Prooxidant-Initiated Lipid Peroxidation in Isolated Rat Hepatocytes

Hepatocytes: Detection of 4-Hydroxynonenal- and. Malondialdehyde-Protein Adducts. Dylan P. Hartley, David J. Kroll, and Dennis R. Petersen*. Departmen...
5 downloads 0 Views 258KB Size
Download PDF
Chem. Res. Toxicol. 1997, 10, 895-905

895

Prooxidant-Initiated Lipid Peroxidation in Isolated Rat Hepatocytes: Detection of 4-Hydroxynonenal- and Malondialdehyde-Protein Adducts Dylan P. Hartley, David J. Kroll, and Dennis R. Petersen* Department of Pharmaceutical Sciences, Molecular Toxicology and Environmental Health Sciences Program, Hepatobiliary Research Center, University of Colorado Health Sciences Center, Denver, Colorado 80262 Received November 1, 1996X

Toxicity associated with prooxidant-mediated hepatic lipid peroxidation is postulated to originate from the interaction of the aldehydic end products of lipid peroxidation with cellular constituents. The principal R,β-unsaturated aldehydic products of lipid peroxidation, 4-hydroxy2-nonenal (4-HNE) and malondialdehyde (MDA), are known to modify proteins through covalent alkylation of lysine, histidine, and cysteine amino acid residues. To detect and characterize the formation of 4-HNE- and MDA-adducted proteins during prooxidant-initiated lipid peroxidation, rabbit polyclonal antibodies were raised to 4-HNE-sulfhydryl, dinitrophenylhydrazine (DNPH)-4-HNE-sulfhydryl, and MDA-amine conjugates of keyhole limpet hemocyanin (KLH). Each antiserum displayed high antibody titers to either 4-HNE-metallothionein, DNPH-albumin, or MDA-albumin adducts when measured by ELISA. To study the formation of 4-HNE- and MDA-protein adducts during prooxidant-initiated cellular injury, isolated hepatocytes were exposed to either carbon tetrachloride or iron/ascorbate for 2 h. Indices of hepatocellular oxidative stress (i.e., cell viability and glutathione status) and lipid peroxidation (i.e., formation of 4-HNE, protein carbonyls, and MDA) were monitored continuously. Hepatocellular viability was affected moderately by carbon tetrachloride, while cellular reduced glutathione status was moderately affected by both iron/ascorbate and carbon tetrachloride. Levels of MDA and protein carbonyls increased dramatically with both prooxidants, whereas 4-HNE levels did not change significantly over the time course studied. In addition, hepatocellular proteins were immunoprecipitated with each antiserum, and aldehyde-modified immunopositive proteins were detected by immunoblotting. Prooxidant-induced increases in MDA corresponded with increases in intensity and number of MDA-adducted proteins over the time course studied. A total of 13 MDA-modified proteins (20, 25, 28, 30, 33, 38, 41, 45, 80, 82, 85, 130, and 150 kDa) were detected with the MDA-amine antiserum. Additionally, both iron/ascorbate- and carbon tetrachloride-induced formation of DNPH-derivatizable protein carbonyls corresponded quantitatively with the ability to detect specific proteins (80, 100, 130, and 150 kDa) with the DNPH-4-HNE-cysteine antiserum. Neither CCl4 nor iron/ascorbate elicited changes in 4-HNE or induced the formation of 4-HNE-modified proteins when assessed by immunoprecipitation-immunoblot analysis with the 4-HNE-sulfhydryl antiserum. In all instances detection of aldehyde-modified proteins was not associated with cell death and may be related to the function of these proteins as aldehyde-binding proteins which sequester electrophilic molecules during oxidative liver injury.

Introduction The peroxidation of polyunsaturated membrane lipids is a well-controlled, ongoing event associated with normal cell turnover and the overall process of aging (1). There is also consensus that the process of uncontrolled cellular lipid peroxidation is associated with hepatic injury as a consequence of exposure to halogenated hydrocarbons (2), excessive alcohol ingestion (3, 4), and acute or chronic iron overload (5, 6). More recently, lipid peroxidation is postulated to be an etiologic factor in such diseases as atherosclerosis (7), diabetes (8, 9), and genetic hemochromatosis (10). The peroxidative destruction of membrane lipids has predictable and direct effects on the structural integrity of cellular membranes. This degradative process also * Author to whom correspondence should be addressed. Tel: (303) 315-6159. Fax: (303) 315-6281. X Abstract published in Advance ACS Abstracts, August 1, 1997.

S0893-228x(96)00181-6 CCC: $14.00

results in the production of R,β-unsaturated aldehydes that, because of their electrophilic nature and facile reactivity, can elicit a diversity of adverse cellular effects. The most abundant byproducts of lipid peroxidation include malondialdehyde (MDA), 4-hydroxynonenal (4HNE), and hexanal (11). The production of these aldehydes in chemical-induced lipid peroxidation is supported by studies which have evaluated MDA and 4-HNE production in conjunction with indices of oxidative stress in animals or cells treated with various prooxidants (1214). Likewise, numerous investigators have attributed a variety of adverse cellular effects to these aldehydes including disruption of second-messenger systems (15), perturbations in calcium homeostasis (16), and inhibition of various enzymes (17). Studies have also appeared supporting the notion that the highly electrophilic R,β-unsaturated aldehydes, MDA and 4-HNE, affect specific cellular systems through interactions with cellular protein. These intermolecular © 1997 American Chemical Society

896 Chem. Res. Toxicol., Vol. 10, No. 8, 1997

interactions result in the formation of aldehyde-protein adducts that directly or indirectly play a central role in the cytotoxic events attributable to uncontrolled lipid peroxidation. To date, reports have been published describing the formation of aldehyde-protein adducts in cryopreserved, prooxidant-challenged hepatocytes (18) and cellular injury resulting from carbon tetrachloride administration (19), in association with hepatic iron overload (4), as a result of chronic ethanol administration (4), and during chemical-induced nephritis (20). The functional relationship between lipid peroxidation, aldehyde-protein adduct formation, and cellular injury is unresolved. Factors underlying this controversy stem from the diversity of animal and cellular models studied, the index of lipid peroxidation measured, and the diversity of procedures used to assess lipid peroxidation. In addition, the general cytotoxic potential of MDA and 4-HNE is dependent upon the rate at which these aldehydes are formed and eliminated. Therefore, in order to begin identifying potential relationships between lipid peroxidation, aldehyde-protein adduct formation, and cellular injury, comprehensive studies must be conducted to assess these parameters simultaneously in isolated viable cells. Further, it is essential that aldehydeadducted proteins be identified in order to establish the potential role of specific proteins in prooxidant-induced cellular injury. This report begins to provide an experimental framework for establishing these relationships.

Methods Reagents and Solutions. All solutions were prepared in deionized and distilled water. Unless otherwise stated, chemicals were from Sigma Chemical Co. (St. Louis, MO). Collagenase type IV used for the isolation of rat hepatocytes was purchased from Worthington Biochemical Corp. (Freehold, NJ). N-Succinimidyl S-acetylthioacetate (SATA) and keyhole limpet hemocyanin (KLH) were purchased from Pierce (Rockford, IL), and hydroxylamine was obtained from Aldrich Chemical Co. (Milwaukee, WI). Vanadyl sulfate (VOSO4) was from Fisher Scientific. Goat anti-rabbit horseradish peroxidase-conjugated IgG and streptavidin-conjugated horseradish peroxidase were obtained from Gibco-BRL (Life Technologies, Gaithersburg, MD). The Ribi Adjuvant System was purchased from Ribi Immunochem Research, Inc. (Hamilton, MT), and Freund’s complete and incomplete adjuvants were from Gibco-BRL (Life Technologies, Inc., Grand Island, NY). (E)-4-Hydroxy-2-nonenal was synthesized and liberated from the diacetal form as previously described (21). Pure 4-HNE was diluted in water and quantitated spectrophotometrically at 224 nm (water ) 13 750 M-1 cm-1). MDA was also liberated from the diacetal form (malonaldehyde tetramethyl diacetal; Eastman Kodak Co., Rochester, NY), diluted in 50 mM KPO4, pH 7.4, and quantitated spectrophotometrically at 266 nm (water ) 31 500 M-1 cm-1). Animals. New Zealand white rabbits were used for the production of polyclonal antisera. Animals were housed in the animal facility of the Veterans Affairs Hospital on the University of Colorado Health Sciences Center campus. Male Sprague-Dawley rats obtained from SASCO, Inc. (Omaha, NE) were housed on aspen bedding in climate-controlled rooms (12 h day/ night cycle) maintained at 25 °C and 50% relative humidity. Animals were given free access to lab chow (Wayne Teklab Premier Rodent Blox, Madison, WI) and tap water. Upon arrival at the animal facility animals were allowed at least 1 week of acclimation prior to experimentation. Synthesis of Haptens and Hapten-Carrier Protein Conjugates. To produce polyclonal antibodies directed against 4-HNE-sulfhydryl adducts, two procedures were used to crosslink 4-HNE to carrier proteins. In each case, numerous lysine residues present in carrier proteins were modified by derivati-

