Presented in part at the Digestive Disease Week 2008 in San Diego, CA (Gastroenterology 2008;134(Suppl 1):A-18).
Potential conflict of interest: Nothing to report.
Oxidative stress is thought to play a major role in the pathogenesis of hepatocellular cancer (HCC), a frequent complication of alcoholic liver disease (ALD). However, the underlying mechanisms are poorly understood. In hepatocytes of ALD patients, we recently detected by immunohistochemistry significantly increased levels of carcinogenic etheno-DNA adducts that are formed by the reaction of the major lipid peroxidation product, 4-hydroxynonenal (4-HNE) with nucleobases. In the current study, we show that protein-bound 4-HNE and etheno-DNA adducts both strongly correlate with cytochrome P450 2E1 (CYP2E1) expression in patients with ALD (r = 0.9, P < 0.01). Increased levels of etheno-DNA adducts were also detected in the liver of alcohol-fed lean (Fa/?) and obese (fa/fa) Zucker rats. The number of nuclei in hepatocytes stained positively for etheno-DNA adducts correlated significantly with CYP2E1 expression (r = 0.6, P = 0.03). To further assess the role of CYP2E1 in the formation of etheno-DNA adducts, HepG2 cells stably transfected with human CYP2E1 were exposed to ethanol with or without chlormethiazole (CMZ), a specific CYP2E1 inhibitor. Ethanol increased etheno-DNA adducts in the nuclei of CYP2E1-transfected HepG2 cells in a concentration-dependent and time-dependent manner, but not in vector mock-transfected control cells. CMZ blocked the generation of etheno-DNA adducts by 70%-90% (P < 0.01). Conclusion: Our data support the assumption that ethanol-mediated induction of hepatic CYP2E1 leading inter alia to highly miscoding lipid peroxidation–derived DNA lesions may play a central role in hepatocarcinogenesis in patients with ALD. (HEPATOLOGY 2009.)
Excess consumption of alcohol is widespread in Western countries and, in the United States, alcohol abuse affects approximately 18 million adults.1 The consumption of more than 80 g ethanol per day for more than 10 years increases the risk for hepatocellular carcinoma (HCC) by approximately fivefold.2 Worldwide, HCC is one of the most common malignant tumors, ranking as the seventh most prevalent in men and the ninth in women. The risk for HCC in decompensated alcohol-induced cirrhosis is approximately 1% to 2% per year.2 Although the evolution of alcohol-related HCC involves several factors,3, 4 including the presence of cirrhosis, oxidative stress, disturbed DNA methylation, and defective retinoic acid signaling, the precise mechanisms at a molecular level by which alcohol causes HCC are poorly understood. Recently, the International Agency for Research on Cancer has classified ethanol as a human (group 1) carcinogen, because it induces HCC (among other tumors) in animals and increases the risk for developing HCC in humans.3, 5, 6
It is widely recognized that reactive oxygen species (ROS) such as superoxide anion and hydrogen peroxide play an important role in alcohol-induced liver injury and in hepatocarcinogenesis.3, 7 Chronic ethanol consumption results in the generation of ROS through multiple pathways leading to lipid peroxidation (LPO) and LPO byproducts such as 4-hydroxy-2-nonenal (4-HNE) and malondialdehyde. These DNA-reactive aldehydes in turn form mutagenic exocyclic DNA adducts, including 1,N6-ethenodeoxyadenosine (εdA) and 3,N4-ethenodeoxycytidine (εdC).8–10
Several enzyme systems are capable of producing ROS, including the mitochondrial respiratory chain, the cytosolic enzymes xanthine oxidase, and aldehyde oxidase, as well as the microsomal cytochrome P450–dependent mono-oxygenases. Among the latter, cytochrome P450 2E1 (CYP2E1) is involved in alcohol-mediated generation of oxidative stress. Expression of CYP2E1 has been shown to correlate significantly with the generation of hydroxyethyl radicals and with LPO products such as 4-HNE and malondialdehyde.11 CYP2E1 is induced by chronic alcohol consumption within a week even at a relatively low ethanol dose (40 g/d), but the degree of CYP2E1 induction varies substantially between individuals.12 Inhibition of CYP2E1 by chlormethiazole (CMZ), a specific CYP2E1 inhibitor, improved alcoholic liver disease (ALD) as shown in the Tsukamoto-French rat model.13 An increase of oxidative DNA adducts and of mutagenic apurinic and apyrimidinic DNA sites were found in wild-type mice chronically fed ethanol, but not in mice knocked out for functional CYP2E1,14 further highlighting the importance of CYP2E1 in the generation of DNA damage after ethanol ingestion.
