Differential Regulation of Redox Responsive Transcription Factors by

Jun 27, 2001 - the oxidative stress induced by TGHQ in renal proximal tubule epithelial cells (LLC-PK1) modulates ... regulation of these transcriptio...
0 downloads 0 Views 205KB Size
814

Chem. Res. Toxicol. 2001, 14, 814-821

Differential Regulation of Redox Responsive Transcription Factors by the Nephrocarcinogen 2,3,5-Tris(glutathion-S-yl)hydroquinone Thomas J. Weber,† Qihong Huang,‡ Terrence J. Monks,‡ and Serrine S. Lau*,‡ Molecular Biosciences, Pacific Northwest National Laboratory, Richland, Washington 99352, and Center for Molecular and Cellular Toxicology, Division of Pharmacology and Toxicology, College of Pharmacy, The University of Texas at Austin, Austin, Texas 78712-1074 Received August 25, 2000

2,3,5-Tris(glutathion-S-yl)hydroquinone [TGHQ] is a potent nephrotoxicant and nephrocarcinogen, and induces a spectrum of mutations in human and bacterial cells consistent with those attributed to reactive oxygen species (ROS). Studies were conducted to determine whether the oxidative stress induced by TGHQ in renal proximal tubule epithelial cells (LLC-PK1) modulates transcriptional activities widely implicated in transformation responses, namely 12-O-tetradecanoyl phorbol 13-acetate (TPA) responsive element (TRE)- and nuclear factor kappa B (NF-κB)-binding activity. TGHQ increased TRE- and NF-κB-binding activity in a concentration- and time-dependent manner. Catalase fully inhibited peak TGHQ-mediated TRE- and NF-κB-binding activity. In contrast, although deferoxamine fully inhibited TGHQmediated TRE-binding activity, it had only a marginal effect on NF-κB-binding activity. Collectively, these data indicate that TGHQ modulates TRE- and NF-κB-binding activity in an ROS-dependent fashion. Cycloheximide and actinomycin D fully inhibited TGHQ-mediated TRE-binding activity, but in the absence of TGHQ increased NF-κB-binding activity. Although protein kinase C (PKC) is widely implicated in stress response signaling, pretreatment of cells with PKC inhibitors (H-89, calphostin C) did not modulate TGHQ-mediated DNA-binding activities. In contrast, pretreatment of cells with (PD098059), a mitogen activated protein kinase kinase (MEK) inhibitor, markedly reduced TGHQ-mediated TRE-binding activity, but enhanced TGHQ-mediated NF-κB-binding activity. We conclude that TGHQ-mediated TRE- and NFκB-binding activities are ROS-dependent. Although there is a common requirement for hydrogen peroxide (H2O2) in the regulation of these DNA-binding activities, there appears to be divergent regulation after H2O2 generation in renal epithelial cells.

Introduction The cellular response to oxygen free radicals is largely influenced by the type and concentration of oxygen radical generated. A single exposure to xanthine/xanthine oxidase predominantly increases smooth muscle cell proliferation, whereas frequent exposures to high levels of xanthine/xanthine oxidase result in cell death (1). Cotreatment studies with superoxide dismutase (SOD)1 and catalase suggest that xanthine/xanthine oxidasedependent superoxide anion (O2•-) is mitogenic, while hydrogen peroxide (H2O2) is cytotoxic (1). In the presence of transition metal ions, and iron in particular, H2O2 is converted to the extremely reactive hydroxyl radical (•OH) which is thought to be the primary toxic species responsible for DNA damage and cell death (2-4). Within this context, we have been investigating the nephrotoxic * To whom correspondence should be addressed. Phone: (512) 4715190. Fax: (512) 471-5002. E-mail: [email protected]. † Molecular Biosciences. ‡ Center for Molecular and Cellular Toxicology. 1 Abbreviations: AP-1, activator protein-1; EMSA, electrophoretic mobility shift assay; ERK, extracellular signal regulated kinase; FBS, fetal bovine serum; H2O2, hydrogen peroxide; •OH, hydroxyl radical; HQ, hydroquinone; MAPK, mitogen activated protein kinase; NF-κB, nuclear factor kappa B; PKC, protein kinase C; ROS, reactive oxygen species; O2•-, superoxide anion; SOD, superoxide dismutase; TPA, 12O-tetradecanoyl phorbol-13-acetate; TRE, TPA responsive element; TGHQ, 2,3,5-tris(glutathion-S-yl)hydroquinone.

and nephrocarcinogenic actions of glutathione (GSH) conjugates (quinol-thioethers) of hydroquinone [HQ (5)]. Quinol-thioether-mediated cytotoxicity is markedly reduced by catalase and deferoxamine, scavengers of H2O2 and Fe3+/Fe2+, respectively, suggesting that the cytotoxic response is mediated by oxygen free radicals and in particular the •OH (4, 6). Reactive oxygen species (ROS) are consistently associated with the regulation of the activator protein-1 (AP1) and nuclear factor kappa B (NF-κB) transcription factors (7-9). AP-1 and NF-κB are implicated as causal in transformation responses (10, 11) leading a number of investigators to speculate a role for ROS-dependent regulation of these transcription factors in carcinogenesis. AP-1 is a heterodimeric complex composed of c-jun (cJun, JunB, and JunD) and c-fos (c-Fos, Fos B, and Fra1) protooncogene family members, as either a Jun:Jun homodimer or Jun:Fos heterodimer that specifically binds to the 12-O-tetradecanoyl phorbol-13-acetate (TPA) responsive element [TRE (12)]. NF-κB DNA-binding activity is associated with at least five different NF-κB family members: NF-κB1 (p105/p50), NF-κB2 (p100/p52), RelA (p65), RelB, and c-Rel (13). The most common NF-κB dimers consist of RelA (p65) and NF-κB1 (p50) or NFκB2 (p52) subunits (14). H2O2, but not O2•-, is thought to regulate NF-κB DNA-binding activity (9). In contrast,

