volume of binding buffer and loaded onto the column. The column was next washed with 20 ml of binding buffer. The IgG was then eluted with 10 ml of elution buffer in 1-ml fractions. The amounts of IgG eluted into each tube were estimated by determining absorbance at 280 nm. The fractions containing the highest amounts of antibody protein were pooled and desalted. The desalting columns were equilibrated with 10 ml of PBS; each column was used for a sample volume of 1 ml. A 1-ml sample of the pooled antiserum was loaded onto the desalting column and was then washed with 10 × 1 ml aliquots of nanopure water. The 1-ml aliquots were collected as fractions, and protein concentrations were determined by measurements at an absorbance of 280 nm. The tubes containing the highest amounts of protein were pooled, and the final protein concentration was determined by the Bradford protein assay. The pooled sample represented the IgG fraction and was used as the specific antibody.

Chemical syntheses. A variety of protein conjugates were prepared using methods similar to those described for preparation of the immunogen but substituting coupling reagents, carrier species or haptens as required to obtain the desired species as illustrated in fig. 2. Glycine, GSH and Tris were coupled to BSA (30 mg) with GLUT by using a concentration of 10 µmol each in a final volume of 2 ml. The GTA-GLUT-OVB complex was synthesized by adding OVB (30 mg) to the reaction mixtures to replace the BSA.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Chemical structures of the synthesized immunogen (GTA-GLUT-BSA) and related conjugates. Details of the syntheses are described in "Materials and Methods." The arrow at the GTA-GLUT-BSA structure denotes the start of the portion of the antigen unique from the GLY-GLUT-BSA conjugate (structure depicted) used for blocking nonspecific recognition. The structure of an expected derivative of EDC coupling, via the amine of GTA, is indicated by GTA-EDC-BSA. Other EDC conjugates, where coupling is achieved by conjugation via either one of the acidic functional groups of the GTA to amines in the carrier protein is also an expected reaction.

Enzyme-linked immunosorbent assay. Noncompetitive and competitive ELISA assays were used to screen the polyclonal rabbit antisera. The assays were carried out in 96-well flat-bottomed polystyrene plates. Unless specified otherwise, each well was coated by allowing 50 µl of carbonate-bicarbonate buffer, pH 9.6, containing 2.5 µg of the GTA-GLUT-BSA to incubate overnight at 4°C. Control wells contained only the carbonate-bicarbonate buffer. The plates were then washed three times with PBS and blocked with BSA (3%, w/v) in a volume of 100 µl of carbonate-bicarbonate buffer per well for 2 h at room temperature. The plates were next washed three times with PBS containing Tween 20 (0.05%). The rabbit polyclonal antibodies were diluted with 1% BSA in PBS-Tween 20 at dilutions ranging from 1:1 to 1:1000. A volume of 50 µl was dispensed into each well and incubated for 2 h at room temperature. The unbound antibody was removed by washing four times with PBS-Tween 20. The primary antibody was then reacted with goat anti-rabbit IgG conjugated with horseradish peroxidase at a dilution of 1:1000 in 1% BSA in PBS-Tween 20 containing 2.5% normal goat serum. Each well was filled with a volume of 50 µl and incubated for 2 h at room temperature. The wells were then washed three times with PBS-Tween 20 and finally with PBS. The color reaction was developed by adding 50 µl of O-phenylenediamine (34 mg/100 ml) in 0.1 M phosphate-citrate buffer containing 40 µl of 30% hydrogen peroxide. This substrate yields a soluble end product with a high extinction coefficient at 490 nm. The reaction was terminated with 50 µl of 4 N sulfuric acid. Absorbance was determined at 490 nm, using an ELISA plate reader.