Recently, we have been able to detect etheno-DNA adducts such as εdA in livers of patients with ALD.15 The aim of this study is to determine whether (1) a correlation exists between CYP2E1, protein-bound 4-HNE, and the occurrence of etheno-DNA adducts in ALD in humans and rats, and (2) CYP2E1 is primarily responsible for the generation of these miscoding etheno-DNA adducts in a cell culture model overexpressing human CYP2E1.
Human liver fine-needle biopsy samples were obtained from 14 patients with ALD for diagnostic purposes at Salem Medical Center, University of Heidelberg, Germany. The liver specimens were immediately frozen in liquid nitrogen and stored at −80°C before analysis. The study was approved by the Ethics Committee of the University of Heidelberg.
Insulin-resistent, leptin-deficient Zucker rats (fa/fa) and their lean littermates (Fa+/?) with a body weight of approximately 350 g were pair fed Lieber DeCarli liquid diets containing 36% of total calories either as ethanol or isocalorically replaced by carbohydrates.16 Obese Zucker rats spontaneously develop metabolic syndrome and fatty liver, which closely resembles that occurring after chronic alcohol ingestion. Zucker rats were used here to study a situation with an extreme hepatic fat content resulting in CYP2E1 induction and oxidative stress, which may be further enhanced by chronic alcohol feeding. After 3 months of feeding these diets, animals were killed, and whole livers were removed and immediately frozen in liquid nitrogen for immunohistochemical and western blot analyses. The animal experiments were approved by the Office of Agriculture and Nature, Department of Veterinary Medicine, Division of Animal Experimentation, Bern, Switzerland.
Cell Culture and Treatment.
HepG2 cell lines, one stably transfected with the human CYP2E1 gene (E47 cells), and one transfected with a control mock vector (C34 cells)17 (provided by A. I. Cederbaum, New York, NY), were used in this study. The cell lines were grown in high-glucose Dulbecco's modified Eagle medium containing 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin. Cells were seeded at a density of 20,000 cells/mL onto Poly-L-Lysine–coated slides (Sigma-Aldrich, Munich, Germany) and cultured within multiwell slide chambers (flexiperm, Greiner Bio-One, Frickenhausen, Germany) until 75% confluence. Cells were washed twice with phosphate-buffered saline (PBS), and the medium was replaced with Dulbecco's modified Eagle medium containing 3% fetal bovine serum. After overnight culture, the medium was removed, and the cells were treated with 2.5, 5, 10, and 25 mM ethanol (diluted in Dulbecco's modified Eagle medium containing 3% fetal bovine serum) for 24, 48, or 72 hours. To prevent ethanol evaporation during treatment, each culture dish was tightly wrapped with Parafilm. All experiments were done in triplicates. At the end of incubation, slides were washed in PBS, fixed in cold acetone for 10 minutes, and air-dried. To inhibit the CYP2E1 activity in E47 cells, 20 μM CMZ18 was added to the culture medium during ethanol treatment.
Immunohistochemical Detection of Etheno-DNA Adducts in Cultured Cells.
Staining was performed on cultured cells using the method developed in our laboratory.15, 19 Acetone-fixed slides with cultured cells were dipped in PBS for 10 minutes, then placed in 0.3% H2O2 in absolute methanol for 10 minutes to quench endogenous peroxidase. Slides were incubated with proteinase K (10 μg/mL) (Roche, Mannheim, Germany) in double distilled H2O at room temperature for 10 minutes to remove histone and nonhistone proteins from DNA, increasing antibody accessibility. After washing with PBS, slides were treated with 10 μg/mL ribonuclease (RNAse; Roche, Mannheim, Germany) (heated for 10 minutes at 80°C to inactivate DNAse) at 37°C for 1 hour to prevent antibody binding to RNA adducts, and then washed in PBS. To denature DNA, cells were treated with 1.5 N HCl for 5 minutes at room temperature and subsequently rinsed in PBS and double distilled water. The pH was neutralized with 50 mM Trisbase buffer, pH 7.4, for 5 minutes at room temperature. Nonspecific binding sites were blocked with 8% bovine serum albumin, 2% horse serum, 0.05% Tween, and 0.05% Triton X-100 for 20 minutes at room temperature. Slides were incubated with the primary monoclonal antibody EM-A-1 against εdA and EM-C-1 against εdC (provided by Drs. P. Lorenz and M. Rajewsky, University of Essen, Essen, Germany), at a dilution of 1:20 at 4°C overnight.20 After washing with PBS, the antibody detection was performed using the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol. Diaminobenzidine was used as a chromogen to visualize the reaction. The reaction was stopped after 5 minutes with H2O. Slides were counterstained with 4′,6-diamidino-2-phenylindole and mounted with Kaiser's glycerin-gelatin. All slides were subjected to the same standardized conditions. Negative controls were performed by omitting the primary antibody.