10.1021/tx000190r CCC: $20.00 © 2001 American Chemical Society Published on Web 06/27/2001

Stress-Responsive Signaling

the AP-1-dependent induction of gadd153 following H2O2 treatment is inhibited by o-phenanthroline and mannitol, raising the possibility that iron-generated •OH regulates AP-1-related signal transduction (8). In addition to ROS, a number of growth- and stress-related signal transduction pathways, including the protein kinase C (PKC) and mitogen activated protein kinase (MAPK) pathways, regulate AP-1 and NF-κB activity (15-17). The MAPK cascade is firmly established in the transformation response to chemical and polypeptide tumor promoters (11, 18-20), and is typically defined on the basis of extracellular signal regulated kinase (ERK1/ERK2) activity (20, 21). PKC represents a group of at least 11 different isoforms that transduce signals from a wide variety of stimuli, and are the receptors for tumor promoting phorbol esters (22). HQ and 2,3,5-tris(glutathion-S-yl)HQ [TGHQ] are nephrocarcinogens (23, 24). In addition, a single treatment of primary renal epithelial cells, isolated from Eker rats, with TGHQ (100 µM, 4 h) resulted in cell transformation as indicated by the acquisition of anchorageindependent growth (25). The spectrum of mutations induced by quinol-thioethers in human and bacterial cells is consistent with those induced by ROS, indicating that ROS play an important role in HQ and TGHQ-mediated mutagenicity and carcinogenicity (26). Moreover, HQmediated nephrocarcinogenicity may involve a cytotoxic mode of action, and it is well established that HQ-GSH conjugates are potent nephrotoxicants (27). Collectively, these observations raise the possibility that HQ-GSH conjugates mediate the nephrocarcinogenic actions of HQ. The present studies were conducted to determine whether TGHQ modulates TRE- and NF-κB-binding activity in renal epithelial cells via the generation of an oxidative stress, and to identify signal transduction pathways that regulate these DNA-binding activities.

Materials and Methods Caution: HQ, TGHQ, and acrylamide are hazardous and should be handled carefully. Chemicals. TGHQ was synthesized as previously described (28) and was >99% pure as determined by HPLC. TRE, NFκB, and AP2 consensus sequences were purchased from Promega (Madison, WI). [γ-32P]-ATP (3000 Ci/mmol) was obtained from New England Nuclear (Beverly, MA). Poly d(I-C) was purchased from Boehringer Mannheim (Indianapolis, IN). PD098059 was from Calbiochem (La Jolla, CA). All other chemicals were from Sigma Chemical (St. Louis, MO). Cell Culture. LLC-PK1 cells were obtained from the American Type Culture Collection (CL101) at passage 181. Cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM; JRH Biosciences; Lenexa, KS) supplemented with 4 g/L D-glucose and 10% fetal bovine serum (FBS; Atlanta Biologicals; Norcross, GA) in 5% CO2:95% air at 37 °C. Cells were subcultured by trypsinization and all experiments were conducted with 5 day postconfluent cultures at passage levels 187-200. Neutral Red Assay. Cell viability was determined using a neutral red assay as previously described (4) with minor modification. Specifically, the volume of extraction solvent used for the 60 mm dish was 4 mL. All 60 mm dishes were placed in a wire basket and processed simultaneously to ensure comparable handling and extraction times. TGHQ Treatment. TGHQ was solubilized in distilled water at a concentration of 10 mg/mL immediately before use and added directly to culture dishes (0.1% FBS DMEM + 20 mM Hepes, pH 7.4; media volume of 3 mL/60 mm dish and 0.5 mL/ well of 24 well plate). A distilled water control was used in all cases. Concentrations are expressed as nmol TGHQ/cm2 surface

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 815 area to present the data in a normalized fashion. For example, in 24 well plates (0.5 mL/well), 300 µM TGHQ treatment is associated with an approximate 50% reduction of cell viability. Under these conditions in a 24 well plate, the volume of media is 0.25 mL/cm2. To maintain the same dose of TGHQ/cm2 in a 60 mm dish, a volume of 6 mL is required which is beyond the practical volume limits of a 60 mm dish. The media volume used in the 60 mm dish was 3 mL (0.13 mL/cm2). Deferoxamine Pretreatment. LLC-PK1 cells were pretreated with 10 mM deferoxamine in 0.1% FBS DMEM for 30 min. Prior to TGHQ treatment, monolayers were washed three times with PBS to remove residual deferoxamine and minimize nonspecific antioxidant effects of preservatives in the chemical stock. Electrophoretic Mobility Shift Assay (EMSA). EMSAs were carried out as described previously (15). The nucleic acid sequence for all the response elements used in this study are available from the manufacturer’s catalog (Promega, Madison, WI). Statistics. Individual comparisons were made using the students t-test or ANOVA and Student’s Newman-Keul post hoc. The p ) 0.05 level was accepted as significant.

