Several genes encoding detoxifying and antioxidative

Nrf2 Is Increased by CYP2E1 in Rodent Liver and HepG2 Cells and Protects Against Oxidative Stress Caused by CYP2E1 Pengfei Gong and Arthur I. Cederbaum Induction of CYP2E1 by ethanol is one pathway through which ethanol generates oxidative stress. Nrf2 is a transcription factor that regulates important antioxidant and phase II detoxification genes. Nrf2 induction by CYP2E1 and its importance in the adaptive response to increased oxidative stress caused by CYP2E1 was studied. Increases in Nrf2 protein and mrna were observed in livers or hepatocytes of chronic alcohol-fed mice or rats and of pyrazole-treated rats or mice, conditions known to elevate CYP2E1. HepG2 cells expressing CYP2E1 (E47 cells) showed increased Nrf2 mrna and protein expression compared with control HepG2 C34 cells. Nrf2 is activated in E47 cells as shown by an increase in nuclear Nrf2 levels and Nrf2 antioxidant-responsive element binding activity, and upregulation of Nrf2-regultated genes, glutamate cysteine ligase catalytic subunit (GCLC), and heme oxygenase 1 (HO-1). Increases in Nrf2 protein and mrna are blocked by inhibitors of CYP2E1 activity and a reactive oxygen species (ROS) scavenger, N-acetylcysteine, which decrease ROS levels as well as Nrf2 mrna induction. Upregulation of GCLC and HO-1 in E47 cells is dependent on Nrf2 and is prevented by sirna- Nrf2. Blocking Nrf2 by sirna-nrf2 decreases glutathione and increases ROS and lipid peroxidation, resulting in decreased mitochondrial membrane potential and loss of cell viability of E47 cells but not C34 cells. These results suggest that Nrf2 is activated and that levels of protein and mrna are increased when CYP2E1 is elevated. In conclusion, Nrf2 plays a key role in the adaptive response against increased oxidative stress caused by CYP2E1. Supplementary material for this article can be found on the HEPATOLOGY Website (http://interscience.wiley.com/jpages/ 0270-9139/suppmat/index.html). (HEPATOLOGY 2006;43:144-153.) Several genes encoding detoxifying and antioxidative stress enzymes are coordinately induced on exposure to electrophiles and reactive oxygen species (ROS). 1-2 This coordinated response is regulated through a cis-acting element called the antioxidant-responsive element (ARE) within the regulatory region of target Abbreviations: ROS, reactive oxygen species; sirna, small interfering RNA; ARE, antioxidant-responsive element; GCL, glutamate cysteine ligase; CYP2E1, cytochrome P450 2E1; Nrf2, nuclear factor erythroid 2-related factor 2; mrna, messenger RNA; GCLC, the catalytic subunit of glutamate cysteine ligase; HO-1, heme oxygenase 1; GSH, glutathione; TBARS, thiobarbituric acid reactive species; cdna, complementary DNA. From the Department of Pharmacology and Biological Chemistry, Mount Sinai School of Medicine, New York, NY. Received July 5, 2005; accepted October 15, 2005. Supported by USPHS Grant AA 03312 from the National Institute on Alcohol Abuse and Alcoholism. Address reprint requests to: Dr. Arthur I. Cederbaum, Department of Pharmacology and Biological Chemistry, Mount Sinai School of Medicine, Box 1603, One Gustave L. Levy Place, New York, NY 10029. E-mail: arthur.cederbaum@ mssm.edu; fax: 212-996-7214. Copyright 2005 by the American Association for the Study of Liver Diseases. Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/hep.21004 Potential conflict of interest: Nothing to report. genes. 3-4 Genes encoding a subset of drug-metabolizing enzymes such as glutathione-s-transferases 3 and NAD(P)H-quinone oxidoreductase 1 4 have been shown to be under ARE regulation, along with a subset of antioxidant genes, such as heme oxygenase 1 (HO-1) 5 and glutamate cysteine ligase (GCL). 6 The signaling system leading to ARE activation has been partly elucidated, and nuclear factor erythroid 2-related factor 2 (Nrf2) has been identified as the key transcriptional factor that transmits the inducer signal to ARE. 7 Nrf2 is normally sequestered in the cytoplasm by Kelch-like ECH-associated protein 1. 8 After activation, Nrf2 dissociates from Kelch-like ECH-associated protein 1 and translocates into the nucleus, where it complexes with other nuclear factors and binds to ARE, activating transcription of many antioxidant genes and phase II detoxification genes that have ARE elements in their promoter regions. 7 Nrf2 is constitutively and ubiquitously expressed in several tissues and cell lines and is responsible for the low-level expression of its target genes observed under physiological conditions. However, in cells exposed to oxidative stress, Nrf2 activity 144

HEPATOLOGY, Vol. 43, No. 1, 2006 GONG AND CEDERBAUM 145 is increased, further driving the transcriptional activation of genes whose expression is essential to protect cells against loss of viability. 9 Cytochrome P450 2E1 (CYP2E1) metabolizes and activates many toxicologically important substrates to more toxic products. 10-12 Induction of CYP2E1 by ethanol is a central pathway by which ethanol generates oxidative stress in hepatocytes. 13-15 Our laboratory established a HepG2 cell line that constitutively overexpresses CYP2E1 (E47 cells) and characterized CYP2E1-dependent toxicity in the presence of ethanol, arachidonic acid, and AA plus iron in these cell lines. 16-17 These toxicity effects were enhanced when cellular glutathione (GSH) levels were lowered by treatment with L -buthionine- (S,R)-sulfoximine. 