15-Deoxy Prostaglandin J2 Enhances Allyl Alcohol–Induced Toxicity in Rat Hepatocytes

Toxicological Sciences, Feb 2004

Allyl alcohol causes hepatotoxicity that is potentiated by small doses of bacterial lipopolysaccharide (LPS) through a cyclooxygenase-2 (COX-2)-dependent mechanism. The COX-2 product prostaglandin D2 (PGD2) increases hepatocyte killing by allyl alcohol in vitro. In the present study the ability of the nonenzymatic product of PGD2, 15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2), to increase the cytotoxicity of allyl alcohol was evaluated. In a concentration-dependent manner, 15d-PGJ2 significantly augmented cell death caused by allyl alcohol in isolated rat hepatocytes. 15d-PGJ2 also increased the cytotoxicity of acrolein, the active metabolite of allyl alcohol. An agonist for the PGD2 receptor neither reproduced the increase in allyl alcohol-mediated cytotoxicity nor altered the response to 15d-PGJ2. Similarly, these responses were not affected by either an agonist or an antagonist for the peroxisome proliferator-activated receptor-γ. The enhancement by 15d-PGJ2 of allyl alcohol-mediated cell killing was unaffected by augmentation or inhibition of cAMP. Protein synthesis was markedly decreased by 15d-PGJ2, but inhibition of protein synthesis alone with cycloheximide did not increase allyl alcohol-mediated cell killing. Allyl alcohol at subtoxic concentrations increased translocation of nuclear factor kappa B (NF-κB), whereas at cytotoxic concentrations no translocation occurred. 15d-PGJ2 inhibited translocation of NF-κB from the cytosol to the nucleus both in the presence and absence of allyl alcohol. Like 15d-PGJ2, MG132, an inhibitor of NF-κB activation, enhanced allyl alcohol-induced hepatocyte death. Together these results indicate that 15d-PGJ2 augments hepatocyte killing by allyl alcohol, and the mechanism may be related to the inhibition of NF-κB activation.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://toxsci.oxfordjournals.org/content/77/2/290.full.pdf

15-Deoxy Prostaglandin J2 Enhances Allyl Alcohol–Induced Toxicity in Rat Hepatocytes

Jane F. Maddox maddox@msu.edu 0 1 Alison C. Domzalski 0 1 Robert A. Roth 0 1 Patricia E. Ganey 0 1 0 cology and Toxicology, 215 Food Safety and Toxicology Building, Michigan State University , East Lansing, MI 48824. Fax: (517) 432-2310 1 Department of Pharmacology and Toxicology, Institute for Environmental Toxicology, Michigan State University , East Lansing, Michigan 48824 Allyl alcohol causes hepatotoxicity that is potentiated by small doses of bacterial lipopolysaccharide (LPS) through a cyclooxygenase-2 (COX-2)-dependent mechanism. The COX-2 product prostaglandin D2 (PGD2) increases hepatocyte killing by allyl alcohol in vitro. In the present study the ability of the nonenzymatic product of PGD2, 15-deoxy- 12,14-prostaglandin J2 (15dPGJ2), to increase the cytotoxicity of allyl alcohol was evaluated. In a concentration-dependent manner, 15d-PGJ2 significantly augmented cell death caused by allyl alcohol in isolated rat hepatocytes. 15d-PGJ2 also increased the cytotoxicity of acrolein, the active metabolite of allyl alcohol. An agonist for the PGD2 receptor neither reproduced the increase in allyl alcohol-mediated cytotoxicity nor altered the response to 15d-PGJ2. Similarly, these responses were not affected by either an agonist or an antagonist for the peroxisome proliferator-activated receptor- . The enhancement by 15d-PGJ2 of allyl alcohol-mediated cell killing was unaffected by augmentation or inhibition of cAMP. Protein synthesis was markedly decreased by 15d-PGJ2, but inhibition of protein synthesis alone with cycloheximide did not increase allyl alcoholmediated cell killing. Allyl alcohol at subtoxic concentrations increased translocation of nuclear factor kappa B (NF- B), whereas at cytotoxic concentrations no translocation occurred. 