Results Comparative Dose Response. To correlate molecular events with cellular response, we investigated TGHQmediated cytotoxicity in 24 well plates and 60 mm dishes over a range of concentrations. In this comparison, we used routine culture volumes of 0.5 mL/well in 24 well plates ()0.25 mL/cm2) and 3 mL/60 mm dish ()0.13 mL/ cm2). To maintain equal media volume/cm2 would require a 6 mL volume in the 60 mm dish which is beyond the practical volume limits of this dish. This is an important consideration since interlaboratory media volumes used for routine cell culture vary. For example, many laboratories use 1 mL media/well in 24 well plates, which would require a 12 mL volume in a 60 mm dish to maintain equal media volume/cm2, a volume that cannot be achieved in standard 60 mm tissue culture dishes. The significance of this observation is illustrated in our comparative dose response. Treatment of LLC-PK1 cells with TGHQ (1001000 µM) for 2 h resulted in a concentration-dependent decrease of cell viability in 24 well plates and 60 mm dishes, as measured using a neutral red assay as described in Materials and Methods (Figure 1). However, the dose of TGHQ required an approximate doubling in the 60 mm dish to achieve a comparable toxic response (Figure 1A). If the same data are expressed as nmol TGHQ/cm2 (i.e., amount of chemical/cell), the large difference in toxicity is no longer apparent, although slight variability exists (Figure 1B). The volume of media/dish (24 well ) 0.25 mL/cm2; 60 mm ) 0.13 mL/cm2) is different and may contribute to this slight variability. Within this context, interexperimental variation can be up to 20%. Therefore, the toxic response to TGHQ is comparable in culture dishes of different size if expressed as nanamoles of TGHQ per squared centimeter, a response that is not apparent when data are expressed as molarity. An application of this model to the correlation of DNA-binding activity with gene expression will be illustrated in the discussion section. TGHQ-induces TRE- and NF-κB-binding activity. Initial studies were conducted to determine the temporal regulation of TRE- and NF-κB-binding activity by TGHQ (62 nmol/cm2) and H2O2. Therefore, we also determined the dose-response for H2O2-mediated cytotoxicity in 60 mm dishes used for the EMSA. LLC-PK1 cells were

816

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Figure 1. TGHQ-mediated cytotoxicity in 24 well plates and 60 mm culture dishes. LLC-PK1 cells were seeded in the respective culture dish and maintained until 5 day postconfluent. Cells were then treated with 100-1000 mM TGHQ in 0.1% FBS DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in Materials and Methods. Panel A shows data expressed as molarity and Panel B shows same data expressed as nmol TGHQ/cm2. Values represent the mean ( SE (n ) 3). (*) Significantly different from control, p ) 0.05. Similar results were observed in two separate experiments.

Weber et al.

Figure 3. Temporal regulation of TRE- and NF-κB-binding activity by TGHQ and H2O2. (A) TRE-binding activity in LLCPK1 cells treated with 62 nmol TGHQ/cm2 (O) and 0.27 mM H2O2 (b) for 1-5 h. (B) NF-κB-binding activity in cells treated with 62 nmol TGHQ/cm2 (circle) and 0.27 mM H2O2 (square) for 1-5 h. Two inducible NF-κB-binding complexes were observed and designated complex 1 (solid symbol) and complex 2 (open symbol). Similar results were observed in two separate experiments.

Figure 2. H2O2-mediated cytotoxicity in 60 mm culture dishes. LLC-PK1 cells were seeded in 60 mm culture dish and maintained until 5 day postconfluent under identical conditions used for EMSA analysis. Cells were then treated with 0.1-0.4 mM H2O2 in 0.1% FBS DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in Materials and Methods. Values represent the mean ( SE (n ) 3). (*) Significantly different from control, p ) 0.05. Similar results were observed in two separate experiments.

Figure 4. Concentration-dependent regulation of TRE- and NF-κB-binding activity by TGHQ. LLC-PK1 cells were treated with 25-126 nmol TGHQ/cm2 for 3 h and processed for measurements of TRE (A)- and NF-κB (B)-binding activity in a standard EMSA as described in Materials and Methods. Similar results were observed in two separate experiments.

treated with 0.1-0.4 mM H2O2 for 2 h and cell viability determined using a neutral red assay as described in the Materials and Methods. H2O2 treatment resulted in a concentration-dependent decrease in cell viability (Figure 2). A moderately toxic concentration of H2O2 (EC50; 0.27 mM) was chosen to determine the effect of H2O2 on TREand NF-κB-binding activity. To compare the temporal regulation of TRE- and NFκB-binding activities by these toxicants, the same nuclear extracts were used in parallel in all of our analyses. Peak TGHQ-mediated TRE- and NF-κB-binding activity was observed between 2 and 4 h (Figure 3, panels A and B).

NF-κB-binding activity consists of two inducible bands termed complex 1 and complex 2 (see figure legend). Complex 1 was the major binding activity induced by TGHQ in all the experiments, while the signal for complex 2 was more variable. Although several NF-κB complexes have been identified (13), it is not known whether the complexes observed in our studies represent unique heterodimers, or degradation products. Subsequently, LLC-PK1 cells were treated with H2O2 to determine whether TRE- and NF-κB-binding activities were directly regulated by ROS. H2O2 (0.27 mM; 1-5 h) increased TRE- and NF-κB-binding activity in a time-

Stress-Responsive Signaling

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 817

Figure 5. Effect of catalase and deferoxamine on peak TGHQ- and TPA-mediated TRE- and NF-κB-binding activity. LLC-PK1 cells were treated with TGHQ (76 nmol/cm2; 3 h) or TPA (10 ng/mL; 1 h) in the presence and absence of catalase (10 U/ml cotreatment) and deferoxamine (10 mM; 30 min pretreatment) and processed for measurements of TRE (A)- and NF-κB (B)-binding activity. Specificity for the binding reaction was confirmed by addition of excess unlabeled target DNA which competitively eliminated the inducible bands, and by addition of excess unlabeled nontarget DNA which was without effect. Similar results were observed in two separate experiments.