18-19 In response to the increased oxidative stress caused by CYP2E1, many antioxidant genes were upregulated in E47 cells, including the catalytic subunit of GCL (GCLC), 20 HO-1, 21 and catalase, 22 possibly reflecting an adaptation to CYP2E1-derived oxidative stress. In this study, we attempted to identify the functional effects of Nrf2 in the upregulation of these antioxidant genes and its importance in protecting against toxicity in cells with high CYP2E1 expression. Materials and Methods Animal Models. Male C57BL/6J mice (30-40 g) and male Sprague-Dawley rats (150-170 g) were housed in a facility approved by The American Association for Accreditation of Laboratory Animal Care and fed with ethanol (35%) and control diets (Bio-Serv, Frenchtown, NJ) that were equicaloric and had the same composition with respect to fat (42% of calories) and protein (16% of calories) for 25 days or 2 months. Hepatocytes were isolated as previously described. 23 To induce liver CYP2E1, male rats (200 mg/kg) or male mice (120 mg/kg) were injected intraperitoneally with pyrazole once per day for 2 days, followed by overnight fast. 23 Cell Culture and Transfection Experiments. HepG2 cells that constitutively express CYP2E1 (E47 cells) or control HepG2 cells (C34 cells) that have undetectable P450 activity 24 were cultured in minimal essential medium containing 10% fetal bovine serum and 0.5 mg/ml of G418 supplemented with 100 U/mL of penicillin and 100 g/ml of streptomycin and 2 mmol/l L-glutamine in a humidified atmosphere in 5% CO 2 at 37 C. Cells were plated and incubated in minimal essential medium overnight, the culture medium was replaced with fresh medium, and the different treatments were initiated. Small interfering RNA (sirna)-nrf2 and sirna-control (a nontargeting sirna) (Santa Cruz Biotechnology, Santa Cruz, CA) were transfected using the sirna transfection regent according to the manufacturer s protocol. To study the in vitro effect of ethanol on Nrf2 protein levels, 1 10 6 C34 or E47 cells were incubated for 0, 1, 2, and 3 days with 100 mmol/l ethanol, and Nrf2 protein levels were determined via Western blot analysis. The 6-well plates were wrapped tightly with parafilm, and fresh ethanol was added daily. General Methodology. Cell viability was measured via MTT assay. 24 ROS were determined via flow cytometry with 2 7 -dichlorofluorescin diacetate. 20-21 Lipid peroxidation was determined by measurement of the concentration of thiobarbituric acid reactive species (TBARS) in cell lysates. 25 GSH was determined as previously described. 20 Mitochondrial membrane potential was analyzed via flow cytometry after double-staining with 5 g/ml rhodamine 123 and 5 g/ml of propidium iodide. 26 Western Blot Analysis. Nrf2, GCLC, HO-1, and catalase proteins were detected via Western blot analysis. 17, 20-22 Sample proteins from rat or mouse liver homogenates (100 g), whole cell extracts (50 g), or nuclear extracts (25 g) were loaded on a 12% denaturing polyacrylamide gel and electroblotted onto 0.2 m nitrocellulose membranes. Protein concentration was determined using the Protein DC-20 Assay Kit (Bio-Rad, Hercules, CA). Protein immunoblot analysis was performed using the following: anti-human HO-1 (1:5,000) monoclonal antibody (StressGen Biotech, San Diego, CA); anti-human Nrf2 (1:5,000) polyclonal antibody (Santa Cruz Biotechnology); anti-human GCLC (1:500) polyclonal antibody (Lab Vision Corp., Fremont, CA); anti-human catalase (1:10,000) polyclonal antibody (Calbiochem, San Diego, CA); and anti-human -actin (1: 10,000) monoclonal antibody (Sigma, St. Louis, MO) as primary antibodies and horseradish peroxidase conjugated goat anti mouse immunoglobulin G (1:4,000) or goat anti rabbit immunoglobulin G (1:10,000) (Sigma) as secondary antibody. Blots were developed using the enhanced chemiluminescence immunoblot-detecting reagent (Amersham, Piscataway, NJ). Northern Blot Analysis. Total RNA was isolated using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA). Ten micrograms of RNA were electrophoresed under denaturing conditions in 0.9% agarose/ formaldehyde gels, transferred onto nylon membranes, and hybridized to random-primed 32 P-labeled Nrf2, GCLC, HO-1, catalase, or GAPDH complementary DNA (cdna) probes. 21 Mouse Nrf2 cdna was produced by digesting the plasmid pef/nrf2 27 with Not1 and BglII. Human HO-1 cdna was produced by digesting the plasmid phho1 21 with Xho1 and Xba1. Human

146 GONG AND CEDERBAUM HEPATOLOGY, January 2006 GCLC cdna, human catalase cdna, and human GAPDH cdna were purchased from American Type Culture Collection (Manassas, VA). Electrophoretic Mobility Shift Assay. Crude nuclear extracts were prepared as described previously. 28 The standard binding reaction mixture (12.5 L) contained 18 mmol/hepes (ph 7.9), 80 mmol/l KCl, 2 mmol/l MgCl 2, 10 mmol/l dithiothreitol, 10% glycerol, 0.2 mg/ml bovine serum albumin, 160 g/ml poly(di-dc), 20,000 cpm [- 32 P]ATP-labeled probe, and 2 g nuclear extract. Reaction mixtures were incubated at 25 C for 20 minutes and analyzed via native 5% polyacrylamide gel electrophoresis and autoradiography as previously described. 