15d-PGJ2 inhibited translocation of NF- B from the cytosol to the nucleus both in the presence and absence of allyl alcohol. Like 15d-PGJ2, MG132, an inhibitor of NF- B activation, enhanced allyl alcoholinduced hepatocyte death. Together these results indicate that 15d-PGJ2 augments hepatocyte killing by allyl alcohol, and the mechanism may be related to the inhibition of NF- B activation. - Allyl alcohol is a chemical used in the manufacture of food flavorings and other agents. Administration of allyl alcohol to animals causes hepatotoxicity, affecting primarily periportal regions of liver lobules. Allyl alcohol must be bioactivated to acrolein by alcohol dehydrogenase to cause hepatocellular injury (Jaeschke et al., 1987; Rikans and Moore, 1987; Serafini-Cessi, 1972), but subsequent events leading to cell death are poorly understood. We recently demonstrated that activation of protein kinase C (PKC ) by allyl alcohol is critical to its cytotoxicity to hepatocytes in vitro (Maddox et al., 2003). Sensitivity of liver to allyl alcohol is increased by inflammation, such that hepatic damage is observed at smaller doses. Indeed, in animals treated with a small dose of bacterial lipopolysaccharide (LPS) that alone did not cause liver toxicity but did initiate an inflammatory response, a normally nontoxic dose of allyl alcohol produced overt hepatotoxicity (Sneed et al., 1997). Several inflammatory mediators have been identified that contribute to the LPS-mediated enhancement of allyl alcohol-induced toxicity. Inhibition of Kupffer cell function with gadolinium chloride abolished injury, suggesting that these cells are critical to the response (Sneed et al., 1997). Neutrophils (PMNs) are also important to injury from LPS/ allyl alcohol, as evidenced by the protective effect of PMN depletion (Kinser et al., 2004). LPS-stimulated Kupffer cells and PMNs release lipid mediators such as prostaglandins (PGs) and thromboxane, the synthesis of which is catalyzed by the enzyme cyclooxygenase-2 (COX-2). COX-2 mRNA in liver tissue and the plasma concentration of PGD2 were increased in rats treated with small doses of LPS and allyl alcohol. In addition, a selective inhibitor of COX-2 protected rats from hepatotoxicity induced by cotreatment with LPS and allyl alcohol (Ganey et al., 2001). These results suggest an essential role for COX-2 products in LPS-mediated enhancement of allyl alcohol liver injury. PGD2, but not PGE2, increased the cytotoxicity of allyl alcohol in isolated hepatocytes (Ganey et al., 2001), raising the possibility that the former mediator contributes to injury in vivo. PGD2 spontaneously dehydrates in aqueous solutions to form PGJ2, which is further metabolized sequentially to 12PGJ2 and 15-deoxy- 12,14-prostaglandin J2 (15d-PGJ2) (Fitzpatrick and Wynalda, 1983). Although PGD2 affects cellular function by binding to plasma membrane receptors, the DP receptor and CRTH2 (chemoattractant receptor-homologous molecule expressed on T-helper type 2 cells) (Hirai et al., 2001), no plasma membrane receptors have been identified for PGs of the J2 series. PGJ2 and its metabolites are endogenous ligands for the nuclear peroxisome proliferator-activated receptor- (PPAR ) (Forman et al., 1995), and binding to this receptor effects functional changes in cells. Of these PGs, 15d-PGJ2 has the greatest affinity for the receptor. Not all actions of 15d-PGJ2 are mediated through PPAR , however, as evidenced by the observations that other PPAR agonists do not mimic all of the effects of 15d-PGJ2 and that 15d-PGJ2 produces some of the same effects in cells with and without PPAR (Castrillo et al., 2000; Guyton et al., 2001; Hortelano et al., 2000; Straus et al., 2000). 15d-PGJ2 inhibits the activation of the transcription factor, NF- B (Castrillo et al., 2000; Guyton et al., 2001; Hortelano et al., 2000; Straus et al., 2000), and this inhibition may contribute to some of the observed actions of 15d-PGJ2. The purpose of these studies was to examine further the enhancement of allyl alcohol-induced cytotoxicity by PGD2. Specifically, the role of the PGD2 metabolite 15d-PGJ2 was evaluated, and studies designed to investigate its mechanism of action were undertaken. MATERIALS AND Materials. BW 245C (5-[6-carboxyhexyl]-1-[3-cyclohexyl-3-hydroxypropyl] hydantoin), PGD2, 15d-PGJ2, and PGJ2 were purchased from Cayman Chemical (Ann Arbor, MI). Cycloheximide, ciglitazone, GW9662, IBMX (3-isobutyl-1-methylxanthine); MG132 (carbobenzoxyl-leucinyl-leucinylleucinal), and Rp-cAMPs (Rp-adenosine-3,5-cyclic mono-phosphorothioate triethylamine salt) were obtained from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA). 