dependent fashion (Figure 3, panels A and B). Control TRE- and NF-κB-binding activities in 5 day postconfluent LLC-PK1 cells did not change over time, consistent with our previous reports (15). TGHQ-Induced TRE- and NF-KB-Binding Activities Are ROS-Dependent. A differential dose-response for peak TRE- and NF-κB-binding activity was observed in LLC-PK1 cells treated with 25-126 nmol TGHQ/cm2 (Figure 4). Maximal TRE-binding activity 3 h following chemical treatment was observed between 25 and 76 nmol TGHQ/cm2, whereas maximal NF-κB-binding activity was observed between 76 and 126 nmol TGHQ/cm2. To determine whether TGHQ-mediated TRE- and NFκB-binding activities were ROS dependent, LLC-PK1 cells were treated with TGHQ in the presence or absence of catalase (H2O2 scavenger) and deferoxamine (Fe2+/Fe3+ chelator). Catalase (10 units/mL) as a cotreatment fully inhibited peak TGHQ (76 nmol/cm2)-mediated TRE- and NF-κB-binding activities indicating a requirement for H2O2 in these responses (Figure 5). Pretreatment with the iron chelator deferoxamine (10 mM, 30 min) fully inhibited inducible TRE-binding activity suggesting that this response is dependent upon iron-mediated •OH generation. Deferoxamine had a lesser effect on NF-κBbinding activity, suggesting this response is regulated by both H2O2 and the •OH (Figure 5), consistent with a number of reports indicating that H2O2 is the primary ROS regulating NF-κB (9, 29). In contrast, peak TPA (10 ng/mL, 1 h)-mediated TRE- and NF-κB-binding activities were not inhibited by catalase or deferoxamine (Figure 5), arguing against nonspecific inhibition of inducible DNA-binding activities by either of these treatments. The catalase used for these experiments is from an antioxidant-free stock, and therefore inhibition of TGHQ-

Figure 6. Effect of cycloheximide and actinomycin D on peak TGHQ-mediated TRE (top panel)- and NF-κB (bottom panel)binding activity. LLC-PK1 cells were pretreated for 10 min with 10 µg/mL actinomycin D and 50 µg/mL cycloheximide, subsequently treated with 76 nmol TGHQ/cm2 for 3 h, and processed for measurements of TRE (A)- and NF-κB (B)-binding activity in a standard EMSA as described in Materials and Methods. Similar results were observed in two separate experiments.

mediated DNA-binding activities by catalase cannot be attributed to a nonspecific antioxidant effect. Specificity for the binding reaction was confirmed by addition of unlabeled target DNA, which competitively eliminated the inducible band, and with unlabeled nontarget DNA which was without effect. TGHQ-Induced TRE-Binding Activity Is Transcriptionally Dependent. TRE- and NF-κB-binding activity in the EMSA can represent activation of existing protein or the transcriptional upregulation of AP-1 and NF-κB. To determine the role of transcription and

818

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Figure 7. TGHQ-mediated TRE- and NF-κB-binding activity is not sensitive to PKC inhibitors. LLC-PK1 cells were pretreated for 30 min with 20 µM H-89 and 100 nM calphostin C, subsequently treated with 76 nmol TGHQ/cm2 for 3 h, and processed for measurements of TRE (top panel)- and NF-κB (bottom panel)-binding activity in a standard EMSA as described in Materials and Methods. Similar results were observed in two separate experiments.

translation in this response, LLC-PK1 cells were pretreated with actinomycin D (10 µg/mL) or cycloheximide (50 µg/mL) for 10 min as previously described (30), subsequently treated with 76 nmol TGHQ/cm2 for 3 h, and TRE- and NF-κB-binding activity determined. TGHQmediated TRE-binding activity was fully inhibited by actinomycin D and cycloheximide, but not DMSO, suggesting this response is transcriptionally dependent (Figure 6). The transcriptional-dependence suggests that the TRE-binding response does not represent the direct

Weber et al.

activation of existing AP-1 by the •OH (Figure 5). Actinomycin D and cycloheximide alone increased NFκB-binding activity. Thus, the dependence of the NF-κBbinding response to TGHQ on transcription/translation could not be determined from these studies. The increase of NF-κB-binding activity by actinomycin D and cycloheximide is observed in other epithelial cell types (31). We have previously shown that prostanoid- and phorbol ester-induced TRE-binding activity are markedly reduced by PKC inhibitors (H-89; calphostin C) in LLCPK1 cells (15, 32). LLC-PK1 cells were pretreated with H-89 (20 µM) and calphostin C (100 nM) for 30 min, subsequently treated with 76 nmol of TGHQ/cm2 for 3 h and TRE- and NF-κB-binding activity determined. H-89 and calphostin C pretreatment did not inhibit TGHQmediated TRE- and NF-κB-binding activities (Figure 7), suggesting these responses are not dependent on H-89and calphostin C-sensitive PKC isoforms. Modulation of TGHQ-Induced TRE- and NF-KBBinding Activity by a MEK Inhibitor. TRE- and NFκB-binding activity are also regulated by the MAPK signal transduction pathway, which is typically defined on the basis of extracellular signal regulated kinase (ERK1/ERK2) activity (33-35). PD098059 inhibits MEK activity, an upstream regulator of the ERKs. To determine whether the TRE- and NF-κB-binding response to TGHQ was dependent on the MAPK pathway, LLC-PK1 cells were pretreated for 60 min with 50 µM PD098059 as described (36), subsequently treated with 12-63 nmol TGHQ/cm2 for 3 h, and TRE- and NF-κB-binding activity determined. TGHQ-mediated TRE-binding activity was markedly reduced by PD098059 (Figure 8). In contrast, PD098059 pretreatment did not inhibit NF-κB-binding

Figure 8. Inhibition of TGHQ-mediated TRE-binding activity by PD098059. LLC-PK1 cells were pretreated for 60 min with 50 µM PD098059, subsequently treated with 12-63 nmol TGHQ/cm2 for 3 h, and processed for measurements of TRE-binding activity in a standard EMSA as described in Materials and Methods. Groups shown represent TRE-binding activity with (O) or without (9) PD098059 pretreatment. Similar results were observed in two separate experiments.