29 Double-stranded oligonucleotides corresponding to a GCLC ARE (5 -GCGCGCGCACCGC-CTCC- GTGACTCAGCGCTTTGTGCG-3, core ARE sequence underlined) was used as probe. In supershift assays, 1 L of preimmune immunoglobulin G or anti- Nrf2 antibody was added to the reaction mixture and incubated for 20 minutes at room temperature before electrophoresis. Nuclear In Vitro Run-on Transcription Assays. Cultured cells were washed twice with phosphate-buffered saline, scraped off, and transferred into 15-mL Falcon tubes for centrifugation at 500g at 4 C for 5 minutes. Pellets were loosened and resuspended in 4 ml of buffer A (10 mmol/l Tris [ph 7.4], 10 mmol/l NaCl, and 3 mmol/l MgCl 2 ). After addition of 0.5% (v/v) Nonidet P-40, samples were vortexed for 10 seconds and kept on ice for 10 minutes. After a second centrifugation, supernatants containing cytoplasmic RNA were discarded, and 4 ml of buffer A was added. Samples were kept on ice for 10 minutes and spun down. Nuclei were collected in 100 L of storage buffer (50 mmol/l Tris [ph 8.3], 40% glycerol, 5 mmol/l MgCl 2, and 0.1 mmol/l EDTA) and immediately stored at 70 C until analyzed. Transcription was assayed in vitro with 100 L nuclei (isolated from 25 10 6 cells), 100 Ci [ 32 P]UTP, and 100 L assay buffer (2.5 mmol/l MnCl 2, 125 mmol/l Tris [ph 7.4], 12.5 mmol/l MgCl 2, 2.5 mmol/l dithiothreitol, 2.5 mmol/l adenosine triphosphate, 1.25 mmol/l guanosine triphosphate (GTP), 1.25 mmol/l cytidine triphosphate (CTP), and 375 mmol/l KCl). The reaction was performed at 25 C for 45 minutes and stopped by adding DNase I (RNase-free) at a final concentration of 20 g/ml. After incubation at 30 C for 5 minutes, RNA was released from the nuclei by treatment with 1% SDS, 5 mmol/l EDTA, 10 mmol/l HEPES (ph 7.5), and 200 g/ml proteinase K at 42 C for 30 minutes. Newly transcribed RNA was isolated using Trizol reagent. Eight micrograms of each cdna of interest was denatured in 0.25 N NaOH and 0.5 mol/l NaCl for 10 minutes at room temperature and diluted in 0.1 SSC and 0.125 N NaOH. Samples were blotted onto GeneScreen membranes, neutralized in 0.5 N NaCl and 0.5 mol/l Tris (ph 7.5), and UV cross-linked. Prehybridization was performed for 6 hours at 42 C, and purified denatured riboprobes (1 10 6 cpm/ml) were used for hybridization for 72 hours at 42 C. The composition of the hybridization buffer was the same as described for the Northern blot assay, using 0.1 mg/ml baker s yeast transfer RNA instead of the salmon sperm DNA. After four washes of 15 min each with 2 SSC and 0.5% SDS at 65 C, membranes were exposed to the phosphor storage screen, and the signals were quantified using the GAPDH mrna signal as a control. Statistical Analysis. The Student t test for unpaired data was used to evaluate the differences between the compared groups. Results Nrf2 Protein and mrna Are Increased in Livers of Mice Fed Ethanol. Little is known about changes of Nrf2, if any, in the liver of animals treated with ethanol. Male C57BL/6J mice (30-40 g) were fed with a control diet or a diet containing up to 35% ethanol for 25 days. Nrf2 and CYP2E1 protein levels in the homogenates of liver were detected via Western blot analysis. Nrf2 and GAPDH mrna levels were determined via Northern blot analysis. Ethanol feeding increased the CYP2E1 protein level approximately fourfold (P.001), the Nrf2 protein level approximately twofold (P.01), and Nrf2 mrna level approximately 1.7-fold (P.01) compared with mice fed a control diet (Fig. 1A-B). Nrf2 Protein and mrna Are Increased in Hepatocytes of Rats Fed Ethanol. Male Sprague-Dawley rats were fed a control diet or a diet containing 35% ethanol for 2 months. Hepatocytes were isolated from the livers. Ethanol feeding increased the CYP2E1 protein level approximately threefold (P.001), the Nrf2 protein level approximately 2.6-fold (P.001), and Nrf2 mrna level approximately 2.5-fold (P.001) compared with rats fed a control diet (Supplementary Fig. 1A-B; supplementary material available at the HEPATOLOGY website [http://interscience.wiley.com/jpages/0270-9139/ suppmat/index.html]. ). These results show that chronic ethanol feeding in mice or rats increases Nrf2 protein and mrna levels, and that these increases can be found in hepatocytes. Induction of CYP2E1 In Vivo by Pyrazole Increases Nrf2 Protein Expression. Pyrazole induces CYP2E1 by

HEPATOLOGY, Vol. 43, No. 1, 2006 GONG AND CEDERBAUM 147 Fig. 1. Nrf2 expression is increased in the livers of mice fed ethanol. Male C57BL/6J mice (30-40 g) were fed with a control diet or a diet containing ethanol (up to 35%) for 25 days. (A) Nrf2 and CYP2E1 protein levels were determined via Western blotting and quantified. (B) Nrf2 and GAPDH mrna levels were determined via Northern blotting and quantified. The amount of Nrf2 mrna was normalized to GAPDH mrna. Results are expressed as arbitrary densitometric units or ratios under the representative protein or mrna bands. Although three samples are shown, the data are expressed as the mean SE (n 6). **P.01. ***P.001 versus control mice. CYP2E1, cytochrome P450 2E1; Nrf2, nuclear factor erythroid 2-related factor 2. stabilizing the protein against degradation. To evaluate whether there were any changes in Nrf2 levels due to the increased CYP2E1 activity in hepatocytes, male rats or male mice were treated with 0.9% NaCl (control) or pyrazole for 2 days, followed by an overnight fast. After sacrifice, livers were removed and CYP2E1 and Nrf2 content were determined via Western blot analysis. CYP2E1 protein was increased by pyrazole treatment approximately threefold in rats and 2.5-fold in mice, while Nrf2 protein was induced approximately 3.5-fold in rats and 2.6-fold in mice compared with saline treatment (Fig. 2). Fig. 2. Nrf2 protein is increased in liver of mice or rats injected with pyrazole. Male rats (200 mg/kg) or male mice (120 mg/kg) were injected intraperitoneally with pyrazole, once per day for 2 days, followed by overnight fast. Nrf2 and CYP2E1 proteins in liver homogenates were detected via Western blotting and quantified. Results are expressed as arbitrary densitometric units under the representative protein bands and represent the mean SE (n 4). ***P.001 versus