Protein immunoblotting. Samples of each conjugate were diluted to a protein concentration of 5 µg in nanopure H₂O and applied to the wells of a slot-blotting apparatus onto a nitrocellulose membrane (Hoefer Slot Blot PR 648 Apparatus, Hoefer Scientific Instruments, San Francisco, CA). The wells were then washed three times with 1 ml of PBS each. The membrane was removed and treated with 3% BSA in PBS-Tween 20 for 2 h with gentle agitation, after which it was washed three times with PBS-Tween 20 for 10 min each. It was then reacted with the IgG (375 ng/ml) diluted in 1% BSA in PBS-Tween 20 and GLY-GLUT-BSA (1 mg/ml final concentration). Incubation was carried out overnight at room temperature with agitation. After this step, the membrane was washed three times with PBS-Tween 20 for 10 min each. It was then reacted with the secondary antibody (goat anti-rabbit IgG conjugated to alkaline phosphatase, 1:1000) in 1% BSA in PBS-Tween 20 and incubated for 2 h at room temperature with gentle agitation. It was washed twice for 10 min each in PBS-Tween 20 and finally with PBS. The color reaction was developed in a solution containing 5-bromo-4-chloro-3-indoylphosphate and p-nitroblue tetrazolium chloride in PBS.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Synthesis and purification of GTA. HPLC analysis of the reaction products of GSH and chloroacetic acid revealed three peaks corresponding to GSH, GTA, and oxidized GSH (GSSG) in order of increasing elution time as reported previously (Dowsley et al., 1995; 1996). GSH, GTA and GSSG eluted at 5.7, 7.6 and 16.5 min, respectively. The spectroscopic data confirmed the identity of GTA. ¹H-NMR (D₂O):  2.02 (q, 2H, J = 7.1 Hz, Glu-), 2.39 (m, 2H, Glu-), 2.90 (m, 2H, Cys-), 3.25 (s, 2H, ---CH₂COOD), 3.69 (t, H, J = 6.4 Hz, Glu-), 3.82 (s, 2H, Gly), 4.45 (dd, 1 H, J = 5.1 Hz, Cys ). Electrospray-MS (positive ion, 1% aq. CH₃COOH, pH 2.4) m/z 366 (M + 1) (11%). These data are in agreement with those reported in previous studies (Dowsley et al., 1995; Liebler et al., 1985).

Production of antibody for recognition of GTA. GTA was synthesized and coupled to BSA with GLUT to yield the GTA-GLUT-BSA conjugate, which was used as the immunogen (fig. 2). A direct or noncompetitive ELISA was used in initial studies to screen antisera from immunized rabbits. All immunized rabbits were shown to be responsive. The relative strength of each antiserum was determined by a direct ELISA, using antisera obtained at 6, 10 and 14 weeks after two, three and four boosters, respectively. Serial dilutions performed to determine a relationship between absorbance at 490 nm and dilutions of the antibody showed linearity at dilutions ranging from 1:1000 to 1:5000 (fig. 3). It was established that the antiserum obtained at 14 weeks was the most effective in binding to the coating antigen at all antibody dilutions (fig. 3). Estimates from titration of concentrations of the coating antigen showed that the optimal amount was about 2.5 µg per well. Data from these experiments were used to identify the antisera with the highest efficiency to detect the antigen.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Representative profile of antisera obtained at 6 (), 12 () and 14 () weeks after initial immunization of a rabbit with the hapten (GTA-GLUT-BSA). Binding to the coating antigen (2.5 µg/well) was evaluated by the noncompetitive ELISA, using serial dilutions of the antisera. Binding of the rabbit antibodies was detected by reaction with goat anti-rabbit IgG conjugated to horseradish peroxidase followed by O-phenylenediamine and hydrogen peroxide.

Characterization of antibody by competitive ELISA. The competitive ELISA was used to assess the efficiency with which the antibodies reacted with the target antigen. Initial screening with dilutions of the antiserum showed that the antigen competed efficiently with the antibodies for binding to the coating antigen at dilutions ranging from 1000 to 10,000 (fig. 4). However, the specificity of the reaction was enhanced when incubations were performed in the presence of GLY-GLUT-BSA as a blocking agent (fig. 4). At a dilution of 1:1000, there was about a 60% decrease in the absorbance level when reactions were carried out in the presence of GLY-GLUT-BSA. The absorbance level was not detectable when the coating antigen was absent. As a result of these findings, we performed reactions in all subsequent experiments using GLY-GLUT-BSA as a blocking agent. After these screening experiments, the antisera were subjected to affinity purification to obtain the IgG fraction.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Evaluation of the relative efficiency of GLY-GLUT-BSA to inhibit nonspecific binding of the antiserum to the antigen, as assessed by the competitive ELISA. The wells of a polystyrene plate either were uncoated () or were coated with the hapten (GTA-GLUT-BSA, 2.5 µg/well). Incubations were performed in the absence () or presence () of the competitive conjugate (GLY-GLUT-BSA), which was added to the well together with the antiserum.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Evaluation of the relative efficiencies of the antiserum (clear bar) and IgG fraction (hatched bars) to bind to the coating antigen (GTA-GLUT-BSA, 2.5 µg/well), as assessed by the competitive ELISA. Preimmune serum (filled bar) was used as a control. The competitive conjugates used were GTA-GLUT-BSA (GTA), GLY-GLUT-BSA (GLY) and GSH-GLUT-BSA (GSH); PBS was used as a control. The competitive conjugates, together with the preimmune serum (1:1000), antiserum (1:1000) or IgG (350 ng/well), were added to the wells and incubated. Detection was achieved by reaction with goat anti-rabbit IgG conjugated to horseradish peroxidase followed by O-phenylenediamine and hydrogen peroxide.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Evaluation of the relative efficiencies of a range of concentrations of the chemically synthesized conjugates to inhibit binding of the IgG (350 ng/well) to the coating antigen (GTA-GLUT-BSA, 2.5 µg/well), using the competitive ELISA. The conjugates, together with the IgG, were added to the wells. The competitive substrates used were GTA-GLUT-BSA (), GTA (), GTA-EDC-BSA (), TRIS-GLUT-BSA (), GSH-GLUT-BSA () and GTA-GLUT-OVB (). The IC50 value for GTA-GLUT-BSA was determined to be 3.75 µg. No inhibitory effects were defined for reactions performed in the presence of PBS (represented by the point on the ordinate).