Stress-Responsive Signaling

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 819

Figure 9. Increased TGHQ-mediated NF-κB-binding activity by PD098059. LLC-PK1 cells were pretreated for 60 min with 50 µM PD098059, subsequently treated with 12-63 nmol TGHQ/cm2 for 3 h, and processed for measurements of NF-κB-binding activity in a standard EMSA as described in Materials and Methods. Groups shown represent NF-κB-binding activity with (open) or without (solid) PD098059 pretreatment for complex 1 (square) and complex 2 (circle). Similar results were observed in two separate experiments.

activity from the same nuclear extracts used for the TREbinding analysis (Figure 9). Qualitatively, PD098059 pretreatment appeared to enhance TGHQ-mediated NFκB-binding activity, suggesting MEK-related signaling may antagonize this DNA-binding response.

Discussion We have investigated the regulation of redox responsive transcription factors, that are widely implicated in carcinogenesis, by TGHQ a redox active nephrocarcinogenic metabolite of HQ (23-25). TGHQ-mediated TREand NF-κB-binding activities are fully inhibited by catalase (Figure 5) and are therefore ROS-dependent. In previous studies, we have demonstrated that deferoxamine inhibits quinol-thioether-mediated cytotoxicity (6, 37), DNA single strand breaks (37), and gadd153 expression, a molecular marker of genotoxic stress (6). Deferoxamine fully inhibits TGHQ-mediated TRE-binding activity (Figure 5) suggesting this response is •OHdependent. Deferoxamine had a lesser effect on TGHQmediated NF-κB-binding activity (Figure 5), suggesting the NF-κB response is regulated by both H2O2 and the •OH. Because TRE-binding activity is transcriptionally dependent (Figure 6) and the temporal regulation of TREand NF-κB-binding activity are comparable (Figure 3), the •OH-dependent regulation of these DNA-binding activities is likely secondary to the genotoxic stress response. A correlation between increasing H2O2-mediated cytotoxicity and decreasing TRE-binding activity has been observed in cultured neuronal cells (38) and AP-1 DNAbinding activity is inhibited by high ROS concentrations via oxidation of a critical cysteine residue (39). Quinol-

thioethers are redox active and are electrophiles that covalently bind and inactivate cellular macromolecules (40). Thus, the reduction of TGHQ-mediated TRE-binding activity at progressively higher concentrations may be related to deregulated gene expression due to overt cytotoxicity (compare Figures 1 and 4), the generation of high concentrations of ROS, or the direct inhibition of AP-1 via covalent binding of TGHQ-derived reactive electrophilic metabolites. The activation of NF-κB by H2O2 is well established (9, 29) and high ROS concentrations also inhibit NF-κB DNA-binding activity, presumably via oxidation of a sensitive thiol. NF-κB-binding activity can be restored by treatment of nuclear extracts with reducing agents such as 2-mercaptoethanol (41). NF-κB-binding activity was increased in a dose-dependent fashion and therefore, appears to be a direct response to quinone-thioether challenge. However, this observation must be interpreted with caution since reducing agent was present in the extraction buffer. A functional NFκB-luciferase assay is required to better understand the correlation between inducible NF-κB-binding activity and NF-κB-dependent transcriptional activity under these experimental conditons. AP-1 is highly responsive to genotoxic stress (42), and quinol-thioether-mediated DNA damage is observed within 15 min of exposure (4). Therefore, genotoxic stress precedes the TRE-binding response (Figure 3). As noted above, deferoxamine inhibits many of the cellular responses to TGHQ, including the expression of gadd153, a molecular marker of genotoxic stress (43-45). Of interest, a role for the •OH in the AP-1-dependent regulation of gadd153 has been reported (8), and our