control mice or rats. CYP2E1, cytochrome P450 2E1; Nrf2, nuclear factor erythroid 2-related factor 2. Fig. 3. Nrf2 protein is increased in E47 cells. (A) Nrf2 protein levels are increased in E47 cells. C34 or E47 cells (1 10 6 cells/well in 6-well plates) were incubated overnight. Nrf2 and CYP2E1 protein levels in the whole cell lysates were detected via Western blotting and quantified. Results are expressed as arbitrary densitometric units under the protein bands and represent the mean SE (n 3). (B) Stability of Nrf2 protein. C34 or E47 cells (1 10 6 cells/well in 6-well plates) were incubated overnight, and cycloheximide (50 g/ml) was added. After incubation with or without cycloheximide for 15, 30, 45, 60, and 120 minutes, cells were collected and Nrf2 protein levels in the whole cell lysates were detected via Western blotting and quantified. Results are expressed as the percentage of untreated control C34 or E47 cells and represent the mean SE (n 3). *P.05 versus control C34 cells at the same time points; ***P.001 versus control C34 cells. CYP2E1, cytochrome P450 2E1; Nrf2, nuclear factor erythroid 2-related factor 2. Nrf2 Protein Is Increased in E47 Cells. To determine whether there is any link between the induction of Nrf2 and CYP2E1, E47 cells or C34 cells were used. Figure 3A shows that CYP2E1 protein was only detected in E47 cells and that Nrf2 protein in E47 cells was 2.8- fold higher than in C34 cells. To evaluate whether this increase of Nrf2 protein in E47 cells is the result of decreased degradation or increased expression, cycloheximide (50 g/ml) was used to block the new synthesis of Nrf2 protein. The degradation rate of Nrf2 protein in E47 cells was slightly decreased at 15 and 30 minutes (apparent half-life of Nrf2 protein decreased from about 15 min to 26 minutes; P.05) compared with that seen in C34 cells (Fig. 3B). Although stabilization of Nrf2 against degradation may contribute somewhat to the elevated levels of Nrf2 in the E47 cells, other mechanisms (e.g., changes in mrna levels) may be more important (see later discussion). Nuclear Nrf2 Protein Level and Nrf2-ARE Binding Activity Are Increased in E47 Cells. To study whether

148 GONG AND CEDERBAUM HEPATOLOGY, January 2006 that it contains Nrf2. Together, these results suggest that Nrf2 is activated in E47 cells. Nrf2 mrna Expression Is Increased in E47 Cells. Nrf2 mrna levels in E47 and C34 cells were assayed via Northern blotting. Figure 5A shows that Nrf2 mrna levels were significantly increased in E47 cells compared with C34 cells. To evaluate whether this increase is caused by increased Nrf2 stability, we added actinomycin D (10 g/ml) to block new mrna transcription. Figure 5B shows that the stability of Nrf2 mrna in C34 and E47 cells was basically similar. Nuclear run-on assay showed Fig. 4. Nuclear Nrf2 levels and Nrf2-ARE binding activity are increased in E47 cells. (A) Nuclear Nrf2 protein levels are increased in E47 cells. C34 and E47 cells (2 10 8 cells in 150-mm dishes) were incubated overnight. Nrf2 protein levels in the nuclear extracts were detected via Western blotting and quantified. Results are expressed as arbitrary densitometry units under the Nrf2 bands and represent the mean SE (n 3). **P.01 versus control C34 cells. (B) Nrf2-ARE binding activity is increased in E47 cells. C34 and E47 cells (2 10 8 cells in 150-mm dishes) were incubated overnight. Nuclear extracts were prepared, and electrophoretic mobility shift assays were performed in the absence or presence of anti-nrf2 antibody according to the method described in Materials and Methods. Nrf2, nuclear factor erythroid 2-related factor 2; ARE, antioxidant-responsive element. increased CYP2E1 expression can activate Nrf2, we examined the nuclear Nrf2 protein levels and Nrf2-ARE binding activity in E47 cells and C34 cells. Figure 4A shows that the level of Nrf2 protein in the crude nuclear extracts of E47 cells was significantly increased compared with C34 cells. To determine whether Nrf2 binding activity to the ARE was affected by CYP2E1, nuclear extracts of C34 and E47 cells were prepared, and Nrf2 binding activity was determined via electrophoretic mobility shift assay with a double-strand DNA containing the ARE sequence as a probe. ARE binding activity was increased in E47 cells compared with C34 cells. Nrf2 antibody can supershift this complex (Fig. 4B), indicating Fig. 5. Nrf2 mrna expression is increased in E47 cells. (A) Nrf2 mrna levels are increased in E47 cells. C34 and E47 cells (1 10 7 cells in 100-mm dishes) were incubated overnight, and Nrf2 and GAPDH mrna were detected via Northern blotting. The amount of Nrf2 mrna was normalized to GAPDH mrna. The ratios of Nrf2/GAPDH mrna are expressed under the blots and represent the mean SE (n 3). **P.01 versus C34 cells. (B) Stability of Nrf2 mrna. C34 or E47 cells (1 10 7 cells in 100-mm dishes) were treated with 10 g/ml actinomycin D for 0, 4, 8, 16, and 24 hours. Nrf2 mrna was detected via Northern blotting. Nrf2 mrna was quantified and the amount of Nrf2 mrna at 0 hours was assigned a value of 100%. Data are expressed as the mean SE (n 3). (C) Nrf2 mrna transcription is increased in E47 cells. C34 and E47 cells (1 10 8 cells) were collected and Nrf2 mrna transcription was determined via nuclear run-on assay as described in Materials and Methods. The amount of labeled Nrf2 mrna was normalized to labeled GAPDH mrna. The ratios of Nrf2/GAPDH mrna are expressed under the blots and represent the mean SE (n 3). **P.01 versus C34 cells. Nrf2, nuclear factor erythroid 2-related factor 2; mra, messenger RNA.