Hartley et al. zation methods to produce functional “cysteine-like” sulfhydryl residues which served as targets for 4-HNE alkylation. A 4-HNE-sulfhydryl hapten cross-linked to KLH was produced by incubating reduced glutathione, GSH (5 mg), with 4-HNE (40 µmol/mL) in 1 mL of 50 mM phosphate buffer, pH 7.4, for 4 h at room temperature. The nucleophilic addition of GSH to 4-HNE was monitored by following the decrease in 4-HNE-dependent absorbance at 224 nm (22) which indicated that approximately 80% of the 4-HNE was converted to the S-alkyl conjugate, GSH-4-HNE. The composition of the S-alkyl conjugate was validated by proton NMR in D2O and subsequent negative ion fast atom bombardment (FAB-) mass spectrometry. NMR analysis revealed loss of the aldehyde proton and trans-vinylic resonances consistent with conjugate addition to the sulfhydryl moiety of the enone system. Mass spectral examination of the deuterated compounds revealed the most abundant ion of M + 18 (m/z 480, FAB-) indicative of the hydrated aldehyde. A second predominant parent ion of m/z 465 (m/z 464, FAB-) was also observed suggesting formation of the hydroxyaldehyde adduct or cyclized hemiacetal. Isotope exchange resulted in less abundant ions of m/z 463 and 462 in FAB- that were also attributable to formation of the hydroxyaldehyde adduct or cyclized hemiacetal Next, an aliquot of the GSH-4-HNE conjugate (1.0 mL of a 10 µmol/mL solution) was added to an equal volume of PBS containing KLH (1.0 mL of a 20 mg/mL solution, ∼40 nmol). To cross-link the hapten (GSH-4-HNE) to the carrier protein (KLH), 3 µL of a 70% glutaraldehyde solution (final concentration 0.2%, v/v) was added to the 2.0 mL incubation and the reaction was allowed to proceed for 1 h at room temperature (32). Unreacted glutaraldehyde was titrated from the reaction over a period of 1 h with the addition of 200 mM glycine, pH 7.2. The product, KLH-glutaraldehyde-GSH-4-HNE, was dialyzed against 1 L of PBS overnight at 4 °C and then stored at -20 °C. An antibody was also developed to recognize protein carbonyls and 4-HNE-derived protein carbonyls by constructing a haptencarrier protein conjugate to 4-HNE-sulfhydryl epitopes derivatized with 2,4-dinitrophenylhydrazine (DNPH). In order to generate a sulfhydryl-containing carrier protein, lysine-rich KLH (10 mg) was incubated with SATA at a concentration of 43 µmol in 100 µL of dimethyl sulfoxide. The reagent, SATA, derivatizes primary amine residues to acetylated derivatives that can be converted to sulfhydryl moieties. The incubation consisted of 1.0 mL of buffer (PBS containing KLH) combined with 1.0 mL of a 20 mg of SATA/mL solution (∼43 nmol of SATA) and was maintained for 2 h at room temperature. Unreacted SATA was removed by applying the entire mixture to a desalting column (Excellulose GF-5, Pierce Chemical Co., Rockford, IL) and eluted in 50 mM NaPO4 buffer, pH 7.5. Protein recovery was monitored spectrophotometrically at 280 nm. A total of 2.0 mL of sample was incubated with 200 µL of deacetylation solution (0.5 M hydroxylamine-HCl, 25 mM EDTA in 50 mM sodium phosphate buffer, pH 7.5) to yield reactive sulfhydryl residues. Excess hydroxylamine was removed by passing the sample through a desalting column. The proteincontaining fraction (2.8 mL) was assayed for sulfhydryl content with Ellman’s reagent (5,5′-diethyldithiobenzoic acid). Total protein and sulfhydryl recovery were calculated to be 4.7 mg of KLH with 1.5 µmol of incorporated sulfhydryl. The resulting KLH-SATA conjugate was incubated with a 10-fold excess of 4-HNE (10.9 µmol of 4-HNE:1.5 µmol of SH) overnight at 4 °C. Unconjugated 4-HNE was removed by passing the sample through the desalting column, and aliquots of the recovered sample were assayed for sulfhydryl content, all of which had been titrated by 4-HNE. Because of the large molecular mass of KLH (estimated to range from 4.5 × 105 to 1.3 × 107), it was not possible to validate adduction of 4-HNE to thioacetylated KLH. Therefore, a model system substituting 3-mercaptopropionic acid (MW ) 106.14 g/mol) for the deacetylation product of SATA was employed to document formation of the respective 4-HNE-thiol adduct. Formation of the 4-HNE-mercaptopropionic acid adduct was determined by negative ion FAB mass spectrometry in incubations containing 1 mL PBS buffer, 250

Prooxidant-Initiated Lipid Peroxidation in Hepatocytes µM 4-HNE, and 500 µM 3-mercaptopropionic acid. The incubation was maintained at room temperature for 2 h prior to analysis. Mass spectrometry revealed an abundant ion of m/z 261 (FAB-, M - 1) documenting formation of the hydroxyaldehyde-sulfhydryl adduct. Adduct formation using this model system is consistent with the molar chemical determination of sulfhydryl incoporation and subsequent 4-HNE adduction during preparation of the KLH-SATA-4-HNE product described below. Having verified the formation of a 4-HNE-sulfhydryl conjugate, the KLH-SATA-4-HNE product (2.6 mL) was then derivatized with an equal volume of DNPH (0.1%, w/v, in 2.0 N HCl) for a 2 h incubation period at room temperature. The reaction was terminated by the addition of an equal volume of 20% (w/ v) trichloroacetic acid (TCA). Precipitated protein was pelleted by centrifugation, and free DNPH was extracted three times with an excess volume (1.0 mL) of ethanol/ethyl acetate (1:1, v/v). The final protein pellet was solubilized in 8.0 M guanidineHCl in 0.1 M Tris-HCl, pH 7.2. The guanidine was dialyzed from the sample against PBS overnight at 4 °C, and aliquots (1.0 mL) of the total sample (5.0 mL at 1.0 mg/mL) were analyzed spectrophotometrically at 370 nm to determine the carbonyl content using an extinction coefficient of 21 000 M-1 cm-1 for DNPH derivatives of protein carbonyls. Residual guanidine was dialyzed from the sample against PBS overnight at 4 °C, and the final hapten-carrier protein conjugate, KLHSATA-DNPH-4-HNE, was stored at -20 °C. Using this procedure, the total 4-HNE-derived carbonyl content averaged approximately 200 nmol of carbonyls/5.0 mg of KLH. Derivatization of 4-HNE-derived carbonyls present in the hapten-carrier protein conjugate is consistent with the mass spectral verification of retention of free aldehyde moieties after the conjugative steps outlined above. A separate approach was taken to generate polyclonal antisera against MDA-amine conjugates. This method relies on the direct alkylation of KLH-associated lysine residues with MDA. Briefly, KLH (20 mg/mL) was allowed to react with MDA (10 µmol) in 1 mL of 50 mM sodium phosphate, pH 7.4, for 2 h. The resulting MDA conjugate was diluted to 1.0 mg/mL in PBS and stored at -20 °C. Production of Antisera. For production of KLH-glutaraldehyde-GSH-4-HNE antisera, antigens were suspended in Ribi Adjuvant System, R-730. All other antigens, KLH-SATA4-HNE-DNPH and KLH-MDA, were suspended in Freund’s complete adjuvant for the priming injection and Freund’s incomplete adjuvant for subsequent booster injections. To initiate an immune response, 2 rabbits/antigen were given primary injections of the antigen in eight sites (sc, im, ip; 50 µg of antigen/site) as described elsewhere (23). After the original priming injection animals were given booster injections every 28 days and test bled 7-10 days after the booster injection. Blood samples (5-10 mL) were obtained from the rabbit’s ear vein. All animals were given a series of booster injections (at least two). After assessment of hapten-specific antibody titer by ELISA, the animals were exsanguinated to collect whole blood (150 mL). Whole blood was refrigerated and allowed to coagulate. Whole serum was collected and stored at -20 °C. Synthesis of Haptens To Assess Antigen-Specific Antibody Titers. 4-HNE-sulfhydryl and KLH-gluturaldehydeGSH-4-HNE antisera were assayed in ELISA (see below) with immobilized horse kidney metallothionein (MT) conjugates of 4-HNE. Antisera produced against KLH-MDA (MDA-amine) haptenic epitopes were assessed by conjugating BSA with MDA or 4-HNE. Each conjugate was produced by incubating the peptide/protein (1.0 mg) with the aldehyde (10 µmol) in 1 mL of 50 mM KPO4 buffer, pH 7.4, for 2 h at 37 °C. These conjugates were diluted in PBS for application to ELISA microtiter plates which were stored overnight at 4 °C. To screen the anti-KLH-SATA-4-HNE-DNPH sera, BSA was oxidized with the hydroxyl radical-generating system VOSO4/ H2O2 and subsequently derivatized with DNPH as previously described (24). Briefly, BSA (1.0 mg/mL/reaction) was dissolved in 10 mM NaPO4 buffer, pH 7.4, and oxidized in the presence