820

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

findings are consistent with a •OH/AP-1-dependent regulation of gadd153. In support of this statement, TGHQmediated TRE-binding activity (Figure 3) and gadd153 gene expression (6) exhibit comparable temporal- and concentration-dependent regulation (50-500 µM TGHQ groups shown in ref 6 are equivalent to 8-80 nmol TGHQ/cm2). Collectively, these observations form a plausible model where •OH-dependent genotoxic stress results in the transcriptional upregulation of AP-1, which in turn, regulates gadd153 expression. As described in the Results, the concentration-dependent correlate for TRE-binding activity and gadd153 expression would not be apparent if the data were expressed as molarity. AP-1 and NF-κB are regulated by a number of signal transduction pathways, including those coupled to ERK activation (33-35) and PKC (15). TGHQ-mediated TREbinding activity is markedly reduced by a MEK inhibitor (Figure 8), suggesting an important role for the MAPK cascade in the TRE-binding response to TGHQ, consistent with our previous findings that quinol-thioethers increase ERK activity in LLC-PK1 cells.2 TRE-binding activity was not fully inhibited by PD098059, suggesting this response may be regulated by additional MEKindependent signaling events, consistent with the observation that interventions targeting both the genotoxicand ERK-stress response pathways dramatically reduce inducible TRE-binding activity (Figures 5 and 8). In contrast, NF-κB-binding activity from the same nuclear extracts was not inhibited by PD098059 (Figure 9). This observation suggests that the NF-κB response is not mediated by MEK-related signaling. Many “nongenotoxic” nephrotoxicants induce tumors in the kidney of mice and rats, suggesting a specific effect on a “preinitiated” population of cells (46, 47). Cytotoxic modes of chemical-induced carcinogenesis have also been implicated for toxicants such as chloroform (48). Therefore, there is precedence for a number of mechanisms by which quinol-thioethers may induce tumors or modulate carcinogenic processes. In particular, quinol-thioethermediated cytotoxicity is associated with a prominent oxidative stress (4), and ROSs are widely implicated in carcinogenic processes (49). Acute oxidative injury is thought to produce cell death and a compensatory increase in cell proliferation, resulting in the clonal expansion of preinitiated cells. Alternatively, a sublethal concentration of ROS could result in damage to DNA which, if mis-repaired, would lead to the generation of newly initiated cells. In this context, TGHQ is mutagenic in a manner consistent with the participation of the •OH in DNA damage (26). Chronic oxidative injury may contribute to carcinogenesis through the modulation of growth-related signal transduction pathways, such as those associated with AP-1 and NF-κB, which are activated by ROS and implicated as causal in transformation responses (8, 9, 11, 16, 18, 29). Our data indicate that quinol-thioether-mediated TRE- and NF-κB-binding activities precedes overt cytotoxicity (Figures 1 and 4) and, therefore, may contribute to the carcinogenic properties of this compound. The induction of TRE- and NF-κB-binding activity is also associated with the regulation of cell death and survival in mammalian cells. NF-κB activation is associ2 Huang, Q., Lau, S. S., and Monks, T. J. (1999) Inhibition of the MAPK pathway blocks quinone-thioether mediated cytotoxicity, but not growth arrest, in LLC-PK1 cells. The Toxicologist 48, 1340.

Weber et al.

ated with apoptosis in rat striatum following kainic acid treatment (50). In contrast, the activation of AP-1 and NF-κB by H2O2 in oligodendrocytes is believed to promote cell survival, while the formation of •OH promotes cell death (51). Prostanoid-mediated activation of AP-1 and NF-κB induces protection against TGHQ-induced cytotoxicity in LLC-PK1 cells (6, 15, 32), suggesting these transcription factors promote survival in renal epithelial cells. Promotion of cell survival by AP-1 and/or NF-κB would have plausible implications in carcinogenic processes and could result in the fixation of mutations induced by quinol-thioethers (26). However, it is important to note that the signal transduction pathways regulating TRE- and NF-κB-binding activity by prostanoids and TGHQ are clearly different. Prostanoidmediated DNA-binding activities are inhibited by H-89 and calphostin C (15), while these PKC inhibitors have no effect on TGHQ-mediated DNA-binding activities (Figure 7). Therefore, additional studies are required to determine whether PKC-independent activation of TREand NF-κB-binding activity will also promote cell survival in renal epithelial cells. In summary, we have shown that TGHQ-mediated TRE- and NF-κB-binding activities are ROS-dependent in LLC-PK1 cells. Peak TGHQ-induced TRE-binding activity is increased in a •OH-dependent fashion, transcriptionally dependent, precedes overt cytotoxicity, and is markedly reduced by a MEK inhibitor. Peak TGHQinduced NF-κB-binding activity appears to be regulated by both H2O2 and the •OH, and is not reduced by a MEK inhibitor. The differential effects of deferoxamine and PD098059 on TGHQ-mediated TRE- and NF-κB-binding activity provide evidence that these transcription factors are regulated by different signaling pathways after H2O2 generation. Collectively, our findings support an ROSdependent effect of TGHQ on signal transduction events implicated in tumor promotion.

Acknowledgment. This work was supported by a grant from the Department of Energy (DE-AC06-76RLO 1830) to Dr. Thomas Weber and by a grant from the National Institute of General Medical Sciences (GM 39338) to Dr. Serrine Lau. This work was also supported by grants from the National Institute of Environmental Health Sciences (T32 ES 07247 and ES 07884).

References (1) Li, P. F., Dietz, R., and von Harsdorf, R. (1997) Differential effect of hydrogen peroxide and superoxide anion on apoptosis and proliferation of vascular smooth muscle cells. Circulation 96, 3602-3609. (2) Dahm-Daphi, J., Sass, C., and Alberti, W. (2000) Comparison of biological effects of DNA damage induced by ionizing radiation and hydrogen peroxide in CHO cells. Int. J. Radiat. Biol. 76, 6775. (3) Polla, B. S. (1999) Therapy by taking away: the case of iron. Biochem. Pharmacol. 57, 1345-1349. (4) Mertens, J. J., Gibson, N. W., Lau, S. S., and Monks, T. J. (1995) Reactive oxygen species and DNA damage in 2-bromo-(glutathionS-yl) hydroquinone-mediated cytotoxicity. Arch. Biochem. Biophys. 320, 51-58. (5) Monks, T. J., and Lau, S. S. (1992) Toxicology of quinonethioethers. Crit. Rev. Toxicol. 22, 243-270. (6) Towndrow, K. M., Mertens, J. J., Jeong, J. K., Weber, T. J., Monks, T. J., and Lau, S. S. (2000) Stress- and growth-related gene expression are independent of chemical-induced prostaglandin E2 synthesis in renal epithelial cells. Chem. Res. Toxicol. 13, 111117. (7) Hsu, T. C., Young, M. R., Cmarik, J., and Colburn, N. H. (2000) Activator protein 1 (AP-1)- and nuclear factor kappaB (NF-

Stress-Responsive Signaling

(8) (9)

(10) (11)

(12) (13) (14) (15)