3H-Leucine and 32P-ATP were purchased from NEN Perkin Elmer Life Sciences, Inc. (Boston, MA). Allyl alcohol, acrolein, and all other chemicals were obtained from Sigma-Aldrich Chemicals (St. Louis, MO). The vehicle for allyl alcohol and acrolein was Williams Medium E with 0.5 mg/ml gentamicin added. The vehicle for Rp-cAMPs was double-distilled H2O. The vehicle for the prostaglandins, ciglitazone, BW245C, MG132, and IBMX was DMSO. Final concentration of DMSO in a given experiment in all wells was equal and ranged from 0.03 to 0.05%, depending on the experiment. Animals and hepatocyte isolation. Male, Sprague-Dawley rats (Crl:CD (SD)IGS BR; Charles River, Portage, MI), weighing 125175 g were used for hepatic parenchymal cell isolation. Rats received humane care according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication 86-23, revised 1985). They were housed under conditions of controlled temperature, humidity, and light (12 h dark:12 h light) and were allowed food (Harlan Teklad Rodent Diet 8640) and water ad libitum. Hepatocytes were isolated from the livers of rats by collagenase perfusion using a modification of the method of Seglen (1973) as described previously (Ganey et al., 1994). Hepatocyte preparations were 85% viable as determined by exclusion of trypan blue. Cells were placed in Williams medium E supplemented with 10% fetal calf serum and 0.5 mg/ml gentamicin and plated either in 96-well tissue culture plates at a density of 1.2 104 cells per well, in 24-well tissue culture plates at a density of 1.0 105 cells per well, or in 100-mm tissue culture dishes at a density of 2.5 106 cells per dish. After a 2- to 3-h adherence period at 37C, the medium was removed and replaced with Williams medium E supplemented only with 0.5 mg/ml gentamicin. Assessment of cytotoxicity and effects of pharmacologic agents. Cells were exposed to various compounds for 0 3 h at 37C (see figure legends for details) before addition of allyl alcohol (0 100 M) or acrolein (0 125 M). They were then incubated for 90 min at 37C. Hepatocellular injury was assessed by measuring the release of alanine aminotransferase (ALT) into the medium. Release of ALT correlates with other measures of cytotoxicity in hepatocytes, such as uptake of trypan blue (Ganey et al., 1994). After 90 min exposure of cells to allyl alcohol or acrolein, 50 l of the medium was transferred to a clean 96-well plate. The remaining medium was aspirated, and the cells were lysed with a 1.0% solution of Triton X-100. The activity of ALT was determined in both the medium and lysates using Sigma Diagnostics kit no. 59. The total cellular ALT activity was calculated from the activity in the medium plus the activity in the lysate, and hepatocellular injury is presented as the percentage of total cellular ALT released into the medium. Measurement of cAMP. Hepatocytes were plated in 10-cm tissue culture dishes and exposed to 50 M IBMX for 2 h at 37C. Supernatants were then aspirated, and 2.5 ml of 0.1 M HCl was added to the cells remaining on the plate. Cells were incubated for 15 min at room temperature, scraped off the plates, and triturated with a pipettor until homogenized. The homogenate was spun at 1000 g for 10 min, and the supernatant was decanted to a clean tube and stored at 80C until assayed. cAMP activity was measured in acetylated samples by commercial EIA (Cayman Chemical) according to manufacturers instructions. Protein synthesis. Hepatocytes were plated in 24-well plates and allowed to adhere for 23 h in Williams medium E containing 10% FBS. The medium was replaced with leucine-free RPMI supplemented with 3H-leucine (1 mCi/ ml) and 0.5 mg/ml gentamicin. Prostaglandins (10 50 M) or cycloheximide (7 mM) were added 15 min later, and cells were incubated at 37C for 2 h. Cells were then treated with allyl alcohol (25100 M) and further incubated for 1.5 h. After this treatment, medium was removed, and ALT activity in the medium was measured. Adherent hepatocytes were rinsed once with ice-cold, phosphate-buffered saline (PBS) and lysed with 0.1% Triton X-100. Proteins were precipitated from the lysate with an equal volume of ice-cold 10% trichloracetic acid. The precipitate was washed three times with 10% trichloracetic acid and solubilized with 1 ml 1.0 N NaOH overnight at room temperature. Scintillation fluid was added, and 3H levels were quantified [counts per minute (cpm)] using a liquid scintillation counter. NF- B electrophoretic mobility shift assay (EMSA). After treatment hepatocytes were washed with ice-cold PBS, dislodged from plates with cell scrapers, and resuspended (1 106 cells) in 400 l hypotonic lysis buffer A (10 mM HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1.0 mM dithiothreitol (DTT), and 0.5 mM PMSF). Cells were allowed to incubate on ice in buffer A for 15 min before solubilization in 25 l 10% Igepal (Sigma Chemical Co., St. Louis, MO) and vortexing. Nuclei were collected by centrifugation at 15,000 g for 30 s. The nuclear pellet was resuspended in 50 l high-salt extraction buffer B (20 mM HEPES, 0.4 M NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 