Chem. Res. Toxicol., Vol. 10, No. 8, 1997 897 of various concentrations of VOSO4 (0-1000 µM) and H2O2 (1.0 mM). Control incubations contained only BSA or BSA with only H2O2. Immediately following the addition of VOSO4, an equivalent volume (1.0 mL) of a DNPH solution (0.5 mM DNPH in 0.1 mM NaPO4 buffer, pH 6.3) was aliquoted into each incubation and the samples were derivatized for 1 h at room temperature. Each sample was diluted in PBS (2 µg/50 µL) for ELISA. Assessment of Antibody Titers with Enzyme-Linked Immunoabsorbent Assays (ELISA). Depending on the antisera under evaluation, specific antigens were applied to Immunolon 4 96-well microtiter immunoassay plates (Dynatech Laboratories, Inc., Chantilly, VA) at a concentration of 2 µg/50 µL of PBS/well, and the protein-aldehyde conjugates were allowed to bind to the well overnight at 4 °C. Subsequently, the antigen solution was removed, wells were rinsed with PBS (3×), and the remaining protein binding sites were blocked with a solution of 3% BSA in PBS (150 µL/well) overnight at 4 °C or for 1 h at room temperature. The solution was removed, the wells were rinsed (3×) with PBS containing 0.05% Tween-20 (PBS-T), and serial dilutions (1:25 to 1:6400 in PBS containing 1% BSA; PBS-BSA) of each antisera were aliquoted into appropriate wells (50 µL/well) and allowed to incubate for 2 h at room temperature. The plates were then rinsed (3×) with PBS-T. Secondary goat anti-rabbit IgG conjugated to horseradish peroxidase was added to the wells and incubated for 30 min (1:5000 in PBS-BSA). Plates were again rinsed (3×) and then treated with horseradish peroxidase solution consisting of 1.0 mL of 10 mg/mL 2,2′-azinobis[3-ethylbenzylthiazoline sulfonate] (ABTS) in water, diluted in 8.95 mL of McIlwain’s (13.3 mL of 0.1 M citric acid and 11.7 mL of 0.2 M sodium phosphate, pH 4.6), and 50 µL of 1% hydrogen peroxide. This substrate solution was added to each well (50 µL/well), and the reaction was allowed to develop for 15 min at 37 °C. Antibody-dependent absorbance was measured spectrophotometrically at 405 nm by a 96-well microplate reader (Molecular Devices). Isolation and Prooxidant Treatment of Rat Hepatocytes. Rats were anesthetized by intraperitoneal injection of pentobarbital sodium (65 mg/kg). Hepatocytes were isolated according to the collagenase-perfusion procedure described previously (53) and were routinely 85% viable or greater as assessed by trypan blue exclusion. Free hepatocytes were suspended in 10 mL of Krebs-Henseleit buffer supplemented with 11.5 mM HEPES [N-(2-hydroxyethyl)piperazine-N′-(2ethanesulfonic acid)], pH 7.4. Isolated hepatocytes (1.0 × 106/ mL in 20 mL total volume) were placed in 25 mL round bottom flasks, maintained at 37 °C with a constant atmosphere of 95% O2, 5% CO2 and allowed to equilibrate for at least 15 min prior to the addition of prooxidants, iron/ascorbate (0.4 mM FeSO4/ 1.0 mM ascorbate), or carbon tetrachloride (4.0 mM). Hepatocytes in separate flasks, without the addition of prooxidants, served as controls. Appropriate aliquots of the cell suspensions were taken during a 2 h time course to monitor cell viability, GSH concentration, and lipid aldehyde production, as well as lipid aldehyde-adduct formation. Quantitation of Malondialdehyde and 4-Hydroxynonenal. Malondialdehyde was assessed as an index of prooxidant-initiated lipid peroxidation by the method of Draper and colleagues (26). Briefly, at various time points after the addition of prooxidants to the hepatocyte suspensions, cells were sampled (0.5 mL, 0.5 × 106 cells) and aliquoted into a solution (1.1 mL) designed to precipitate cellular protein and prevent spurious lipid peroxidation [10% trichloroacetic acid (TCA)/50 µM butylated hydroxytoluene (BHT; dissolved in methanol)]. Cellular protein was removed by centrifugation, the acidified cellular extract (0.25 mL) was derivatized with a 250 µL thiobarbituric acid (TBA)-saturated solution, and the mixture was heated at 70 °C for 30 min. The subsequent TBA-derivatized products were extracted in an excess volume of n-butanol (0.75 mL) from which the organic layer was retained for sample analysis by RPLC. Finally, the samples were applied to a 10 µm Bondaclone C18 reversed-phase RPLC column (300 × 3.9 mm; Phenomenex), and the TBA derivative of MDA was separated isocratically (mobile phase: 15% acetonitrile, 0.6% tetrahydrofuran in 5.0