Protein immunoblotting. Slot blots were prepared to determine the efficacy of the IgG fraction to recognize the hapten (fig. 7). In the first series of experiments, immunoblotting was performed to assess the ability of GLY-GLUT-BSA to block nonspecific binding of our chemically synthesized conjugates to the antiserum (fig. 7). Panel A depicts blots of conjugates or controls performed in the absence of the GLY-GLUT-BSA conjugate. Strong immunoreactivity was elicited with GTA-GLUT-BSA (lane 6); however, positive signals, albeit at considerably lower levels, were also found in reactions with GTA-EDC-BSA (lane 1), GSH-GLUT-BSA (lane 3), TRIS-GLUT-BSA (lane 4) and GLY-GLUT-BSA (lane 5). Panels B and C depict blots obtained in incubations performed with GLY-GLUT-BSA as the blocking reagent. Immunoreactivity was also specific for the GTA-GLUT-BSA conjugate (panel B, lane 6; panel C, lanes 1), and was low for the GTA-EDC-BSA conjugate (panel B, lane 1; panel C, lanes 2). In contrast, no observable protein bands were detected with the other conjugates (panel B, lanes 3-5 and 7; panel C, lanes 3-5) or with BSA (panel B, lane 2; panel C, lanes 6) or OVB (panel B, lane 8). Thus the antiserum was highly specific for GTA-GLUT-BSA, especially after blocking with GLY-GLUT-BSA.

View larger version (42K): [in this window] [in a new window]   Fig. 7.   Protein immunoblotting for chemically synthesized conjugates or controls. Samples of conjugates (5 µg in 50 µl H2O) were applied to the wells of a slot-blotting apparatus. In panel A, the membrane was reacted with IgG (375 ng/ml) diluted in PBS-Tween-20 containing 1% BSA. In panels B and C, the membranes were reacted with IgG (375 ng/ml) diluted in PBS-Tween-20 containing 1% BSA and GLY-GLUT-BSA (1 mg/ml). After rinsing, the membranes were incubated with IgG conjugated to alkaline phosphatase. The lanes in panels A and B are duplicates of the same samples. The lanes were loaded as follows: lane 1, GTA-EDC-BSA; lane 2, BSA; lane 3, GSH-GLUT-BSA; lane 4, TRIS-GLUT-BSA; lane 5, GLY-GLUT-BSA; lane 6, GTA-GLUT-BSA; lane 7, GTA-GLUT-OVB; lane 8, OVB. The lanes in the top and bottom rows of panel C are duplicates of the same samples and were loaded as follows: lane 1, GTA-GLUT-BSA; lane 2, GTA-EDC-BSA; lane 3, GSH-GLUT-BSA; lane 4, GLY-GLUT-BSA; lane 5, TRIS-GLUT-BSA; lane 6, BSA.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Reactive intermediates or electrophilic metabolites are derived from metabolic activation of xenobiotics and are linked to the pathologic lesions associated with a variety of toxicities in diverse tissues (Hinson and Roberts, 1992; Nelson and Pearson, 1990; Boyd, 1980; Mitchell and Jollow, 1975). These metabolites are the products of metabolic modification, and enzymes of the cytochrome P450 system have frequently been implicated in this process. Furthermore, the activation appears in many cases to be isozyme-selective, involving either single or multiple P450 isoforms. Accordingly, the constitutive and inducible contents of individual P450 isozymes in target tissues determine, in part, the amounts of electrophilic metabolites formed. Their formation has been strongly correlated with the manifestation of cytotoxicities mediated by numerous drugs and chemicals. The fate of these electrophiles has been of considerable interest, partly because of the hypothesized role of covalent binding of reactive metabolites to tissue macromolecules, which is believed to be a critical event leading to chemically induced cytotoxic cellular damage (Mitchell et al., 1973a; Jollow et al., 1973). The fundamental concepts of this mechanism and its role in drug-induced toxicities have been derived, in part, from studies with acetaminophen, an analgesic drug that produces hepatic centrilobular necrosis when administered in high doses to both humans and experimental animals (Mitchell et al., 1973b; Hinson, 1980; Prescott, 1983; Black, 1984). Substantial evidence has accrued to demonstrate that acetaminophen-induced hepatoxicity is mediated by N-acetyl-p-benzoquinone imine, a reactive metabolite produced by cytochrome P450-dependent metabolism (Miner and Kissinger, 1979; Dahlin et al., 1984; Harvison et al., 1988). The oxidative reaction is catalyzed by CYP2E1 and CYP1A2 (Raucy et al., 1989; Morgan et al., 1983). Detoxification is mediated by conjugation of the reactive metabolite to GSH, and after GSH depletion, covalent binding and tissue injury ensue (Mitchell et al., 1973b; Potter et al., 1974).