Immunohistochemical Detection of Etheno-DNA Adducts in Human Liver Biopsy Specimes and in Rat Liver.
The protocol for immunohistochemical staining of etheno-DNA adducts in liver biopsy sections was similar to the protocol described for cultured cells, with three modifications: First, for DNA denaturation, 4 N HCl was used. Second, for RNAse digestion, a concentration of 20 μg RNAse/mL was used. Third, in rat tissue, a cross-adsorbed horse anti-mouse antibody, conjugated to biotin (Vector Laboratories, Burlingame, CA), was employed to prevent unspecific binding to endogenous rat immunoglobulin.
Immunohistochemical Staining of CYP2E1 and Protein-Bound 4-HNE in Human Liver Biopsy Sections.
Frozen liver biopsy samples were cut into 6-μm sections, placed on 3-aminopropyl-triethoxysilan–coated glass microscope slides, fixed in acetone at −20°C for 10 minutes, and air-dried.
Sections were treated with 0.5% H2O2 in absolute methanol for 10 minutes to quench endogenous peroxidase activity. Thereafter, sections were incubated for 2 hours at room temperature with the primary antibody (rabbit anti-human CYP2E1, 1:100, Chemicon, Hofheim, Germany) or with rabbit anti–4-HNE (1:500, Alexis, Lörrach, Germany) and 5% normal goat serum to block nonspecific binding. Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA) was used for detection according to the manufacturer's protocol. Staining was developed by incubating the sections for 5 minutes in diaminobenzidine. Sections were counterstained with hematoxylin and mounted in Aquatex mounting medium. Negative controls were performed by omitting the primary antibody.
Imaging and Semi-quantitative Analysis of Etheno-DNA Adducts, CYP2E1 Expression, and Protein-Bound 4-HNE.
Representative pictures were taken at a magnification of 200× with a Leica Image Manager 50 (Leica, Solms, Germany) and analyzed using Image J software (Toronto Western Research Institute, UK). The frequency of εdA and εdC positively stained nuclei was expressed as a percentage of stained cell nuclei over the total number of cells counted
Staining intensity for CYP2E1 and 4-HNE was assessed according to the scale devised by Tsutsumi et al.,21 whereby 3+, 2+, 1+, and 0 denote intense, moderate, slight, and no specific immunostaining, respectively.
Quantitative Determination of CYP2E1 by Western Blot in Rat Liver.
For quantitative assessment of CYP2E1 expression, 50 μg liver protein homogenate in radioimmunoprecipitation assay buffer was separated on a 10% polyacrylamide gel. Protein bands were transferred onto a nitrocellulose membrane. Equal loading was verified by Ponceau S staining followed by blocking of nonspecific protein binding with 5% nonfat milk in PBS-Tween 0.05% at room temperature for 1 hour. Rabbit-anti Cyp2E1 antibody 1:1000 in 5% nonfat milk (Chemicon, Schwalbach, Germany) was incubated overnight at 4°C. A horseradish peroxidase conjugate (goat-anti rabbit immunoglobulin–horseradish peroxidase, 1:5000 [Sigma-Aldrich, Munich, Germany]), was applied for 60 minutes at room temperature followed by washing and detection with enhanced chemiluminescence (Amersham, Buckinghamshire, UK). Beta-actin (1:5000, Sigma-Aldrich, Munich, Germany) was used as loading control. Western blots were analyzed densitometrically. Results are expressed as CYP2E1/ beta-actin ratio.
Quantitative Determination of 4- HNE–Protein Adducts in Rat Liver.
A quantitative OxiSelect HNE-His Adduct enzyme-linked immunosorbent assay kit for HNE-protein adducts detection was used (Cell Biolabs, San Diego, CA). Briefly, bovine serum albumin standards or protein samples from the liver tissues (10 μg/mL) were absorbed onto a 96-well plate overnight at 4°C. After washing, 100 μL diluted anti–4-HNE-His antibody (1:1000) was added to each well and incubated for 1 hour at room temperature, followed by 100 μL of horseradish peroxidase–conjugated secondary antibody (1:1000) for the next 1 hour. After washing, 100 μL substrate solution was added, and the reaction was stopped after 10 minutes by adding 100 μL of stop solution. The absorbance was measured at 450 nm on a microplate reader. The 4-HNE–protein adduct content in unknown samples is determined according to a standard curve prepared from predetermined 4-HNE–bovine serum albumin standards.
Results were calculated using Microsoft Excel and SPSS 12.0. Correlations are represented by Spearman's rho and a two-tailed test of significance. If appropriate, statistical significance was documented by student t test at P < 0.05. Levels of 4-HNE adducts in the four groups of obese and lean Zucker rats were compared with one-way analysis of variance.