1.0 mM DTT, and 0.5 mM PMSF) and mixed for 15 min at 4C, and the nuclear extract was obtained after removing the debris by centrifugation for 5 min at 15,000 g. Nuclear protein concentration was quantified using the Bradford assay (Bio-Rad Laboratories, Hercules, CA). For the EMSA, double-stranded NF- B oligonucleotide probes were endlabeled with 32P-ATP using T4 polynucleotide kinase (Promega, Madison, WI) and separated from the unincorporated label by minicolumn chromatography (QIAquick, QIAGEN, Inc., Valencia, CA). Nuclear protein (10 g) was incubated for 30 min on ice with radiolabeled oligonucleotide probes (2 6 104 cpm) in 20 l reaction buffer containing 1 g poly(dI-dC), 10 mM TrisHCl, pH 7.8, 40 mM KCl, 1 mM EDTA, 1 mM DTT and 10% glycerol. Nucleoprotein-oligonucleotide complexes were resolved by electrophoresis in a 4% nondenaturing polyacrylamide gel. After electrophoresis, gels were dried on 3MM paper in a vacuum dryer and exposed to radiographic film. The oligonucleotide sequence used for the detection of binding activity to the NF- B site was 5-AGTTGAGGGGACTTTCCCAGG-3 (Promega). Cold competition assays were conducted by adding a 200-fold molar excess of unlabeled dsDNA oligonucleotide simultaneously with the labeled probe. For supershift assays, p50 antibody (H-119; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and/or p65 antibody (A; Santa Cruz Biotechnology, Inc.) was added FIG. 1. Metabolites of PGD2 increase the sensitivity of isolated hepatocytes to allyl alcohol. Hepatocytes were plated in 96-well tissue culture plates and exposed to (A) 15 M 15dPGJ2 or vehicle or (B) 15 M PGJ2 or vehicle for 2 h. Allyl alcohol (0 100 M) was then added, and the cells were incubated for an additional 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 35. aSignificantly different from vehicle at the same concentration of allyl alcohol. bSignificantly different from the same treatment at 0 M allyl alcohol. to select nuclear protein reaction mixtures before the addition of radiolabeled probe. Statistical analysis. Results are presented as the mean SEM. Data presented as percentages were first transformed (arcsin square root transformation) before analysis by two-way repeated measures ANOVA. Post hoc comparisons were made using Tukeys test. The criterion for significance was p 0.05. Allyl alcohol caused a concentration-dependent cell killing of hepatocytes as assessed from increased ALT release (Fig. 1). 15d-PGJ2 (15 M) augmented hepatocyte killing by allyl alcohol significantly (Fig. 1A). In the presence of 25 M allyl alcohol, 15d-PGJ2 exposure caused a two-fold increase in ALT release, and at 50 M allyl alcohol, 15d-PGJ2 increased cell killing by almost 70%. In previous studies, PGD2, a precursor of 15d-PGJ2, also increased sensitivity to allyl alcohol cytotoxicity, but the effect was smaller in magnitude than that observed with 15d-PGJ2 (Ganey et al., 2001). It was of interest to examine the activity of the more immediate precursor, PGJ2. PGJ2 (15 M) possessed similar activity in enhancing hepatocyte killing by allyl alcohol as 15d-PGJ2 (Fig. 1B). The ability of 15d-PGJ2 to increase cytotoxicity of allyl alcohol was concentration-related (Fig. 2). Using 50 M allyl alcohol, 10 M 15d-PGJ2 did not significantly increase ALT release. All larger concentrations of 15d-PGJ2 examined enhanced hepatocyte killing. Allyl alcohol requires bioactivation to acrolein to produce hepatocellular toxicity (Jaeschke et al., 1987; Rikans and Moore, 1987; Serafini-Cessi, 1972). To investigate whether 15d-PGJ2 increased cytotoxicity independent of effects on allyl alcohol bioactivation, experiments were performed with acrolein. Acrolein caused a concentration-dependent cell killing of hepatocytes (Fig. 3). 15d-PGJ2 (15 M) augmented ALT release caused by acrolein. Studies were undertaken to evaluate whether 15d-PGJ2 was acting through a receptor-mediated mechanism to increase the sensitivity of hepatocytes to allyl alcohol. To examine whether 15d-PGJ2 could be acting via the PGD2 receptor (DP receptor), BW245C, a DP receptor agonist (Harris and Phipps, 2002), was used. The reported IC50 of BW245C for the DP receptor in rat platelets is 250 nM (Town et al., 1983). In the present study, BW245C was used at 100 nM, 1 M, 10 M (data not shown), and 30 M (Fig. 4A). This highest concentration of BW245C produced maximal stimulation of cAMP in bovine tracheal cells (Crider et al., 1999). Both the ability of BW245C to increase the toxicity of allyl alcohol (DP receptor activation) and the possibility that BW245C could affect the response to 15d-PGJ2 (interference through occupation of the receptor) were examined. BW245C had no effect on allyl alcohol-induced killing of hepatocytes (Fig. 4A) at any of the concentrations used. In addition, BW245C did not affect 15d-PGJ2mediated enhancement of allyl alcohol cytotoxicity. Because 15d-PGJ2 has been identified as a ligand for PPAR FIG. 2. 