898 Chem. Res. Toxicol., Vol. 10, No. 8, 1997 mM phosphate buffer, pH 7.0) using a Beckman model 110 pump system (flow rate ) 1.0 mL/min). The MDA derivative was detected with a Shimadzu fluorescence detector (515 nm excitation, 550 nm emission) which was coupled to a HewlettPackard 3390A integrator (chart speed ) 2.0 mm/min). Using these analytical separation conditions, the TBA derivative of MDA was found to elute at 5.94 min. Formation of the lipid peroxidation product 4-HNE during the time course of prooxidant-induced hepatocellular injury was also evaluated by RPLC detection of fluorometric cyclohexanedione (CHD) derivatives (25). Hepatocyte suspensions were sampled (1.0 × 106 cells) at designated time points and added to equivalent volumes of extraction buffer (50% acetonitrile, 50% 50 mM KPO4, pH 3.0) containing 1.0 nmol/mL 4-hydroxy-2octenal which served as the internal standard. Cellular protein was precipitated by addition of 100 µL of 70% PCA and pelleted by centrifugation at 10 000 rpm. A 1 mL aliquot of the acidified sample was derivatized as described (25). The CHD derivatized samples were then applied to a 5 µm Ultramex (300 × 3.9 mm) C18 Phenomenex column. The CHD derivative of 4-HNE was eluted with a gradient mobile phase system (mobile phase A, water; mobile phase B, 100% acetonitrile). The flow rate was maintained at 1.0 mL/min as mobile phase B was increased linearly from 0% to 40% over 50 min. The CHD derivatives were detected using a fluorescence detector (380 nm excitation, 415 nm emission ). Using these analytical conditions, the CHD derivative of 4-HNE eluted in 38.5 min. The retention time of the internal standard, CHD-4-hydroxyoctenal, was 37.5 min. All retention times were confirmed by comparisons with synthetic standards. Detection and Quantitation of Protein Carbonyls as 2,4-Dinitrophenylhydrazine (DNPH) Derivatives. Hepatocellular protein-bound carbonyl derivatives were evaluated using a modification of the methods first described elsewhere (28, 29). Briefly, an aliquot of the cell suspension (0.5 mL, 0.5 × 106 cells) was added to an equivalent volume (0.5 mL) of 0.1% DNPH in 2.0 N HCl (w/v) and allowed to incubate for 1 h at room temperature. This reaction was terminated and total cellular protein precipitated by the addition of an equivalent volume (1.0 mL) of 20% TCA (w/v). Cellular protein was rapidly pelleted by centrifugation at 10 000 rpm, and the supernatant was decanted. Excess unincorporated DNPH was extracted (3×) with an excess volume (0.5 mL) of ethyl acetate:ethanol (1:1, v/v). Following the extraction, the recovered cellular protein was dried under a nitrogen stream and solubilized in a solution (1.0 mL) of Tris-buffered 8.0 M guanidine-HCl, pH 7.2. Proteinbound DNPH was assessed spectrophotometrically between 190 and 400 nm, where the maximum wavelength for absorption of protein-bound hydrazones is 366-370 nm and served as an index of total protein-bound carbonyls. Quantitation of Cellular Glutathione. Reduced GSH was detected and quantitated as a 2,4-dinitrofluorobenzene (DNFB) derivative as described previously (30). The GSH derivative was applied to a reversed-phase Maxsil 10 µm, 250 × 4.6 mm C18NH2 column (Phenomenex) and separated with a gradient mobile phase system [mobile phase A, 80% methanol; mobile phase B, 80% methanol, 20% acetic acid/ammonium acetate solution (756 mL of glacial acetic acid, 244 mL of water, 308 g of ammonium acetate)]. The analysis and quantitation of DNBGSH were performed with a Shimadzu 10AV detector set to 350 nm, coupled to a Shimadzu LC-600 dual pump system (flow rate ) 1.0 mL/min) and a Shimadzu CR601 integrator (chart speed ) 2.0 mm min-1). Using these separation conditions the DNFB derivatives of GSH eluted at 31.4 min. Immunoprecipitation of 4-HNE and MDA Adducts Formed in Peroxidized Hepatocytes. Formation of lipidderived aldehyde products was assessed by immunoprecipitation-immunoblotting analyses using the antibodies generated against MDA-, 4-HNE-, and DNPH-derivatized proteins. Immediately after an aliquot (1.0 × 106/sample) of cell suspension was obtained from control, iron/ascorbate-treated, or carbon tetrachloride-treated hepatocytes, the cells were rapidly pelleted at 1000 rpm and the supernatant was removed. The cells were

Hartley et al. resuspended in ice cold modified RNase/DNase buffer (31) (180 µL/sample of a stock solution containing 1.0 mg/mL RNase, 1.0 mg/mL DNase I, 5 mM MgCl2, all in 500 mM Tris-HCl, pH 7.5, and 45% glycerol). The cell suspensions in RNase/DNase solution were lysed by the addition of a 2.5% SDS (w/v) solution (20 µL/sample) and mixed by repeated inversion. Insoluble cellular material was solubilized, and SDS was neutralized by the addition of an immunoprecipitation (IP) buffer to each sample (200 µL/sample of a solution containing 100 mM Tricine, pH 8.2, 300 mM NaCl, 0.75% Triton X-100, 10 mM EDTA, 0.02% NaN3, 1.0 mM benzamidine, and 1.0 mM phenylmethanesulfonyl fluoride; PMSF). Peptide protease-mediated protein degradation in the sample was inhibited by the addition of a cocktail of specific peptide protease inhibitors (20 µg/sample from a solution containing 1.0 mg/mL each of aprotinin, pepstatin, leupeptin, and antipain dissolved in 50% ethanol) in addition to the benzamidine and PMSF present in the IP buffer. An aliquot (4 µL) of antiserum (MDA-KLH or HNESH) was added to the 400 µL of total sample volume containing 1.0 × 106 cells, and the samples were allowed to incubate overnight at 4 °C. When the anti-KLH-SATA-4-HNE-DNPH sera were used to immunoprecipitate aldehyde-bound proteins, an aliquot of DNPH solution (0.5 mM DNPH in 0.1 M NaPO4 buffer, pH 6.3) was added to the cell suspension in IP buffer and incubated for 1 h at room temperature to derivatize protein-bound aldehyde moieties. Subsequently, the anti-4-HNE-DNPH serum (4 µL) was added and allowed to incubate with the derivatized hepatocellular proteins overnight at 4 °C. In a separate tube, the protein A-Sepharose CL-4B matrix was prepared by initial hydration of the matrix in 1.0 mL of deionized/distilled water. Once hydrated the matrix was preadsorbed in a solution of IP buffer containing 1% BSA. This mixture was agitated overnight at room temperature. An aliquot (50 µL) of the protein A-Sepharose CL-4B slurry (6%, w/v) in IP buffer was added to each sample to immunoprecipitate protein-aldehyde-antibody interactions, and the mixture was agitated for 1 h at room temperature. After immunoprecipitation with protein ASepharose, the matrix was washed 3 times with IP buffer followed by a final single wash with 10 mM Tricine, pH 7.8. Control experiments were also performed to ensure that each antibody immunoprecipitated protein-aldehyde conjugates. Synthetic conjugates of MDA-BSA, GSH-4-HNE, or GSH-4-HNEDNPH were prepared by incubating the respective peptide/ protein (1.0 mg) with aldehyde (10 µmol) in 1 mL of 50 mM KPO4 buffer, pH 7.4, overnight at 4.0 °C; 250 µL of each conjugate was added to the samples obtained from hepatocytes incubated for 120 min in the absence or presence of iron/ ascorbate or carbon tetrachloride. Electrophoresis and Electroimmunoblot (Western) Analysis of 4-HNE- and MDA-Modified Proteins. Electrophoresis sample buffer (50 µL of a solution containing 62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.02% bromphenol blue, 5% β-mercaptoethanol) was added to each immunoprecipitation, and the samples were incubated at 100 °C for 10 min. The samples were subjected to standard SDS-PAGE (32). Proteins separated by SDS-PAGE were transferred to nitrocellulose in 192 mM glycine, 25 mM Tris-HCl, 1.3 mM SDS, pH 8.2, over 2 h at 0.7 A. Subsequently, the blotted nitrocellulose was equilibrated in PBS-T and incubated in 1% BSAPBS-T for 1 h. Primary antisera were diluted in PBS-T (1:500, antibody:PBS-T), applied to the nitrocellulose, and incubated for 1 h. Subsequently, the nitrocellulose was rinsed (3×) for periods of 5 min in fresh PBS-T followed by incubation of the nitrocellulose with a biotin-conjugated secondary antibody (goat anti rabbit IgG) diluted in PBS-T (1:5000) for 0.5 h. The nitrocellulose was again rinsed (3×) and incubated for 0.5 h with the amplification system, which consisted of streptavidinconjugated horseradish peroxidase diluted in PBS-T (1:5000). Positive interactions were visualized with the enhanced chemiluminescence substrate for horseradish peroxidase (AmershamECL). Immunopositive interactions were quantitated using the NIH Image Version 1.6 software.

Prooxidant-Initiated Lipid Peroxidation in Hepatocytes Statistical Analysis. Data were analyzed by two-way ANOVA (time by treatment) to detect differences in the various biochemical parameters over the entire time course of prooxidant exposure. Post-hoc comparisons of mean values for determination of significant differences over time (significance level of p < 0.05 or greater) were by the Tukey b-test. These analyses were conducted using the Version 4 Statistical Package (Crunch Software Corp., Oakland, CA).