HEPATOLOGY, Vol. 43, No. 1, 2006 GONG AND CEDERBAUM 149 Fig. 6. CYP2E1 inhibitors and ROS scavengers reduce ROS and block the increase of Nrf2 mrna by CYP2E1. C34 and E47 cells (1 10 7 cells in 100-mm dishes) were untreated (labeled as C) or treated with 5 mmol/l 4-methylpyrazole, 211 mmol/l dimethylsulfoxide, or 5 mmol/l N-acetyl cysteine for 2 days. (A) ROS production was assayed via flow cytometry with 2 7 -dichlorofluorescin diacetate. (B) Nrf2 and GAPDH mrna were detected via Northern blotting. The amount of Nrf2 mrna was normalized to the GAPDH mrna and the results are expressed as fold induction compared with control C34 cells. The fold induction of untreated C34 cells was taken as 1. Data are expressed as the mean SE (n 3). **P.01. ***P.001 versus control E47 cells without any treatment. ROS, reactive oxygen species; 4-MP, 4-methylpyrazole; DMSO, dimethylsulfoxide; NAC, N-acetyl cysteine; Nrf2, nuclear factor erythroid 2-related factor 2; mrna, messenger RNA. that the transcription of Nrf2 mrna is approximately 2.3-fold higher in E47 cells than in C34 cells (Fig. 5C), suggesting that the increased Nrf2 mrna level in E47 cells is caused by increased Nrf2 mrna transcription. Inhibitors of CYP2E1 and the ROS Scavenger N- Acetyl Cysteine Block the Increase in Nrf2 mrna in E47 Cells. To examine whether the increase of Nrf2 in E47 cells was caused by an increase in ROS production, the effect of CYP2E1 inhibitors such as 4-methylpyrazole and dimethylsulfoxide 30 and the ROS scavenger N-acetyl cysteine on Nrf2 mrna levels was determined. Control E47 cells showed an elevated basal ROS production compared with control C34 cells (Fig. 6A). Incubation with 4-methylpyrazole, or dimethylsulfoxide, or N-acetyl cysteine for 2 days significantly inhibited the increase in ROS production in E47 cells, whereas these compounds did not significantly affect ROS production in C34 cells (Fig. 6A). In a similar manner to the inhibition of CYP2E1 activity and decrease in ROS production, the increase of Nrf2 mrna levels in E47 cells was also prevented by these inhibitors (Fig. 6B). Ethanol Treatment Increases Nrf2 Protein Level in E47 Cells. To determine whether ethanol treatment affects Nrf2 protein expression, C34 and E47 cells were treated with 100 mmol/l ethanol for 1, 2, and 3 days and Nrf2 protein level in the cell lysates was detected via Western blot analysis. Nrf2 protein levels were unchanged in C34 cells, but the already elevated levels of Nrf2 protein in the E47 cells could be modestly increased further after ethanol treatment for 3 days (Supplementary Fig. 2). sirna-nrf2 Blocks the Induction of GCLC and HO-1 Expression in E47 Cells. To study whether upregulation of the antioxidant genes GCLC, HO-1, and catalase in E47 cells is mediated by Nrf2, SiRNA-Nrf2 was used to block the effects of Nrf2. The nontarget sirna, sirna-control, was used as a control. The transfection efficiency of sirna-nrf2 and sirna-control in C34 and E47 cells was similar (50.1 4.6%). After transfection of sirna-nrf2, Nrf2 mrna levels were decreased in C34 cells, and more dramatically in E47 cells compared with C34 or E47 cells transfected with sirnacontrol (Fig. 7A). GCLC and HO-1 mrna levels also showed some decreases in C34 cells when transfected with sirna-nrf2. Importantly, the increased GCLC and HO-1 mrna expression in E47 cells was completely blocked after transfection with sirna-nrf2 (Fig. 7A). Although catalase mrna was induced in E47 cells, sirna-nrf2 had no effect on catalase mrna levels in both C34 and E47 cells (Fig. 7A), indicating some specificity in the actions of Nrf2 on antioxidant genes. Nrf2, GCLC, and HO-1 protein levels were decreased in a time-dependent manner in C34 cells (Fig. 7B) and more dramatically in E47 cells (Fig. 7C) by sirna-nrf2. The decline in GCLC and HO-1 proteins parallels the decline in Nrf2 protein in both cell lines, suggesting an association between Nrf2 levels and these two antioxidant proteins. Catalase protein level in C34 and E47 cells transfected with sirna-nrf2 was unchanged compared with sirna-control (Fig. 7B-C). CYP2E1 protein levels were unchanged by transfection of sirna-control or sirna-nrf2 (data not shown). SiRNA-Nrf2 Decreases GSH and Increases ROS and Lipid Peroxidation in E47 Cells. Because sirna- Nrf2 blocks the induction of some antioxidant genes (e.g., GCLC and HO-1) by CYP2E1, the effects of sirna-nrf2 on the content of GSH, ROS levels, and lipid peroxidation were determined. E47 cells transfected with sirnacontrol had higher levels of GSH compared with C34 cells transfected with sirna-control (Fig. 8A), consistent with upregulation of the rate-limiting enzyme in GSH synthesis, GCLC. sirna-nrf2 did not significantly decrease GSH levels in C34 cells but significantly lowered GSH levels in E47 cells compared with C34 or E47 cells transfected with sirna-control (Fig. 8A). The ROS level in E47 cells transfected with sirna-control was higher