15d-PGJ2 increases the cytotoxicity of allyl alcohol in a concentration-dependent manner. Hepatocytes were plated in 96-well tissue culture plates and exposed to 0 50 M 15dPGJ2 for 2 h, then to 50 M allyl alcohol for an additional 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 3. aSignificantly different from 0 M 15d-PGJ2. FIG. 3. 15d-PGJ2 increases the sensitivity of isolated hepatocytes to acrolein. Hepatocytes were plated in 96-well tissue culture plates and exposed to 15 M 15dPGJ2 for 2 h, then to acrolein at the indicated concentrations for an additional 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 3. aSignificantly different from vehicle at the same concentration of acrolein. bSignificantly different from the same treatment at 0 M acrolein. (Galli et al., 2000), the effects of a PPAR agonist (Fig. 4B) and antagonist (Fig. 4C) were examined. The concentrations of ciglitazone, PPAR agonist, used were based on a report demonstrating inhibition of platelet-derived growth factor-induced stimulation of proliferation of hepatic stellate cells (Galli et al., 2000). At concentrations of 1, 10 (data not shown), and 20 M (Fig. 4B), ciglitazone did not affect hepatocyte killing by allyl alcohol and did not affect 15d-PGJ2-mediated augmentation of killing. At larger concentrations, ciglitazone was cytotoxic without the addition of allyl alcohol (data not shown). A PPAR antagonist, GW9662, was also employed. GW9662, which exhibits an IC50 value for PPAR of approximately 7.6 nM in cell-based assays (Leesnitzer et al., 2002), was added to cells at 100 nM (not shown) and 1 M (Fig. 4C). The receptor antagonist had no effect on allyl alcohol-mediated cytotoxicity or on the 15d-PGJ2-mediated increase in cytotoxicity at either concentration. 15d-PGJ2 has been reported to induce functional changes in macrophages via activating cAMP (Vaidya et al., 1999); therefore, the effect of manipulation of cAMP on the response to 15d-PGJ2 was examined. IBMX is a phosphodiesterase inhibitor that increases cAMP. Incubation with IBMX increased cAMP activity by approximately 80% (data not shown), but did not affect the 15d-PGJ2-mediated enhancement of allyl alcohol cytotoxicity (Fig. 5A). Similarly, inhibition of cAMP activity with an inactive analog, Rp-cAMPs, did not affect 15d-PGJ2 enhancement of allyl alcohol-induced cell killing (Fig. 5B). It has been reported that metabolites of PGD2 inhibit protein synthesis in hepatic stellate cells (Miyahara et al., 2000); FIG. 4. DP and PPAR receptor agonists and a PPAR receptor antagonist do not increase sensitivity of hepatocytes to allyl alcohol and do not affect the response to 15d-PGJ2. Hepatocytes were plated in 96well tissue culture plates. (A) Cells were exposed to vehicle, 30 M 15dPGJ2 or 30 M BW245C, alone or in combination, for 2 h. Cells were then exposed to allyl alcohol at the indicated concentrations for 90 min. (B) Hepatocytes were exposed to vehicle, 30 M 15dPGJ2 or 10 M ciglitazone, alone or in combination, for 2 h, and then to allyl alcohol at the indicated concentrations for 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 3. aSignificantly different from vehicle at the same concentration of allyl alcohol. bSignificantly different from the same treatment at 0 M allyl alcohol. cEach point is significantly different from same treatment at 0 M allyl alcohol. (C) Cells were exposed to 1 M GW9662 for 3 h. Cells were then exposed to 15d-PGJ2 for 2 h, and finally to allyl alcohol at the indicated concentrations for 90 min. Shown is one experiment representative of three. FIG. 5. Manipulation of cellular cAMP does not affect the 15d-PGJ2mediated increase in sensitivity to allyl alcohol. Hepatocytes were plated in 96-well tissue culture plates. (A) Cells were exposed to vehicle, 15 M 15d-PGJ2, or 50 M IBMX, alone or in combination, for 2 h. Cells were then exposed to allyl alcohol at the indicated concentrations for an additional 90 min. (B) Cells were exposed to vehicle, 15 M 15d-PGJ2, or 5 M Rp-cAMPs, alone or in combination, for 2 h. Cells were then exposed to allyl alcohol at the indicated concentrations for an additional 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 4 5. aSignificantly different from vehicle at the same concentration of allyl alcohol. bSignificantly different from the same treatment at 0 M allyl alcohol. therefore, the possibility that inhibition of protein synthesis by 15d-PGJ2 contributes to increased sensitivity to allyl alcohol in hepatocytes was examined. 