Results Characterization of Anti-4-HNE-Sulfhydryl Serum. The structures of the hapten-carrier protein conjugates are illustrated in Figure 1 with the corresponding ELISA results. Antisera to 4-HNE-sulfhydryl adducts (4-HNESH) were immunoreactive against immobilized metallothionein-4-HNE (MT-HNE) conjugates where hapten-specific antibody response titered out with decreasing antisera concentrations (Figure 1a). Preimmune serum for these antisera were not immunopositive to the MT-HNE conjugate. On the other hand, postimmune serum displayed specific immunopositive responses for 4-HNE-MT-sulfhydryl epitopes, which were approximately 10-fold greater than that of the respective preimmune serum. ELISA and immunoblotting experiments demonstrated the absence of cross-reactivity of the HNESH antisera with MDA-amine protein haptens. Additional experiments were performed to determine if the HNESH antisera were specific for sulfhydryl conjugates of 4-HNE. This was accomplished by forming the 4-HNE conjugate of BSA which, because of the absence of any free sulfhydryl groups, serves as a representative model for 4-HNE-amine and 4-HNE-histidine epitopes. Western blot analysis and ELISA demonstrated that the HNESH antiserum did not recognize Schiff’s base or histidine Michael adduct epitopes generated by incubating 4-HNE with BSA. These results suggest that, at the level of immunoblotting and ELISA, the HNESH antisera are specific for sulfhydryl adducts. Characterization of Anti-KLH-SATA-4-HNEDNPH Sera. Additional antisera were generated against KLH-SATA-4-HNE-DNPH hapten-carrier protein conjugates (HNESHDNP). The requirement for the use of these antisera is that samples must be derivatized with DNPH. As a positive control BSA was oxidized with increasing concentrations of VOSO4/H2O2 and derivatized with DNPH. These DNPH-BSA conjugates were then used to screen the antibody titers of antisera developed toward the KLH-SATA-4-HNE-DNPH hapten. Following immobilization of the BSA-DNPH conjugates in microtiter plate wells, the antisera were screened by ELISA. Each antiserum displayed significant antibody titers to general carbonyl moieties in oxidized and DNPHderivatized BSA (Figure 1b), and the immunospecificity for DNPH-derivatized protein in the postimmune serum was approximately 10-fold greater than that observed with the preimmune serum. Furthermore, the immunopositive response was enhanced by increasing the concentration of oxidant (VOSO4) where this functions to increase the concentration of DNPH-derivatizable protein carbonyl moieties in BSA. Although these antisera were generated to hydrazone derivatives of protein-bound 4-HNE, it was not possible to differentiate between types of protein-DNPH hydrazones, a characteristic reported by other investigators preparing antibodies against protein-DNPH hydrazones (24). Characterization of Anti-KLH-MDA Sera. In addition to the battery of 4-HNE-protein adduct-specific

Chem. Res. Toxicol., Vol. 10, No. 8, 1997 899

antisera, antisera were generated to MDA-protein adducts. Antisera produced against the KLH-MDA haptencarrier protein conjugates were exquisitely immunoreactive against MDA-protein conjugates in ELISA (Figure 1c) where postimmune serum was 10-20-fold more reactive with immobilized BSA-MDA conjugates when compared to the preimmune serum. Additional ELISA and immunoblotting experiments revealed that antisera produced against the KLH hapten-carrier protein did not cross-react with 4-HNE-sulfhydryl haptens. Cell Viability, Glutathione Status, and Lipid Peroxidation. To assess the effect of prooxidantinduced lipid peroxidation in isolated hepatocytes and establish conditions for the use of these antisera for adduct detection, cell viability and cellular reduced GSH status were measured. As presented in Figure 2, throughout the 2 h experiment cell viability remained above 80% in control, untreated, hepatocyte suspensions. The addition of iron/ascorbate did not significantly affect cell viability over the 2 h time course evaluated, whereas CCl4 caused a significant decrease (p < 0.0002) in cell viability. The data presented in Figure 3 show that the concentrations of GSH in control cells did not change significantly during the 2 h time course of the experiment. However, cellular reduced GSH was significantly reduced by CCl4 (p < 0.002) or iron/ascorbate (p < 0.0001) over the 2 h time course with the largest reduction occurring in the cells exposed to iron/ascorbate. To ensure that the prooxidant systems used to initiate hepatocellular lipid peroxidation were indeed causing such an effect, simultaneous measurements of lipid peroxidation were performed during prooxidant-induced oxidative stress. As noted from Figure 4, there was no significant accumulation of MDA in untreated hepatocyte suspensions. However, significant increases in MDA accumulation during the 2 h time course of the experiments were observed in incubations containing iron/ ascorbate (p < 0.000 01) and CCl4 (p < 0.0004). The difference in concentrations of MDA present in the incubations at 120 min suggests that iron/ascorbate was approximately 2-fold more efficient at initiating lipid peroxidation as compared to CCl4. A sensitive fluorometric RPLC assay was used to detect the CHD derivative of 4-HNE. It is apparent from Figure 5 that the concentrations of this aldehyde in hepatocyte incubations exposed to iron/ascorbate or CCl4 increased above those measured in control incubations. However, because of the marked variation in the concentration of 4-HNE in the incubations, these trends were not statistically significant. This observation is consistent with previous reports concerning the presence of multiple and very efficient pathways for the metabolism of 4-HNE in isolated hepatocytes (12, 33). The data presented in Figure 6 establish that when hepatocytes were incubated in the absence of either prooxidant, significant elevations in protein hydrazone formation were not observed. The data in this figure do, however, demonstrate that during the time course of exposure to either iron/ascorbate or CCl4, protein hydrazone formation in hepatocytes was proportional to MDA production and inversely proportional to the depletion of cellular glutathione. During the 120 min time course of the experiments, protein-bound carbonyl content of iron/ascorbate hepatocytes increased approximately 2-fold (p < 0.000 01), while protein-bound carbonyls in hepatocytes incubated with CCl4 increased approximately 1.5-fold (p < 0.002).

900 Chem. Res. Toxicol., Vol. 10, No. 8, 1997

Hartley et al.

Figure 1. ELISA analysis of antisera produced against 4-HNESH (a), 4-HNESHDNP (b), and MDA-KLH (c) protein haptens. For each antigen, the triangles and circles represent antibody titers measured in preimmune sera obtained from two individual rabbits. The inverted triangles and squares are antibody titers detected in postimune sera from the same rabbits.

Detection of Aldehyde-Adducted Proteins by Immunoprecipitation and Immunoblotting. Results obtained by immunoprecipitation and immunoblot analysis using the MDA-KLH antisera demonstrate that iron/ ascorbate-initiated lipid peroxidation caused a more pronounced formation of MDA-adducted protein as

compared to CCl4 (Figure 7). These data also show that both immunopositive response and the total number of immunopositive proteins increased with time of exposure to both iron/ascorbate and CCl4. Again, iron/ascorbate was more effective than CCl4 at initiating this apparent sequestration of MDA by specific cellular nucleophiles.

Prooxidant-Initiated Lipid Peroxidation in Hepatocytes

Chem. Res. Toxicol., Vol. 10, No. 8, 1997 901

Figure 2. Time course for the viability of isolated hepatocyte suspensions during CCl4- or iron/ascorbate-induced injury. At each time point, circles, squares, and triangles represent cellular viability in control, CCl4-treated, or iron/ascorbate-treated hepatocytes, respectively. Values are means ( standard error of the mean.

Figure 4. Time course of MDA formation in isolated hepatocyte suspensions during exposure to CCl4 or iron/ascorbate. At each time point, circles, squares, and triangles represent MDA formed in control, CCl4-treated, or iron/ascorbate-treated hepatocytes, respectively. Values are means ( standard error of the mean for 5 individual experiments.

Figure 3. Reduced glutathione status of isolated hepatocyte suspensions during exposure to CCl4 or iron/ascorbate. At each time point, circles, squares, and triangles represent cellular glutathione in control, CCl4-treated, or iron/ascorbate-treated hepatocytes, respectively. Values are means ( standard error of the mean for five individual experiments.

Figure 5. Time course of 4-HNE formation in isolated hepatocyte suspensions during exposure to CCl4 or iron/ascorbate. At each time point, circles, squares, and triangles represent 4-HNE in control, CCl4-treated, or iron/ascorbate-treated hepatocytes, respectively. Values are means ( standard error of the mean for five individual experiments.