150 GONG AND CEDERBAUM HEPATOLOGY, January 2006 than C34 cells transfected with sirna-control. sirna- Nrf2 did not significantly increase ROS levels in C34 cells but significantly increased ROS levels in E47 cells (Fig. 8B). Lipid peroxidation was determined via TBARS assay. The TBARS level in E47 cells transfected with sirna-control was higher than in C34 cells transfected with sirna-control. sirna-nrf2 did not significantly increase TBARS level in C34 cells, but significantly increased the TBARS level in E47 cells (Fig. 8C). Fig. 8. sirna-nrf2 decreases GSH levels and increases ROS and lipid peroxidation in E47 cells. C34 and E47 cells (2.5 10 5 cells/well in 6-well plates) were transfected with sirna-control or sirna-nrf2 according to the method described in Materials and Methods. Cells were collected 30 hours after transfection, and (A) GSH, (B) ROS, and (C) lipid peroxidation were determined according to the methods described in Materials and Methods. Results are expressed as the mean SE (n 3). **P.01. ***P.001 compared with C34 cells transfected with sirna-control. ### P.001 compared with E47 cells transfected with sirna-control. GSH, glutathione; sirna, small interfering RNA; Nrf2, nuclear factor erythroid 2-related factor 2; ROS, reactive oxygen species; TBARS, thiobarbituric acid reactive species. Fig. 7. sirna-nrf2 blocks the induction of GCLC and HO-1 mrna and protein expression in E47 cells. C34 and E47 cells (2.5 10 5 cells/well in 6-well plates) were transfected with sirna-control or sirna-nrf2 according to the method described in Materials and Methods. (A) Cells were collected 30 hours after transfection. Nrf2, GCLC, HO-1, catalase, and GAPDH mrna were detected via Northern blotting and quantified. The amount of Nrf2, GCLC, HO-1, and catalase mrna was normalized to the GAPDH mrna, and the results are expressed as fold induction compared with C34 cells transfected with sirna-control. Data are expressed as the mean SE (n 3). *P.05. ***P.001 versus C34 cells transfected with sirna-control. ### P.001 versus E47 cells transfected with sirna-control. (B,C) Cells were collected at 0, 1, 2, and 3 days after transfection. Nrf2, GCLC, HO-1, and catalase proteins in the cell lysates were detected via Western blotting and quantified. The amount of each protein in C34 cells transfected with sirna-control at day 0 was assigned a value of 1. Data are expressed as the mean SE (n 3). *P.05. **P.01. ***P.001 versus corresponding protein level at day 0. mrna, messenger RNA; Nrf2, nuclear factor erythroid 2-related factor 2; GCLC, the catalytic subunit of glutamate cysteine ligase; HO-1, heme oxygenase 1. SiRNA-Nrf2 Decreases the Mitochondrial Membrane Potential in E47 Cells. Damage to the mitochondria and decreases in mitochondrial membrane potential are targets of CYP2E1-oxidative stress. 18,31 Mitochondrial membrane potential was assayed via flow cytometry after double-staining with rhodamine 123 and propidium iodide. As shown in Supplementary Fig. 3, most of the C34 and E47 cells transfected with sirnacontrol appear on the low propidium iodide and high rhodamine 123 fluorescence field (lower right quadrant), indicative of intact, viable cells with high mitochondrial membrane potential. sirna-nrf2 did not change the flow cytometry graph pattern of C34 cells, because the percentage of cells in the M1 low mitochondrial membrane potential zone was not altered, but it decreased the mitochondrial membrane potential of E47 cells as the number of cells in the M1 zone increased from 5.52% to 17.34% (P.01). The cells with lower mitochondrial membrane potential are still viable (low propidium iodide staining), suggesting that the loss in potential caused by sirna-nrf2 occurs before the loss of cell viability shown below. SiRNA-Nrf2 Decreases the Cell Viability of E47 Cells. What are the cosequences of the decrease in GSH and certain antioxidant mrnas, the increase in ROS production and lipid peroxidation, and the decrease in the

HEPATOLOGY, Vol. 43, No. 1, 2006 GONG AND CEDERBAUM 151 mitochondrial membrane potential on the viability of E47 cells? C34 and E47 cells were transfected with sirna-control or sirna-nrf2. After transfection, ROS level and cell viability was determined at 0, 1, 2, and 3 days. sirna-nrf2 only had a small tendency to increase the ROS level and decrease the viability of C34 cells (P.05), but significantly increased the ROS level and decreased the viability of E47 cells in a time-dependent manner (P.01) compared with C34 and E47 cells transfected with sirna-control (Supplementary Fig. 4A- B). There appears to be a close time dependence between the increase in ROS production caused by sirna-nrf2 and the decrease in E47 cell viability. Because the efficiency of transfection with sirna was about 50%, it is tempting to speculate that the 50% E47 cells that remained viable 3 days after sirna-nrf2 treatment are those cells that did not take up the sirna; however, this requires further evaluation (e.g., via cell sorting and assaying Nrf2 levels). Discussion Nrf2 regulates many important antioxidant genes and phase II detoxification genes. Oxidative stress plays an important role in alcoholic injury. 32-33 However, little is known about changes of Nrf2 in alcoholic liver disease. The present study shows that in chronic ethanol-fed mice or rats, liver and hepatocyte Nrf2 protein and mrna expression is increased. This increase of Nrf2 may be caused by induction of CYP2E1 by alcohol, although future experiments using inhibitors of CYP2E1 and CYP2E1 null mice are planned to study this. Induction of CYP2E1 by injecting pyrazole into rats or mice also increased Nrf2 protein expression. CYP2E1 levels were elevated 2.5- to threefold in mice and rats after pyrazole treatment. These increases were associated with similar increases in Nrf2 protein levels. As mentioned above for ethanol treatment, future experiments using inhibitors of CYP2E1 and CYP2E1 null mice are planned to validate the role of CYP2E1 in these actions of pyrazole treatment. Pyrazole treatment alone can induce oxidant stress in these mice and rats as shown by increased protein carbonyl formation, malondialdehyde levels, and in situ superoxide production. 