15d-PGJ2 caused a concentrationdependent inhibition of protein synthesis in the range of 10 50 M in the absence or presence of allyl alcohol (Fig. 6A). In the absence of 15d-PGJ2, allyl alcohol dose-dependently inhibited protein synthesis as well (Fig. 6A). There did not appear to be a synergistic effect of allyl alcohol and 15d-PGJ2 on protein synthesis. Cycloheximide, at a concentration that maximally inhibited protein synthesis, did not increase allyl alcohol-mediated cell killing (Fig. 6B). Thus, it appears that inhibition of protein synthesis is not sufficient to augment allyl alcoholinduced cell death. Because 15d-PGJ2 has been reported to inhibit NF- B activation (Castrillo et al., 2000; Guyton et al., 2001; Hortelano et al., 2000; Straus et al., 2000), and NF- B is involved in cell survival pathways, the activation status of NF- B was examined in hepatocytes treated with 15d-PGJ2 and allyl alcohol. In vehicle-treated cells, some NF- B was detected in nuclear extracts (Fig. 7A), suggesting that in isolated hepatocytes some translocation of the transcription factor occurs. Treatment of the cells for 1 h with 15d-PGJ2 (15 M) decreased the amount of nuclear NF- B detected. Concentrations of allyl alcohol (25 and 50 M) that alone did not cause cytotoxicity (see ALT values in Fig. 7A) increased NF- B translocation to the nucleus. Addition of 15d-PGJ2 inhibited the translocation of NF- B in hepatocytes treated with allyl alcohol at these concentrations and increased release of ALT. A toxic concentration of allyl alcohol (100 M) by itself inhibited NF- B translocation, an effect that was augmented by 15d-PGJ2. Supershift assays with antibodies to the p50 and/or p65 components of NF- B were performed. The results indicated that p65, and not p50, is the effected subunit (Fig. 7B). To investigate FIG. 6. Protein synthesis is inhibited by 15d-PGJ2 and by allyl alcohol. (A) Hepatocytes in 3H-leucine-labeled medium were incubated with vehicle, 15d-PGJ2, or 7 M cycloheximide for 2 h. The cells were then exposed to allyl alcohol at the indicated concentrations for 90 min. Medium was removed, cells were lysed, protein precipitated, and radioactivity in the cell protein precipitates was quantified. (B) ALT activity in the medium of cells treated with allyl alcohol and vehicle or cycloheximide was determined. n 4 5 (A) and 3 (B). aSignificantly different from vehicle at the same concentration of allyl alcohol. bSignificantly different from the same treatment at 0 M allyl alcohol. FIG. 7. 15d-PGJ2 inhibits activation of NF- B. Hepatocytes were plated in 100-mm tissue culture dishes and (A) exposed to vehicle or 25100 M allyl alcohol in the absence or presence of 15 M 15dPGJ2 for 1 h. NF B in the nuclear protein fraction was detected as described in Materials and Methods. Representative of an experiment performed three times. Values under each lane represent the percentage of total ALT activity that was released into the medium for each sample. (B) Cells were exposed to vehicle (Williams medium E) for 1 h. Nuclear protein reactions were incubated with: (A) no antibody; (B) NF- B p50 antibody; (C) NF- B p65 antibody; (D) NF- B p50 and p65 antibodies. further the role of inhibition of NF- B in enhancement of allyl alcohol-mediated cell death, hepatocytes were treated with MG132, a proteasome inhibitor that prevents the translocation of NF- B to the nucleus. MG132 enhanced allyl alcoholinduced hepatocellular death (Fig. 8A). Similarly, MG132 augmented cell killing by acrolein (Fig. 8B). LPS enhances allyl alcohol-induced hepatotoxicity in rats through a mechanism dependent on COX-2 (Ganey et al., 2001). Although the specific products of COX-2 that contribute to liver injury in vivo have not been identified, the plasma concentration of PGD2 increased within 2 h after treatment with LPS plus allyl alcohol (Ganey et al., 2001). Moreover, PGD2, but not PGE2, increased the sensitivity of hepatocytes to allyl alcohol. This study was undertaken to examine further the enhancement of allyl alcohol-induced cytotoxicity by PGD2. Both PGJ2 and 15d-PGJ2, which are metabolites of PGD2, increased the sensitivity of hepatocytes to allyl alcohol (Fig. 1). This effect was concentration-dependent