Immunoblots paralleled those results obtained with biochemical measures of lipid peroxidation. In general, the molecular weights of each immunopositive protein, determined relative to molecular weight standards, were 20, 25, 28, 30, 33, 38, 41, 45, 80, 82, 85, 130, and 150 kDa. Control experiments were performed to establish that each of the three antibodies could be immunoprecipitated by addition of their respective conjugates to immunoprecipitation incubations containing peroxidized hepatocytes. The addition of MDA-BSA, HNE-GSH, or GSH-4-HNE-DPNH conjugates to immunoprecipitation incubations prepared from peroxidized hepatocytes abolished the appearance of immunopositive proteins routinely detected by immunoprecipitation-immunoblot analysis (data not shown). These results further demonstrated a specificity of the antisera to identify aldehyde-specific epitopes in oxidized hepatocellular proteins. Immunoprecipitation-immunoblot analyses with the HNESH did not detect a prooxidant-dependent formation of 4-HNE-modified proteins (Figure 8), a result that is concordant with measurements for 4-HNE accumulation. Interestingly, several adducted proteins were detected in

cells as a function of time but not prooxidant exposure. In both control incubations, and incubations including CCl4 or iron/ascorbate, proteins of 80, 100, and 150 kDa were detected after incubating the hepatocyte suspensions for 60 min. Notably, these proteins correspond to some of the higher molecular weight proteins detected with the MDA-KLH antiserum. Immunoprecipitation-immunoblot analyses with the HNESHDNP antisera (Figure 9) paralleled the treatment- and time-dependent formation of protein-bound carbonyls presented in Figure 6. It is noteworthy that fewer proteins and a weaker immunoreactive response to aldehyde-modified proteins were observed with this antiserum as compared to the MDA-KLH antiserum. Interestingly, the immunopositive proteins detected with HNESHDNP correspond to those higher molecular weight proteins detected with the MDA-KLH antiserum (Figure 7). The estimated molecular weight of these proteins is 80, 100, 130, and 150 kDa. Overall, it appears that iron/ ascorbate is a more effective prooxidant than CCl4 at initiating formation of aldehyde-modified cellular proteins since the immunoreactivity of proteins detected

902 Chem. Res. Toxicol., Vol. 10, No. 8, 1997

Figure 6. Time course for formation of protein carbonyls detected as 2,4-dinitrophenylhydrazine derivatives in isolated hepatocyte suspensions during exposure to CCl4 or iron/ascorbate. At each time point, circles, squares, and triangles represent protein carbonyls formed in control, CCl4-treated, or iron/ ascorbate-treated hepatocytes, respectively. Values are means ( standard error of the mean for five individual experiments.

Figure 7. Immunoprecipitation-immunoblot analysis of hepatocellular protein obtained from hepatocytes exposed to iron/ ascorbate (Fe/Asc) or CCl4 and untreated, control hepatocytes. Hepatocytes were sampled at 0, 30, 60, and 120 min as indicated and immunoprecipitated with the MDA-KLH antisera. The estimated molecular weights (kDa) of immunoreactive proteins are as labeled (left) and were calculated from their Rf values in relation to those of known standards. The IgG carried through the immunoprecipitation is labeled at 55 kDa.

from iron/ascorbate-treated cells occurs at earlier time points (30 min) and with greater intensity. Consistent with the results obtained using MDA-KLH and HNESH antisera in immunoprecipitation-immunoblot analyses, specific proteins were detected in control hepatocyte suspensions as a function of time.

Discussion Immunization of rabbits with synthetic 4-HNE- or MDA-protein haptens resulted in generation of high titers of antibodies which were shown by ELISA to recognize aldehyde-specific epitopes. As presented in Figure 1, this was observed for antibodies prepared against 4-HNE-sulfhydryl haptens and MDA-amine haptens. As noted, antisera prepared against 4-HNE-DNPH

Hartley et al.

Figure 8. Immunoprecipitation-immunoblot analysis of hepatocellular protein obtained from hepatocytes exposed to iron/ ascorbate (Fe/Asc) or CCl4. The last four lanes display hepatocellular protein profiles of untreated, control hepatocytes. Hepatocytes were sampled at 0, 30, 60, and 120 min as indicated and immunoprecipitated with the HNESH antisera. The estimated molecular weights (kDa) of immunoreactive proteins are as labeled (left) and were calculated from their Rf values in relation to those of known standards. The IgG carried through the immunoprecipitation is labeled at 55 kDa.

Figure 9. Immunoprecipitation-immunoblot analysis of hepatocellular protein obtained from untreated, control hepatocytes or from hepatocytes exposed to CCl4 or iron/ascorbate (Fe/Asc). Hepatocytes were sampled at 0, 30, 60, and 120 min as indicated and immunoprecipitated with the HNESHDNP antisera. The estimated molecular weights (kDa) of immunoreactive proteins are as labeled (right) and were calculated from their Rf values in relation to those of known standards (left). The IgG carried through the immunoprecipitation is labeled at 55 kDa.

were not aldehyde selective but were specific for aldehyde-adducted proteins that were prepared as hydrazone derivatives. Reports have appeared describing the generation of antisera-directed 4-HNE-LDL (34), 4-HNEKLH (35), MDA-LDL (36), or MDA-BSA (37). However, prior to the present communication, the antibodies previously produced against 4-HNE-protein haptens were directed toward 4-HNE-amine epitopes of lysine-rich LDL or KLH. While these antisera have proven useful in detecting aldehyde-protein adducts in tissue sections (6, 14, 19, 38), there has been very limited success in identifying specific cellular protein targets. Antibodies have also been produced against irreversible 2-pentylpyrrole adducts with lysine residues in model proteins exposed to 4-HNE (39, 40). For the studies described

Prooxidant-Initiated Lipid Peroxidation in Hepatocytes

here, it was postulated that 4-HNE would, under physiological conditions, react most rapidly with sulfhydryl residues of proteins, and a novel 4-HNE-sulfhydryl hapten-carrier protein was therefore prepared for antibody production. Likewise, as far as we are aware, this report is the first to describe the production of antisera against the MDA-KLH carrier protein. Our ELISA and immunoblot characterizations revealed that the antisera were aldehyde-specific; the 4-HNESH antisera did not recognize MDA-adducted haptens, and the MDA-amine antisera did not recognize 4-HNE-protein haptens. The results presented here corroborate those from previous studies documenting the ability of iron/ascorbate or CCl4 to initiate lipid peroxidation in isolated hepatocytes. However, in the present study, the association between cellular toxicity, cellular GSH status, lipid peroxidation, and lipid aldehyde-protein adducts was evaluated. Collectively, our results demonstrated that CCl4 and iron/ascorbate initiated moderate and severe lipid peroxidation, respectively. This difference in peroxidative potential was apparent in biochemical assays measuring formation of MDA, 4-HNE, and aldehydeprotein adducts by DNPH derivatization as well as by immunoprecipitation-immunoblot analyses. The results presented here are in agreement with those reported previously (40, 41), where the presence of aldehydeprotein adducts as measured by chemiluminescence and fluorescence (42) were associated with the formation of the carbonyl byproducts of lipid peroxidation, MDA and ethane. Also, the data in Figure 2 indicate that CCl4 caused cell viability to decrease approximately 30-35% over the 2 h exposure period, an observation consistent with previous reports (41-42). Iron/ascorbate caused only mild GSH depletion and no cytotoxicity in this study or in a previous study (37). A previous report demonstrated the use of antibodies generated to KLH-4-HNE haptens for detection of aldehyde-adducted proteins in isolated hepatocytes cryopreserved at -70 °C prior to treatment with potent oxidizing agents (18). These investigators observed immunopositive staining of several proteins in control hepatocytes that increased significantly in intensity following prooxidant exposure. Consistent with that report, we noted the occurrence of certain immunopositive hepatocellular proteins in control hepatocytes and identified novel proteins specifically alkylated by the aldehydic products of lipid peroxidation as a result of prooxidant-induced injury. However, since hepatocytes are known to have extremely efficient and highly integrated pathways for the oxidation of aldehydic products of lipid peroxidation (12, 33), our experimental design emphasized the use of freshly isolated hepatocytes in order to characterize the time course of prooxidantinduced lipid peroxidation and covalent alkylation of cellular macromolecules by MDA or 4-HNE. It is evident from the data presented here that hepatocytes undergo a low level of oxidative stress in the absence of prooxidants in that control hepatocytes displayed slight increases in indices of oxidative stress as indicated by increased concentrations of MDA, proteinbound carbonyls, and immunoreactive adducted proteins. However, these indices of oxidative stress in control incubations were significantly less prominent than the accumulation of MDA or alkylated proteins, especially those alkylated with MDA, after a 120 min exposure to either prooxidant (Figures 4, 6, and 7). In direct contrast to the effect of prooxidants on MDA accumulation and