34-35 In addition, Nrf2 protein and mrna levels are increased in HepG2 cells overexpressing CYP2E1 but not in HepG2 cells transfected with vector plasmid control. Ethanol treatment further increases the Nrf2 protein level in E47 cells but not C34 cells. Compared with C34 cells, overexpression of CYP2E1 in E47 cells causes an increase in nuclear levels of Nrf2 protein and increased Nrf2-ARE binding activity. The Nrf2-regulated antioxidant genes GCLC and HO-1, which have ARE in their promoter regions, 9 are also upregulated in E47 cells. These results demonstrate that Nrf2 is activated by increased expression of CYP2E1 in HepG2 cells. Levels of Nrf2 were also elevated in the livers of mice and rats chronically fed ethanol, and it is interesting to speculate that this increase may play a role in the elevated mrna levels of HO-1 and GCL caused by chronic alcohol treatment. 21,36 Increased expression of CYP2E1 in HepG2 cells causes an increase in Nrf2 mrna and Nrf2 protein. This increase in Nrf2 protein levels could be due to increased protein expression or decreased protein degradation. Previous studies 9 have reported that upregulation of Nrf2 gene expression at the transcriptional level and increased Nrf2 protein stability are important mechanisms for the activation of Nrf2. We found that degradation of Nrf2 protein in E47 cells is slightly decreased compared with C34 cells, suggesting that an increase of Nrf2 protein stability in E47 cells may contribute to the increase in Nrf2 protein levels. On the other hand, Nrf2 mrna transcription is also increased by CYP2E1, but the degradation of Nrf2 mrna is not changed compared with the control C34 cells. Thus, the increase of Nrf2 protein expression and stability both contribute to increased Nrf2 protein levels in E47 cells. The increase in Nrf2 protein levels may be the main reason for the activation of Nrf2 by CYP2E1. Increased ROS production caused by CYP2E1 16-17 (Fig. 6A) possibly mediates the increase of Nrf2 protein and mrna by CYP2E1. The inhibitors of CYP2E1 activity (4-methylpyrazole and dimethylsulfoxide) and the ROS scavenger N-acetyl cysteine effectively decreased ROS levels and also significantly blocked the increase of Nrf2 mrna in E47 cells, suggesting that ROS may play an important role in mediating the effects of CYP2E1 on Nrf2 mrna and protein levels. Upregulation of GCLC and HO-1 by CYP2E1 is dependent on Nrf2. Both GCLC and HO-1 have ARE in their promoter region and can be regulated by Nrf2. 7 Previous studies have shown that GCLC, HO-1, and catalase are all upregulated by increased expression of CYP2E1. 20-22 sirna-nrf2, which blocks Nrf2, also blocked the mrna expression of GCLC and HO-1, but not catalase. These results suggest that induction of the GCLC and HO-1 genes by CYP2E1 is dependent on Nrf2, but induction of the catalase gene is not Nrf2-dependent. Nrf2 is important in the adaptive response against the increased oxidative stress caused by CYP2E1. Because Nrf2 regulates many important antioxidant genes such as GCL, HO-1, and thioredoxin reductase 1 and detoxification enzymes such as NAD(P)H-quinone oxidoreductase 1 and glutathione-s-transferase, Nrf2 upregulation by

152 GONG AND CEDERBAUM HEPATOLOGY, January 2006 CYP2E1 may be an important adaptive response against the increased oxidative stress produced by elevated CYP2E1 expression. The cell viability of E47 cells was decreased after blocking Nrf2 with sirna-nrf2. The blocking of important antioxidant gene responses by sirna-nrf2 could be one of the mechanisms responsible for this decrease in E47 cell viability. Treatment with sirna-nrf2 blocks the induction of GCLC and HO-1 caused by CYP2E1, reduces cellular GSH levels, and increases cellular ROS levels and lipid peroxidation. These actions are followed by decreases in mitochondrial membrane potential and, subsequently, cell death. GCL is the rate-limiting enzyme in the process of GSH synthesis, and HO-1 has been proven to be an important mechanism against CYP2E1-related oxidative injury. 37 In conclusion, increased expression of CYP2E1 causes an enhanced level of Nrf2 protein and mrna. Elevated ROS production as a result of increased CYP2E1 activity possibly mediates these increases in Nrf2 and Nrf2 activation. Activated Nrf2 upregulates important antioxidant genes and detoxification enzymes, therefore helping to maintain cellular protective/survival pathways. The increase of Nrf2 protein and mrna levels in the liver of mice or rats fed ethanol may be caused by the induction of CYP2E1 by ethanol. Further in vivo studies are needed to elucidate the importance of Nrf2 in alcoholic oxidative injury. Acknowledgment: We thank Dr. Natalia Nieto, Dr. Yongke Lu, and Dr. Xiaodong Wang (Mount Sinai School of Medicine) for providing the samples of ethanolfed rats, pyrazole-treated rats, and pyrazole-treated mice. References 1. Talalay P, Dinkova-Kostova AT, Holtzclaw WD. Importance of phase 2 gene regulation in protection against electrophile and reactive oxygen toxicity and carcinogenesis. Adv Enzyme Regul 2003;43:121-134. 2. Primiano T, Sutter TR, Kensler TW. Antioxidant-inducible genes. Adv Pharmacol 1997;38:293-328. 3. Friling RS, Bensimon A, Daniel V. Xenobiotic-inducible expression of murine glutathione S-transferase Ya subunit gene is controlled by an electrophile-responsive element. Proc Natl Acad Sci U S A 1990;87:6258-6262. 4. Rushmore TH, Morton MR, Pickett CB. The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity. J Biol Chem 1991;266: 11632-11639. 5. Alam J, Camhi S, Choi AM. Identification of a second region upstream of the mouse heme oxygenase-1 gene that functions as a basal level and inducer-dependent transcription enhancer. J Biol Chem 1995;270:11977-11984. 