(16)

(17) (18)

(19)

(20)

(21) (22) (23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

kappaB)-dependent transcriptional events in carcinogenesis. Free. Radical Biol. Med. 28, 1338-1348. Guyton, K. Z., Xu, Q., and Holbrook, N. J. (1996) Induction of the mammalian stress response gene GADD153 by oxidative stress: role of AP-1 element. Biochem. J. 314, 547-554. Schmidt, K. N., Amstad, P., Cerutti, P., and Baeuerle, P. A. (1996) Identification of hydrogen peroxide as the relevant messenger in the activation pathway of transcription factor NF-kappaB. Adv. Exp. Med. Biol. 387, 63-68. Ding, M., Li, J. J., and Leonard, S. S., et al. (1999) Vanadateinduced activation of activator protein-1: role of reactive oxygen species. Carcinogenesis 20, 663-668. Li, J. J., Westergaard, C., Ghosh, P., and Colburn, N. H. (1997) Inhibitors of both nuclear factor-kappaB and activator protein-1 activation block the neoplastic transformation response. Cancer Res. 57, 3569-3576. Curran, T., and Franza, B. R., Jr. (1988) Fos and Jun: the AP-1 connection. Cell 55, 395-397. Wahl, C., Liptay, S., Adler, G., and Schmid, R. M. (1998) Sulfasalazine: a potent and specific inhibitor of nuclear factor kappa B. J. Clin. Invest. 101, 1163-1174. Siebenlist, U., Franzoso, G., and Brown, K. (1994) Structure, regulation and function of NF-kappa B. Annu. Rev. Cell Biol. 10, 405-455. Weber, T. J., Monks, T. J., and Lau, S. S. (1997) PGE2-mediated cytoprotection in renal epithelial cells: evidence for a pharmacologically distinct receptor. Am. J. Physiol. 273, F507-F515. Barchowsky, A., Munro, S. R., Morana, S. J., Vincenti, M. P., and Treadwell, M. (1995) Oxidant-sensitive and phosphorylationdependent activation of NF-kappa B and AP-1 in endothelial cells. Am. J. Physiol. 269, L829-L836. Forrest, D., and Curran, T. (1992) Crossed signals: oncogenic transcription factors. Curr. Opin. Genet. Dev. 2, 19-27. Dong, Z., Birrer, M. J., Watts, R. G., Matrisian, L. M., and Colburn, N. H. (1994) Blocking of tumor promoter-induced AP-1 activity inhibits induced transformation in JB6 mouse epidermal cells. Proc. Natl. Acad. Sci. U.S.A. 91, 609-613. Huang, C., Ma, W. Y., Young, M. R., Colburn, N., and Dong, Z. (1998) Shortage of mitogen-activated protein kinase is responsible for resistance to AP-1 transactivation and transformation in mouse JB6 cells. Proc. Natl. Acad. Sci. U.S.A. 95, 156-161. Watts, R. G., Huang, C., Young, M. R., Li, J. J., Dong, Z., Pennie, W. D., and Colburn, N. H. (1998) Expression of dominant negative Erk2 inhibits AP-1 transactivation and neoplastic transformation. Oncogene 17, 3493-3498. Huang, C., Ma, W. Y., and Dong, Z. (1999) The extracellularsignal-regulated protein kinases (Erks) are required for UVinduced AP-1 activation in JB6 cells. Oncogene 18, 2828-2835. Nishizuka, Y. (1992) Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase C. Science 258, 607-614. Kari, F. W., Bucher, J., Eustis, S. L., Haseman, J. K., and Huff, J. E. (1992) Toxicity and carcinogenicity of hydroquinone in F344/N rats and B6C3F1 mice. Food. Chem. Toxicol. 30, 737747. Lau, S. S., Monks, T. J., Everitt, J. I., Kleymenova, E., and Walker, C. L. (2001) Carcinogenicity of a nephrotoxic metabolite of the “nongenotoxic” carcinogen hydroquinone. Chem. Res. Toxicol. 14, 25-33. Yoon, H. S., Monks, T. J., Walker, C. L. and Lau, S. S. (2001) Transformation of kidney epithelial cells by a quinol thioether via inactivation of the tuberous sclerosis-2 tumor suppressor gene. Mol. Carcinog. 31, 37-45. Jeong, J. K., Wogan, G. N., Lau, S. S., and Monks, T. J. (1999) Quinol-glutathione conjugate-induced mutation spectra in the supF gene replicated in human AD293 cells and bacterial MBL50 cells. Cancer Res. 59, 3641-3645. Peters, M. M., Jones, T. W., Monks, T. J., and Lau, S. S. (1997) Cytotoxicity and cell-proliferation induced by the nephrocarcinogen hydroquinone and its nephrotoxic metabolite 2,3,5-(trisglutathion-S-yl)hydroquinone. Carcinogenesis 18, 2393-2401. Lau, S. S., Hill, B. A., Highet, R. J., and Monks, T. J. (1988) Sequential oxidation and glutathione addition to 1,4-benzoquinone: correlation of toxicity with increased glutathione substitution. Mol. Pharmacol. 34, 829-836. Meyer, M., Schreck, R., and Baeuerle, P. A. (1993) H2O2 and antioxidants have opposite effects on activation of NF-kappa B and AP-1 in intact cells: AP-1 as secondary antioxidantresponsive factor. EMBO J. 12, 2005-2015. Jeong, J. K., Stevens, J. L., Lau, S. S., and Monks, T. J. (1996) Quinone thioether-mediated DNA damage, growth arrest, and gadd153 expression in renal proximal tubular epithelial cells. Mol. Pharmacol. 50, 592-598.