Chem. Res. Toxicol., Vol. 10, No. 8, 1997 903

protein adduct formation, prooxidant exposure did not cause 4-HNE to accumulate or form protein adducts. The procedures employed here to derivatize and quantitate 4-HNE were validated using synthetic standards, and our analysis included the internal standard 4-hydroxyoctenal. Thus, the results presented in Figure 5 accurately reflect profiles of 4-HNE in hepatocytes exposed to iron/ascorbate or CCl4. The absence of a significant accumulation of 4-HNE in the hepatocyte incubations is most likely related to the efficiency of 4-HNE metabolism by oxidative, reductive, and conjugative pathways reported to exist in hepatocytes (33). Taken together, the data in Figures 4 and 5 suggest that during CCl4- or iron-induced oxidative stress, MDA is the primary aldehydic product of lipid peroxidation that accumulates and, in turn, aldehyde-protein adducts increase proportionally. The observation of increased immunoreactive protein-bound carbonyls in hepatocytes exposed to CCl4 and iron/ ascorbate suggests that the antibodies directed against the DNPH-derivatized 4-HNE epitopes are not aldehydespecific but recognize aldehyde-adducted proteins derivatized with DNPH. Our mass spectral analysis of the GSH-HNE adduct suggests that the aldehyde moiety exists primarily in the hydrated form but was also present in the hydroxyaldehyde or possibly hemiacetal configuration. Thus, it is possible that the polyclonal antibodies produced against the KLH-GSH-HNE epitopes could recognize protein-HNE epitopes having the free aldehyde or hemiacetal moiety. More specifically, antibodies against MDA-amine, 4-HNE-sulfhydryl, and DNPH-derivatized 4-HNE conjugates detected several common proteins alkylated during the time course of lipid peroxidation when used in immunoprecipitation-immunoblot assays. These results clearly demonstrate the formation of aldehydeadducted proteins in freshly isolated hepatocytes undergoing oxidative stress induced by iron/ascorbate or CCl4. Further, the observation that certain of the same proteins are identified by antibodies directed against MDA or 4-HNE epitopes is also noteworthy and novel. The reasons for choosing to use these combined immunoprecipitation-antibody procedures rather than standard Western blotting were severalfold. First, the formation of aldehyde-protein conjugates was postulated to be a selective event with respect to alkylation of protein nucleophiles. Thus, clearing the sample of nonadducted protein with the antiserum increases the potential for detecting an aldehyde-protein adduct. This approach is particularly important in the case where an abundant, nonadducted protein might have precluded binding to the nitrocellulose of similarly sized adducted proteins present in far lower concentrations. Second, the assay conditions were quite stringent where the immunoprecipitation was carried out in buffer containing a high concentration of NaCl. Therefore in using this immunoprecipitation methodology, we were able to enhance sensitivity to highaffinity protein-aldehyde conjugates while reducing nonspecific immunoreactivity. One limitation of this approach is the inability to detect immunoreactive proteins of 50-60 kDa. This is due to the presence of IgG which is carried through the immunoprecipitation and is detected at 55 kDa during the subsequent immunoblot. This abundant 55 kDa band did, however, serve as an internal standard for both application of protein to SDS-PAGE and successful immunoblotting. It is apparent from these immunochemical data that MDA is produced extensively and is efficiently seques-

904 Chem. Res. Toxicol., Vol. 10, No. 8, 1997

tered by proteins during prooxidant-initiated lipid peroxidation. In contrast, accumulation of free 4-HNE in hepatocyte incubations was insignificant, and 4-HNEadducted proteins were observed at much lower levels as compared to MDA-adducted proteins. It is interesting that two separate antibodies developed to detect 4-HNEprotein adducts, HNESH and HNESHDNP, and the MDA-KLH all detected proteins of 80, 130, 100, and 150 kDa. This may suggest that in instances of uncontrolled lipid peroxidation lipid aldehydes are sequestered by a battery of specific protein nucleophiles. Interestingly, we observed that the formation of specific lipid aldehyde-protein adducts did not correspond to cell death. Even though the apparent formation of covalent adducts to specific proteins increases with time of exposure to prooxidants, cell death did not appear to be an immediate consequence of protein alkylation by MDA or 4-HNE. On the other hand, protein alkylation, as well as glutathione depletion, did precede cell death. However, the cause-effect relationship of these two events with cell death remains to be established. Further studies which extend the time course beyond 2 h may better resolve this question. It is important to emphasize that the antibodies used in the present studies were developed against 4-HNE-HNESH epitopes and thus would not recognize the 4-HNE-derived pyrrole adducts formed with lysine residues of model proteins described elsewhere (39, 40). While the 4-HNE-pyrrole adducts are reported to be irreversible although minor products of 4-HNE-protein modification, their involvement in the oxidative stress mediated by prooxidants such as CCl4 and iron warrants further investigation. These studies further extend that which is known about the generalized aspects of prooxidant-induced lipid peroxidation in isolated hepatocytes and demonstrate that once formed, aldehydic products of lipid peroxidation, especially MDA, act further by binding to specific cellular macromolecules. Although the consequence of this is still unresolved, attention can now be focused on the identification of these proteins with regard to their function in the cell and whether covalent alkylation has any adverse consequences to cell function. Alternatively, it is possible that there are relatively abundant lysineor cysteine-rich cellular proteins that perform a noncatalytic, chaperon function by adducting and eliminating potentially toxic aldehydic products of lipid peroxidation. Certainly, identification and characterization of these proteins will be essential to determine their specific roles in prooxidant-induced liver injury. Also, it will be important to extend these ex vivo observations to that which occurs in vivo during prooxidant-initiated hepatic injury.

Acknowledgment. This work was supported by National Institute of Health AA-09300 and DK34914 (D.R.P.) and Predoctoral Fellowship AA-05370 (D.P.H.).

References (1) Stadtman, E. R. (1992) Protein oxidation and aging. Science 257, 1220-1224. (2) Comporti, M. (1985) Biology of disease: lipid peroxidation and cellular damage in toxic liver injury. Lab. Invest. 53, 599-622. (3) Kawase, T., Kato, S., and Lieber, C. S. (1989) Lipid peroxidation and antioxidant defense systems in rat liver after chronic ethanol feeding. Hepatology 10, 815-821. (4) Tsukamoto, H., Horne, W., Kamimura, S., Niemela¨, O., Parkkila, S., Yla¨-Herttuala, S., and Brittenham. (1995) Experimental liver cirrhosis induced by alcohol and iron. J. Clin. Invest. 96, 620630.