The proteins to which the acetaminophen metabolite N-acetyl-p-benzoquinone imine is covalently bound are believed to be cysteine sulfhydryl groups on proteins. The major protein adduct formed in livers of mice treated with toxic doses of acetaminophen has been identified by mass spectral analysis as 3-(cystein-S-yl) acetaminophen, which constituted over 70% of the adducts detected in protein hydrolysates (Hoffman et al., 1985). More recently, an immunological approach has been adopted to detect proteins that formed adducts with acetaminophen, and a polyclonal antiserum that recognized 3-(cystein-S-yl)acetaminophen has been used as a reagent to investigate mechanisms of acetaminophen toxicity as well as to identify the sites of metabolite formation (Roberts et al., 1987; Potter et al., 1989). A dose-dependent relationship was found between formation of 3-(cystein-S-yl) acetaminophen adducts and manifestation of hepatotoxicity (Pumford et al., 1989). Interestingly, acetaminophen-bound protein adducts were detected in serum after toxic doses of the drug, and their appearance coincided with the onset of hepatotoxicity. 3-(Cystein-S-yl) acetaminophen adducts were preferentially localized in the centrilobular region of the hepatic lobule, a site of acetaminophen-induced hepatotoxicity (Roberts et al., 1991) and were present at high levels in plasma membranes and mitochondria (Pumford et al., 1990). In addition, reactive metabolites of acetaminophen have formed adducts with various hepatic proteins after the administration to mice of hepatotoxic doses of the drug (Bartolone et al., 1987, 1988; Birge et al., 1990). Those that have been identified include a cytosolic selenium-binding protein (Pumford et al., 1992; Bartalone et al., 1992), a microsomal subunit of glutamine synthetase (Bulera et al., 1995), the cytosolic protein N-10-formyl tetrahydrofolate dehydrogenase (Pumford et al., 1997) and the nuclear protein laminin (Hong et al., 1994). Other studies have also used an immunochemical assay to detect the formation of mercapturic acids derived from conjugated products of GSH with the 1,2-epoxide of naphthalene, a reactive intermediate believed to be responsible for mediating bronchiolar cytotoxicity (Marco et al., 1993).

Here we have adopted an alternative approach to those used previously and have prepared an antiserum against GTA, a GSH-conjugate derived from the DCE-epoxide. This strategy was designed to address our long-term objective of identifying the cellular sites of DCE bioactivation in tissues such as the lung and liver. Conjugation to GSH occurs rapidly because of the high reactivity of the epoxide, producing GTA, and we hypothesized that the availability of specific antibodies to this GSH conjugate will render it feasible to identify the site of GTA formation. This is particularly relevant in the case of the lung, a tissue with considerable cell heterogeneity, because the nonciliated Clara cells, which are selective targets of DCE-induced cytotoxicity, are damaged early after chemical treatment and are rapidly exfoliated into the airway lumen (Forkert et al., 1986). We rationalized that formation of GTA is likely to occur soon after DCE treatment and at a time when the Clara cells are still adherent to the airway wall. On the basis of these considerations, we anticipated that development of a specific antibody reagent will permit detection of cell types that participated in the formation of GTA and, accordingly, in epoxide formation. Moreover, the identification of target cells will enable us to determine whether the sites of GTA formation coincided with those that are damaged and thereby to establish indirectly whether the DCE-epoxide mediates cytotoxicity.