and occurred at concentrations of 15d-PGJ2 greater than 10 M (Fig. 2). The concentration of PGD2 in peripheral plasma 2 h after treatment with allyl alcohol in LPS-treated rats is about 0.5 M (Ganey et al., 2001); however, the concentration in the microenvironment of the liver, near the source of production, has been estimated to be closer to 10 M (Neuschafer-Rube et al., 1993), suggesting that the effect of 15d-PGJ2 to increase the cytotoxicity of allyl alcohol occurs at relevant concentrations. Allyl alcohol-induced toxicity only occurs after bioactivation to acrolein. The sensitivity of hepatocytes to the cytotoxicity of acrolein was also increased by 15d-PGJ2 (Fig. 3), indicating that the effect on cytotoxicity of allyl alcohol cannot be explained simply by increased conversion to acrolein. The effect of 15d-PGJ2 to increase the sensitivity of hepatocytes to allyl alcohol does not appear to be mediated through the DP receptor, because BW245C, over a range of concentrations including one-hundred times greater than the IC50, did not reproduce the increase in cytotoxicity of allyl alcohol and did not interfere with the ability of 15d-PGJ2 to enhance allyl alcohol-mediated cytotoxicity. PPAR also does not appear to be involved in the 15d-PGJ2-mediated increase in allyl alcoholinduced cell death, because neither a PPAR -agonist nor a PPAR antagonist affected the response to allyl alcohol. PPAR -independent effects of 15d-PGJ2 have been reported previously. For example, in macrophages 15d-PGJ2 inhibited FIG. 8. MG132 increases sensitivity of hepatocytes to allyl alcohol and to acrolein. Hepatocytes were plated in 96-well tissue culture plates and exposed to 5 M MG132 or vehicle for 2 h, then to (A) allyl alcohol or (B) acrolein at the indicated concentrations for 90 min. Leakage of ALT was measured as a determinant of cytotoxicity. n 4. aSignificantly different from vehicle at the same concentration of allyl alcohol or acrolein. bSignificantly different from the same treatment at 0 M allyl alcohol or 0 M acrolein. production of tumor necrosis factor- and thromboxane B2 (Guyton et al., 2001) and increased production of superoxide anion (Hortelano et al., 2000) through mechanism(s) independent of PPAR . The mechanism by which PGD2 and its metabolites increase the sensitivity of hepatocytes to allyl alcohol-induced cytotoxicity is unknown. In hepatocytes or isolated livers, PGD2 altered the profile of protein phosphorylation, increased release of intracellular calcium, enhanced production of inositol trisphosphate, stimulated glycogen phosphorylase and glucose release, and inhibited gluconeogenesis (Altin and Bygrave, 1988; Casteleijn et al., 1988a,b), demonstrating that PGD2 has diverse effects on intracellular signaling and glucose metabolism. A well-known action of PGD2 is to increase intracellular cAMP. Modulation of cell cycle by PGD2 or 15d-PGJ2 has been reported to occur by cAMP-dependent (Okuda-Ashitaka et al., 1990) and -independent (Hughes-Fulford and Fukushima, 1993) mechanisms. Furthermore, increasing cAMP through inhibition of phosphodiesterase activity with IBMX enhanced 15d-PGJ2-mediated inhibition of neutrophil adhesion and superoxide anion production (Vaidya et al., 1999). On the other hand, 15d-PGJ2-induced apoptosis of neutrophils was not dependent on cAMP (Ward et al., 2002). In the current study, neither increasing cAMP nor inhibiting generation of cAMP affected the increased sensitivity of hepatocytes to allyl alcohol in response to 15d-PGJ2 (Fig. 5), suggesting that this effect is not dependent on cAMP. PGD2 affects hepatocellular proteins, and in the current study, both 15d-PGJ2 and allyl alcohol caused concentrationdependent inhibition of protein synthesis (Fig. 6). However, when the same degree of inhibition of protein synthesis that occurred with a concentration of 15d-PGJ2 that enhanced allyl alcohol cytotoxicity was reproduced with cycloheximide, the sensitivity of hepatocytes to allyl alcohol was not altered. This observation indicates that inhibition of protein synthesis is not sufficient for the increased sensitivity to allyl alcohol. Another mechanism by which 15d-PGJ2 may affect allyl alcohol-induced cell death is through inhibition of activation of the transcription factor, NF- B. In quiescent cells, NF- B exists in the cytosol in association with an inhibitory protein, I B. Phosphorylation of I B results in its ubiquitination and degradation via a proteasome pathway. This activates NF- B, which is then free to migrate into the nucleus and modulate