Hartley et al. (5) Britton, R. S., O’Neill, R., and Bacon, B. R. (1990) Hepatic mitochondrial malondialdehyde metabolism in rats with chronic iron overload. Hepatology 11, 93-97. (6) Houglum, K., Filip, M., Witztum, J. L., and Chojkier, M. (1990) Malondialdehyde and 4-hydroxynonenal protein adducts in plasma and liver of rats with iron overload. J. Clin. Invest. 86, 19911998. (7) Palinski, W., Rosenfeld, M. E., Yla¨-Herttuala, S., Gurtner, G. C., Socher, S. S., Butler, S. W., Parthasarathy, S., Carew, T. E., Steinberg, D., and Witztum, J. L. (1989) Low density lipoprotein undergoes oxidative modification in vivo. Proc. Natl. Acad. Sci. U.S.A. 86, 1372-1376. (8) Lung, C. C., Pinnas, J. L., Yahya, M. D., Meinke, G. C., and Mooradian, A. D. (1992) Malondialdehyde modified proteins and their antibodies in the plasma of control and streptozotocin induced diabetic rats. Life Sci. 52, 329-337. (9) Shah, G., Pinnas, J. L., Lung, C. C., Mahmoud, S., and Mooradian, A. D. (1994) Tissue-specific distribution of malondialdehyde modified proteins in diabetes mellitus. Life Sci. 55, 1343-1349. (10) Young, I. S., Trouton, T. G., Torney, J. J., McMaster, D., Callender, M. E., and Trimble, E. R. (1994) Antioxidant status and lipid peroxidation in hereditary haemochromatosis. Free Radical Biol. Med. 16, 393-397. (11) Esterbauer, H., Zollner, H., and Schaur, R. (1990) Aldehydes formed by lipid peroxidation: mechanisms of formation, occurrence, and determination. In Membrane Lipid Peroxidation (VigoPelfrey, C., Ed.) Vol. I, pp 240-268, CRC Press, Boca Raton, FL. (12) Poli, G., Dianzani, M. U., Cheeseman, K. H., Slater, T. F., Lang, J., and Esterbauer, H. (1985) Separation and characterization of the aldehydic products of lipid peroxidation stimulated by carbon tetrachloride or ADP-iron in isolated rat hepatocytes and rat liver microsomal suspensions. Biochem. J. 227, 629-638. (13) Grune, T., Siems, W. G., and Schneider, W. (1993) Accumulation of aldehydic lipid peroxidation products during postanoxic reoxygenation of isolated rat hepatocytes. Free Radical Biol. Med. 15, 125-132. (14) Benedetti, A., Esterbauer, H., Ferrali, M., Fulceri, R., and Comporti, M. (1982) Evidence for aldehydes bound to liver microsomal protein following CCl4 and BrCCl3 poisoning. Biochim. Biophys. Acta 711, 345-356. (15) Paradisi, L., and Dianzani, M. U. (1975) Cyclic nucleotide levels in the liver of rats treated with CCl4. Chem.-Biol. Interact. 26, 1-9. (16) Benedetti, A., Fulceri, R., and Comporti, M. (1984) Inhibition of calcium sequestration activity of liver microsomes by 4-hydroxyalkenals originating from the peroxidation of liver microsomal lipids. Biochim. Biophys. Acta 793, 489-493. (17) Siems, W. G., Hapner, S. J., and Van Kuijk, F. J. G. M. (1996) 4-Hydroxynonenal inhibits Na+-K+-ATPase. Free Radical Med. Biol. 20, 215-223. (18) Uchida, K., Szweda, L. I., Chae, H.-Z., Statdman, E. R. (1993) Immunochemical detection of 4-hydroxy-2-nonenal protein adducts in oxidized hepatocytes. Proc. Natl. Acad. Sci. U.S.A. 90, 8742-8746. (19) Bedossa, P., Houglum, K., Trautwein, C., Holstege, A., and Chojkier, M. (1994) Stimulation of R1(I) gene expression is associated with lipid peroxidation in hepatocellular injury: a link to tissue fibrosis? Hepatology 19, 1262-1271. (20) Neale, T. J., Ojha, P. P., Exner, M., Poczewski, H., Ruger, B., Witztum, J.L., Davis, P., and Kerjaschki, D. (1994) Proteinuria in passive heymann nephritis is associated with lipid peroxidation and formation of adducts on type IV collagen. J. Clin. Invest. 94, 1577-1584. (21) Mitchell, D. Y., and Petersen, D. R. (1991) Inhibition of rat hepatic mitochondrial aldehyde dehydrogenase-mediated acetaldehyde oxidation by trans-4-hydroxy-2-nonenal. Hepatology 13, 728-734. (22) Ishikawa, T., Esterbauer, H., and Sies, H. (1986) Role of cardiac glutathione transferase and of the glutathione s-conjugate export system in biotransformation of 4-hydroxynonenal in the heart. J. Biol. Chem. 261, 1576-1581. (23) Harlow, E., and Lane, D. (1988) In Antibodies: A Laboratory Manual (Harlow, E., and Lane, D., Eds.) Cold Spring Harbor, NY. (24) Keller, R. J., Halmes, N. C., Hinson, N. C., Hinson, J. A., and Pumford, N. R. (1993) Immunochemical detection of oxidized proteins. Chem. Res. Toxicol. 6, 430-433. (25) Molde´us, P., Hogberg, J., and Orrenius, S. (1978) Isolation and use of liver cells. Methods Enzymol. 60-71. (26) Draper, H. H., Squires, E. J., Mahmoodi, H., Wu, J., Agarwal, S., and Hadley, M. (1993) A comparative evaluation of thiobarbituric acid methods for the determination of malodialdehyde in biological materials. Free Radical Biol. Med. 15, 353-363. (27) Yoshino, K., Matsuura, T., Sano, M., Saito, S., and Tomita, I. (1986) Fluorometric liquid chromatographic determination of aliphatic aldehydes arising from lipid peroxides. Chem. Pharm. Bull. 34, 1694-1700.

Prooxidant-Initiated Lipid Peroxidation in Hepatocytes (28) Benedetti, A., Fulceri, R., Ferrali, M., Ciccoli, L., Esterbauer, H., and Comporti, M. (1982) Detection of carbonyl functions in phospholipids and liver microsomes in CCl4- and BrCCl3-poisoned rats. Biochim. Biophys. Acta 712, 628-638. (29) Uchida, K., and Stadtman, E. R. (1993) Covalent attachment of 4-hydroxynonenal to glyceraldehyde-3-phosphate dehydrogenase: a possible involvement of intra- and intermolecular crosslinking reaction. J. Biol. Chem. 268, 6388-6393. (30) Reed, D. J., Babson, J. R., Beatty, P. W., Brodie, A. E., Ellis, W. W., and Potter, D. W. (1980) High-performance liquid chromatography analysis of nanomole levels of glutathione, glutathione disulfide, and related thiols and disulfides. Anal. Biochem. 106, 55-62. (31) Kroll, D. J., and Rowe, T. C. (1991) Phosphorylation of DNA topoisomerase II in a human tumor cell line. J. Biol. Chem. 266, 7957-7961. (32) Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. (33) Hartley, D. P., Ruth, J. A., and Petersen, D. R. (1995) The hepatocellular metabolism of 4-hydroxynonenal by alcohol dehydrogenase, aldehyde dehydrogenase and glutathione S-transferase. Arch. Biochem. Biophys. 316, 197-205. (34) Chen, Q., Esterbauer, H., and Jurgens, G. (1992) Studies on epitopes on low-density lipoprotein modified by 4-hydroxynonenal: Biochemical characterization and determination. Biochem. J. 288, 249-254. (35) Quinn, M. T., Linner, J. G., Siemsen, D., Dratz, E. A., Buescher, E. S., and Jesaitis, A. J. (1995) Immunocytochemical detection of lipid peroxidation in phagosomes of human neutrophils: correlation with expression flavocytochrome b. J. Leukocyte Biol. 57, 415-421. (36) Niemela, O., Parkkila, S., Yla-Herttuala, S., Halsted, S., Witztum, C., Lanca, A., and Isreal, Y. (1994) Covalent protein adducts in

Chem. Res. Toxicol., Vol. 10, No. 8, 1997 905

(37)

(38)

(39)

(40)

(41)

(42)

(43)

liver as a result of ethanol metabolism and lipid peroxidation. Lab. Invest. 70, 537-549. Rush, G. F., Gorski, J. R., Ripple, M. G., Sowinski, J., Bugelski, P., and Hewitt, W. R. (1985) Organic hydroperoxide-induced lipid peroxidation and cell death in isolated hepatocytes. Toxicol. Appl. Pharmacol. 78, 473-483. Danni, O., Chiarpotto, E., Aragno, M., Biasi, F., Comogliom, A., Belliardo, F., Dianzani, M. U., and Poli, G. (1991) Lipid peroxidation and irreversible cell damage: synergism between carbon tetrachloride and 1,2-dibromoethane in isolated rat hepatocytes. Toxicol. Appl. Pharmacol. 110, 216-222. Nadkarni, D. V., and Sayre, L. M. (1995) Structural definition of early lysine and histidine adduction chemistry of 4-hydroxynonenal. Chem. Res. Toxicol. 8, 284-291. Sayre, L. M., Sha, W., Xu, G., Kaur, K., Nadkarni, D., Subbanagounder, G., and Salomon, R. G. (1996) Immunochemical evidence supporting 2-pentylpyrrole formation on proteins exposed to 4-hydroxynonenal. Chem. Res. Toxicol. 9, 1194-1201. Smith, M. T., Thor, H., Hartzell, P., and Orrenius, S. (1982) The measurement of lipid peroxidation in isolated hepatocytes. Biochem. Pharmacol. 31, 19-26. Keffalas, V., and Stacey, N. H. (1991) Potentiating effects of chlorinated hydrocarbons on carbon tetrachloride toxicity in isolated rat hepatocytes and plasma membranes. Toxicol. Appl. Pharmacol. 109, 171-179. Fulceri, R., Pompella, A., Benedetti, A., and Comporti, M. (1990) On the role of lipid peroxidation and protein-bound aldehydes in the haloalkane-induced inactivation of microsomal glucose 6 phosphatase. Res. Commun. Chem. Pathol. Pharmacol. 68, 73-88.

TX960181B