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 821 (31) Newton, R., Adcock, I. M., and Barnes, P. J. (1996) Superinduction of NF-kappa B by actinomycin D and cycloheximide in epithelial cells. Biochem. Biophys. Res. Commun. 218, 518-523. (32) Weber, T. J., Monks, T. J., and Lau, S. S. (2000) DDM-PGE2mediated cytoprotection in renal epithelial cells by a thromboxane A2 receptor coupled to NF-kappaB. Am. J. Physiol. Renal. Physiol. 278, F270-F278. (33) Zhao, Q., and Lee, F. S. (1999) Mitogen-activated protein kinase/ ERK kinase kinases 2 and 3 activate nuclear factor-kappaB through IkappaB kinase-alpha and IkappaB kinase- beta. J. Biol. Chem. 274, 8355-8358. (34) Ben-Ari, E. T., Bernstein, L. R., and Colburn, N. H. (1992) Differential c-jun expression in response to tumor promoters in JB6 cells sensitive or resistant to neoplastic transformation. Mol. Carcinog. 5, 62-74. (35) Bernstein, L. R., Ferris, D. K., Colburn, N. H., and Sobel, M. E. (1994) A family of mitogen-activated protein kinase-related proteins interacts in vivo with activator protein-1 transcription factor. J. Biol. Chem. 269, 9401-9404. (36) Boden, S. E., Bertsche, T., Ammon, H. P., and Safayhi, H. (2000) MEK-1/2 inhibition prevents 5-lipoxygenase translocation in N-formylpeptide-challenged human neutrophils. Int. J. Biochem. Cell Biol. 32, 1069-1074. (37) Jeong, J. K., Dybing, E., Soderlund, E., Brunborg, G., Holme, J. A., Lau, S. S., and Monks, T. J. (1997) DNA damage, gadd153 expression, and cytotoxicity in plateau-phase renal proximal tubular epithelial cells treated with a quinol thioether. Arch. Biochem. Biophys. 341, 300-308. (38) Iwata, E., Asanuma, M., Nishibayashi, S., Kondo, Y., and Ogawa, N. (1997) Different effects of oxidative stress on activation of transcription factors in primary cultured rat neuronal and glial cells. Brain Res. Mol. Brain Res. 50, 213-220. (39) Xanthoudakis, S., and Curran, T. (1996) Redox regulation of AP1: a link between transcription factor signaling and DNA repair. Adv. Exp. Med. Biol. 387, 69-75. (40) Monks, T. J., Hanzlik, R. P., Cohen, G. M., Ross, D., and Graham, D. G. (1992) Quinone chemistry and toxicity. Toxicol. Appl. Pharmacol. 112, 2-16. (41) Brennan, P., and O’Neill, L. A. (1996) 2-Mercaptoethanol restores the ability of nuclear factor kappa B (NF kappa B) to bind DNA in nuclear extracts from interleukin 1-treated cells incubated with pyrollidine dithiocarbamate (PDTC). Evidence for oxidation of glutathione in the mechanism of inhibition of NF kappa B by PDTC. Biochem. J. 320, 975-981. (42) Davis, T. W., Meyers, M., and Wilson-Van Patten, C., et al. (1998) Transcriptional responses to damage created by ionizing radiation. In DNA Damage and Repair (Nickoloff, J. A., and Hoekstra, M. F., Eds.) pp 223-262, Humana Press, Totowa, NJ. (43) Jackman, J., Alamo, I., Jr., and Fornace, A. J., Jr. (1994) Genotoxic stress confers preferential and coordinate messenger RNA stability on the five gadd genes. Cancer Res. 54, 5656-5662. (44) Amundson, S. A., Zhan, Q., Penn, L. Z., and Fornace, A. J., Jr. (1998) Myc suppresses induction of the growth arrest genes gadd34, gadd45, and gadd153 by DNA-damaging agents. Oncogene 17, 2149-2154. (45) Luethy, J. D., and Holbrook, N. J. (1994) The pathway regulating GADD153 induction in response to DNA damage is independent of protein kinase C and tyrosine kinases. Cancer Res. 54, 19021906. (46) Hiasa, Y., Ohshima, M., Kitahori, Y., Konishi, N., Fujita, T., and Yuasa, T. (1982) beta-Cyclodextrin: promoting effect on the development of renal tubular cell tumors in rats treated with N-ethyl-N-hydroxyethylnitrosamine. J. Natl. Cancer Inst. 69, 963-967. (47) Frank, D. W., Gray, J. E., and Weaver, R. N. (1976) Cyclodextrin nephrosis in the rat. Am. J. Pathol. 83, 367-382. (48) Butterworth: B. E., Templin, M. V., Borghoff, S. J., Conolly, R. B., Kedderis, G. L., and Wolf, D. C. (1995) The role of regenerative cell proliferation in chloroform-induced cancer. Toxicol. Lett. 8283, 23-26. (49) Klaunig, J. E., Xu, Y., Isenberg, J. S., et al. (1998) The role of oxidative stress in chemical carcinogenesis. Environ. Health Perspect. 106 (Suppl. 1), 289-295. (50) Nakai, M., Qin, Z. H., Chen, J. F., Wang, Y., and Chase, T. N. (2000) Kainic acid-induced apoptosis in rat striatum is associated with nuclear factor-kappaB activation. J. Neurochem. 74, 647658. (51) Vollgraf, U., Wegner, M., and Richter-Landsberg, C. (1999) Activation of AP-1 and nuclear factor-kappaB transcription factors is involved in hydrogen peroxide-induced apoptotic cell death of oligodendrocytes. J. Neurochem. 73, 2501-2509.

TX000190R