6. Mulcahy RT, Wartman MA, Bailey HH, Gipp JJ. Constitutive and betanaphthoflavone-induced expression of the human gamma-glutamylcysteine synthetase heavy subunit gene is regulated by a distal antioxidant response element/tre sequence. J Biol Chem 1997;272:7445-7454. 7. Nguyen T, Sherratt PJ, Pickett CB. Regulatory mechanisms controlling gene expression mediated by the antioxidant response element. Annu Rev Pharmacol Toxicol 2003;43:233-260. 8. Itoh K, Wakabayashi N, Katoh Y, Ishii T, Igarashi K, Engel JD, et al. Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding to the amino-terminal Neh2 domain. Genes Dev 1999;13:76-86. 9. Nguyen T, Yang CS, Pickett CB. The pathways and molecular mechanisms regulating Nrf2 activation in response to chemical stress. Free Radic Biol Med 2004;37:433-441. 10. Guengerich FP, Kim DH, Iwasaki M. Role of human cytochrome P-450 IIE1 in the oxidation of many low molecular weight cancer suspects. Chem Res Toxicol 1991;4:168-179. 11. Yang CS, Yoo JS, Ishizaki H, Hong JY. Cytochrome P450IIE1: roles in nitrosamine metabolism and mechanisms of regulation. Drug Metab Rev 1990;22:147-159. 12. Koop DR. Oxidative and reductive metabolism by cytochrome P450 2E1. FASEB J 1992;6:724-730. 13. Castillo T, Koop DR, Kamimura S, Triadafilopoulos G, Tsukamoto H. Role of cytochrome P-450 2E1 in ethanol-, carbon tetrachloride- and iron-dependent microsomal lipid peroxidation. HEPATOLOGY 1992; 16:992-996. 14. Morimoto M, Zern MA, Hagbjork AL, Ingelman-Sundberg M, French SW. Fish oil, alcohol, and liver pathology: role of cytochrome P450 2E1. Proc Soc Exp Biol Med 1994; 207: 197-205. 15. Nanji AA, Zhao S, Sadrzadeh SM, Dannenberg AJ, Tahan SR, Waxman DJ. Markedly enhanced cytochrome P450 2E1 induction and lipid peroxidation is associated with severe liver injury in fish oil-ethanol-fed rats. Alcohol Clin Exp Res 1994;18:1280-1285. 16. Kessova I, Cederbaum AI. CYP2E1: biochemistry, toxicology, regulation and function in ethanol-induced liver injury. Curr Mol Med 2003;3:509-518. 17. Caro AA. Cederbaum AI. Oxidative stress, toxicology, and pharmacology of CYP2E1. Annu Rev Pharmacol Toxicol 2004;44:27-42. 18. Sakurai K, Cederbaum AI. Oxidative stress and cytotoxicity induced by ferric-nitrilotriacetate in HepG2 cells that express cytochrome P450 2E1. Mol Pharmacol 1998;54:1024-1035. 19. Wu D, Cederbaum AI. Ethanol cytotoxicity to a transfected HepG2 cell line expressing human cytochrome P4502E1. J Biol Chem 1996;271: 23914-23919. 20. Mari M, Cederbaum AI. CYP2E1 overexpression in HepG2 cells induces glutathione synthesis by transcriptional activation of gamma-glutamylcysteine synthetase. J Biol Chem 2000;275:15563-15571. 21. Gong P, Cederbaum AI, Nieto N. Increased expression of cytochrome P450 2E1 induces heme oxygenase-1 through ERK MAPK pathway. J Biol Chem 2003;278:29693-29700. 22. Mari M, Cederbaum AI. Induction of catalase, alpha, and microsomal glutathione S-transferase in CYP2E1 overexpressing HepG2 cells and protection against short-term oxidative stress. HEPATOLOGY 2001;33: 652-661. 23. Wu D, Cederbaum AI. Glutathione depletion in CYP2E1-expressing liver cells induces toxicity due to the activation of p38 mitogen-activated protein kinase and reduction of nuclear factor-kappab DNA binding activity. Mol Pharmacol 2004;66:749-760. 24. Chen Q, Cederbaum AI. Cytotoxicity and apoptosis produced by cytochrome P450 2E1 in Hep G2 cells. Mol Pharmacol 1998;53:638-648. 25. Fraga CG, Leibovitz BE, Tappel AL. Halogenated compounds as inducers of lipid peroxidation in tissue slices. Free Radic Biol Med 1987;3:119-123. 26. Bai J, Rodriguez AM, Melendez JA, Cederbaum AI. Overexpression of catalase in cytosolic or mitochondrial compartment protects HepG2 cells against oxidative injury. J Biol Chem 1999;274:26217-26224. 27. He CH, Gong P, Hu B, Stewart D, Choi ME, Choi AM, et al. Identification of activating transcription factor 4 (ATF4) as an Nrf2-interacting protein. Implication for heme oxygenase-1 gene regulation. J Biol Chem 2001;276:20858-20865. 28. Gong P, Ogra Y, Koizumi S. Inhibitory effects of heavy metals on transcription factor Sp1. Ind Health 2000;38:224-227. 29. Gong P, Stewart D, Hu B, Vinson C, Alam J. Multiple basic-leucine zipper proteins regulate induction of the mouse heme oxygenase-1 gene by arsenite. Arch Biochem Biophys 2002;405:265-274.

HEPATOLOGY, Vol. 43, No. 1, 2006 GONG AND CEDERBAUM 153 30. Feierman DE, Cederbaum AI. Role of cytochrome P-450 IIE1 and catalase in the oxidation of acetonitrile to cyanide. Chem Res Toxicol 1989;2:359-366. 31. Marí M, Bai J, Cederbaum AI. Adenovirus-mediated overexpression of catalase in the cytosolic or mitochondrial compartment protects against toxicity caused by glutathione depletion in HepG2 cells expressing CYP2E1. J Pharmacol Exp Ther 2002;301:111-118. 32. Nordmann R, Ribiere C, Rouach H. Implication of free radical mechanisms in ethanol-induced cellular injury. Free Radic Biol Med 1992;12: 219-240. 33. Tsukamoto H, Lu SC. Current concepts in the pathogenesis of alcoholic liver injury. FASEB J 2001;15:1335-1349. 34. Lu Y, Wang X, Cederbaum AI. Lipopolysaccharide-induced liver injury in rats treated with the CYP2E1 inducer pyrazole. Am J Physiol Gastrointest Liver Physiol 2005;289:G308-G319. 35. Wang X, Lu Y, Cederbaum AI. Induction of cytochrome P450 2E1 increases hepatotoxicity caused by Fas agonistic Jo2 antibody in mice. HEPA- TOLOGY 2005;42:400-410. 36. Lu SC, Huang Z, Yang JM, Tsukamoto H. Effect of ethanol and high-fat feeding on hepatic gamma-glutamylcysteine synthetase subunit expression in the rat. HEPATOLOGY 1999;30:209-214. 37. Gong P, Cederbaum AI, Nieto N. Heme oxygenase-1 protects HepG2 cells against cytochrome P450 2E1-dependent toxicity. Free Radic Biol Med 2004;36:307-318.