gene expression. Numerous reports suggest that inhibition of activation of NF- B permits or induces cell death in response to a variety of agents (Botchkina et al., 1999; Hatano and Brenner, 2001; Izban et al., 2001; Ni et al., 2001; Russo et al., 2001). For example, liver injury induced by bile duct ligation was greater in mice in which activation of NF- B was inhibited compared with mice in which NF- B was activated (Miyoshi et al., 2001). In addition, exposure to bile acids that did not activate NF- B caused death in rat hepatoma cells, whereas cells exposed to bile acids that activated NF- B survived (Rust et al., 2000). In particular, 15d-PGJ2 inhibited the activation of NF- B stimulated by LPS or tumor necrosis factor- , and this inhibition was associated with cell death (Ward et al., 2002). Thus, it is clear that inhibition of NF- B activation can be associated with enhanced tissue injury and cell death. In the present study, allyl alcohol caused nuclear translocation of NF- B at noncytotoxic concentrations (Fig. 7), but not at a concentration that resulted in cell death. A similar effect on NF- B activation was observed in acrolein-treated human lung adenocarcinoma cells (Horton et al., 1999). The mechanism by which allyl alcohol activates NF- B was not investigated in the present study; however, allyl alcohol activates PKC in hepatocytes (Maddox et al., 2003), and PKC is critical to NF- B activation in response to a variety of stimuli (Minami et al., 2003; Page et al., 2003; Vancurova et al., 2001). Thus, it is possible that allyl alcohol activates NF- B through a PKC mediated pathway. The reason why NF- B was not activated at larger, cytotoxic concentrations of allyl alcohol is unknown. 15d-PGJ2 inhibited nuclear translocation of NF- B in hepatocytes exposed to allyl alcohol and caused cell death at normally noncytotoxic concentrations of allyl alcohol. Thus, under all conditions in which allyl alcohol induced cell death, activation of NF- B was inhibited. It has been reported that 15d-PGJ2 inhibits NF- B through both PPAR -dependent and -independent mechanisms (Castrillo et al., 2000; Guyton et al., 2001; Hortelano et al., 2000; Straus et al., 2000). One possible mechanism is through inhibition of I B degradation (Mullally et al., 2001; Ward et al., 2002), which precedes activation of NF- B (Chen et al., 1995). Proteasome activity is necessary for the degradation of I B; thus, inhibition of proteasome activity could contribute to the inhibition of NF- B activation. Inhibition of proteasome activity with MG132, which prevents activation of NF- B (Fiedler et al., 1998; Hellerbrand et al., 1998), increased the sensitivity of hepatocytes to allyl alcohol as well as to acrolein (Fig. 8). These results support the hypothesis that inhibition of NF- B by 15d-PGJ2 contributes to the enhancement of allyl alcohol-mediated cytotoxicity. Alternative mechanisms by which MG132 increases the sensitivity to allyl alcohol cannot be ruled out, however. For example, MG132 increased toxicity of 6-hydroxydopamine in neuronal cells through inhibition of protein degradation (Elkon et al., 2001). In summary, the PGD2 metabolites, PGJ2 and 15d-PGJ2, increased the sensitivity of hepatocytes to allyl alcohol-induced cytotoxicity. This effect does not appear to be mediated through the DP receptor or PPAR . Increased cytotoxicity was associated with inhibition of NF- B activation, and another agent that inhibits activation of this transcription factor also increased the cytotoxicity of allyl alcohol. Taken together, these results suggest that inhibition of NF- B activation by 15d-PGJ2 contributes to the increased sensitivity to allyl alcohol. The mechanisms by which inhibition of NF- B leads to enhancement of allyl alcohol-induced cytotoxicity remain to be investigated. ACKNOWLEDGMENTS This work was supported by grants ES08789 and AA014134 from NIH. J.F.M. was supported, in part, by an Arthritis Investigator Award. 15-d-PGJ(2), and peroxisome proliferator activated receptor- agonists induce apoptosis in transformed, but not normal, human T lineage cells. Immunology 105, 2334. ization of prostaglandin-F2 alpha-binding sites on rat hepatocyte plasma membranes. Eur. J. Biochem. 211, 163169.


This is a preview of a remote PDF: https://toxsci.oxfordjournals.org/content/77/2/290.full.pdf

Jane F. Maddox, Alison C. Domzalski, Robert A. Roth, Patricia E. Ganey. 15-Deoxy Prostaglandin J2 Enhances Allyl Alcohol–Induced Toxicity in Rat Hepatocytes, Toxicological Sciences, 2004, 290-298, DOI: 10.1093/toxsci/kfh028