The current report describes the production and binding specificity of a polyclonal antiserum raised against GTA, a product formed by reaction of GSH with the DCE-epoxide. The strategy used for production of the antiserum entailed conjugation of the target molecule GTA to a BSA carrier using GLUT as a chemical cross-linker. BSA was chosen as a carrier molecule because of the presence of abundant functional groups (primary amine) for the coupling chemistry utilized, and because of its ability to induce a characteristic immune response in rabbits. As a consequence of the relatively high solubility of BSA even after conjugation to molecules of modest solubility, it can also be used for preparation of chemically related antigens for characterization of binding specificity. Further, BSA has the significant advantage of allowing application at high concentrations, making it practical for blocking chemistry and therefore increased binding specificity. GLUT is well recognized as being highly immunogenic in addition to being a highly effective cross-linking reagent. Although this characteristic is a disadvantage in the production of antibodies to many haptens (Harlow and Lane, 1988), in this instance it was anticipated that the presence of a highly antigenic molecule adjacent to the GTA (which was not a priori predicted to be very antigenic) might facilitate GTA antibody production. In addition, the stability of the amine bonds formed during the conjugation reaction is an important advantage for the preparation of peptidyl haptens (such as GTA) that may be subject to considerable proteolytic degradation.

Several conjugates related to the GTA-GLUT-BSA complex used for antibody production were also synthesized (fig. 2). Both the unpurified serum and the affinity-purified IgG recognized all of the conjugates prepared, as a result of reaction with the carrier, the linker or the target hapten. In order to evaluate the binding to GTA, we have applied an immunoadsorption technique to block antibody reaction to the carrier molecule and the linker. As shown in figures 6 and 7, the GLY-GLUT-BSA conjugate was effective at preventing antibody recognition to the carrier and linker but allowed binding to the residual GTA component of the immunogen; the portion of the GTA-GLUT-BSA that is identical in structure to that of GLY-GLUT-BSA is depicted in figure 2 (see arrow). The addition of GLY-GLUT-BSA to the blocking steps in both the ELISA and the immunoblotting protocol allowed residual antibody recognition to that portion of the molecule depicted to the right of the arrow shown for the GTA-GLUT-BSA conjugate.

To further confirm the specificity of the antiserum, we examined the ability of both the antiserum and the affinity-purified IgG to recognize GTA coupled to carrier molecules different from those used for preparation of the antigen. Two controls [with GTA coupled to an alternative carrier protein (GTA-GLUT-OVB) or with an alternative linking reagent (GTA-EDC-BSA)], as illustrated in figure 2, were reactive to both of the antibody preparations (fig. 6). Although the reactivity to these preparations was markedly less than that to the GTA-GLUT-BSA conjugate, such a decrease in affinity may result from steric alterations in the complex with respect to BSA, particularly in regard to recognition of the GTA within the GTA-EDC-BSA conjugate. As a negative control, the conjugation of Tris to BSA with GLUT, resulting in the TRIS-GLUT-BSA conjugate shown in figure 2, did not react to the affinity-purified IgG in the presence of GLY-GLUT-BSA (figs. 5, 6 and 7); this demonstrates the GTA recognition domain is not within the GLUT (linker) or BSA (carrier) region of the GTA-GLUT-BSA conjugate. Antibody recognition of the GSH-GLUT-BSA complex was considerably less than binding to GTA-GLUT-BSA, but it was consistently detected (figs. 5 and 7). Although these data suggest that some antibody cross-reactivity to GSH may be expected, it is minimal, and appropriate antibody dilution and blocking systems could probably be developed to enhance the specificity of binding for use within, for example, an immunohistochemical protocol. Indeed, we have obtained preliminary data that suggest that this antibody is highly specific for the GTA adducts formed from GSH at the cellular levels. The inability of unconjugated GTA to inhibit antibody recognition (fig. 6) has provided further evidence regarding the functional groups participating in the antibody interaction. A significant difference between the conjugated and the free hapten is the presence of the free primary -amine of the glutamate residue in free GTA vs. a secondary amine within the conjugated species. Consistent with the premise that this position participated in antibody recognition is the observation that the EDC conjugate of GTA, yielding an amide bond, is recognized by the antibody.

In conclusion, we have developed highly specific polyclonal antibodies that recognize GTA, a GSH conjugate derived from the DCE-epoxide. The success of this effort was due in part to our adopting a strategy in which GTA is conjugated to a BSA carrier and to our using GLUT as a chemical cross-linker, thereby producing a stable conjugate resistant to proteolysis. It is anticipated that the approach we have used will serve as a model for synthesis of a variety of GSH conjugates for investigating mechanisms associated with chemically induced cytotoxicities. It is also anticipated that application of these antibodies in current and future studies will permit precise identification of the cellular sites of formation of GTA, and hence DCE-epoxide, and characterization of the targets of DCE metabolism and cytotoxicity.