Etheno-DNA Adducts Closely Correlate with CYP2E1 and Protein-Bound 4-HNE in Liver Biopsy Specimens of ALD Patients.
We first investigated the correlation between CYP2E1 expression, the presence of protein-bound 4-HNE, a major LPO product, and etheno-DNA adducts in human liver samples from patients with ALD. Nine male and five female patients (Table 1) aged 44 to 73 years were studied. All had a daily ethanol intake of more than 60 g and an average of 146 g/day. All patients underwent liver biopsy, and their fibrosis stage according to Kleiner et al.22 ranged between 0 and 4; two patients had liver cirrhosis.
Table 1. Patient Characteristics
Fibrosis Score (Kleiner)
NA, not available because of lack of material.
Highly significant positive correlations were observed between εdA and εdC (r = 0.97, P < 0.01) between εdA and CYP2E1 (r = 0.93, P < 0.01), and between εdC and CYP2E1 (r = 0.95, P < 0.01) (Fig. 1A-C). Furthermore, significant correlations were found between CYP2E1 and protein-bound 4-HNE (r = 0.95, P < 0.01), between εdA and protein-bound 4-HNE (r = 0.85, P < 0.01), and between εdC and protein-bound 4-HNE (r = 0.83, P < 0.01) (Fig. 1D-F). Figure 2 shows the juxtaposition of immunostaining for CYP2E1, protein-bound 4-HNE, and etheno-DNA adducts (εdA, εdC) in two representative ALD patients with low and high hepatic levels of these markers.
Etheno-DNA Adducts Correlate with CYP2E1 in a Rat Model of ALD.
We then studied whether the correlations found in human liver biopsy specimens can also be observed in liver samples of obese and lean Zucker rats after chronic alcohol administration. Hepatic CYP2E1 determined by western blot as well as εdA and εdC (not shown) determined by immunohistology were found to be significantly increased in lean (Fa+/?) as well as in obese Zucker rats (fa/fa) after chronic ethanol consumption as compared with control diets (Fig. 3A-C). However, in lean Zucker rats, DNA-adducts were almost undetectable. There was a significant correlation between the number of cells positively stained for εdA and CYP2E1 expression determined by western blots in all animals (r = 0.828, P = 0.0001). The correlation between εdA and CYP2E1 in Fa+/? rats was r = 0.63 (P = 0.027) and in fa/fa rats r = 0.775 (P = 0.003), respectively.
The 4-HNE-protein adducts were found to be significantly increased in the livers of alcohol-fed obese Zucker rats (fa/fa) compared with those receiving a nonalcoholic control diet (6.99 ± 0.82 μg/mL versus 3.42 ± 0.28 μg/mL; P < 0.001). However, there was no significant difference in hepatic 4-HNE protein adducts between lean (Fa+/?) Zucker rats with and without ethanol.
CYP2E1 Is Involved in the Formation of Exocyclic Etheno-DNA Adducts in Transfected HepG2 Cells.
To establish a causal role between CYP2E1 and etheno-DNA adduct formation, we analyzed stably transfected HepG2 cells overexpressing CYP2E1 (E47 cells). Ethanol treatment resulted in a concentration-dependent and time-dependent increase of εdA in their nuclei but not in vector mock-transfected control HepG2 (C34) cells (Fig. 4 and 5A, B). To investigate whether this increase in adduct formation is mainly attributable to CYP2E1 activity, we used CMZ, a specific CYP2E1 inhibitor. CYP2E1 was inhibited at 20 μM CMZ, a dose known from the literature to completely block CYP2E1.18 As a consequence, ethanol-induced generation of εdA was reduced by 70% to 90% (P < 0.01) at ethanol concentrations of 2.5 to 25 mM. (Fig. 6A,B).
Excess ethanol consumption has been classified as a human (group 1) carcinogen5, 6; however, the exact mechanisms by which ethanol induces HCC are unclear. In the current study, we show a significant correlation between hepatic CYP2E1 expression in ALD patients and the quantity of exocyclic etheno-adducts, DNA lesions in the human liver with known mutagenic and carcinogenic properties. These results were confirmed in a rat model of fatty liver after chronic ethanol administration and in ethanol-exposed cultured HepG2 cells that overexpress CYP2E1. Moreover, specific inhibition of CYP2E1 by CMZ markedly blocked the formation of etheno-DNA adducts.
It has been shown that both acute and short-term ethanol feeding of rodents also increased the hepatic concentrations of etheno-DNA adducts, εdA and εdC, approximately twofold.23 More recently, excess etheno-DNA adduct levels were detected by immunohistochemistry in liver biopsy specimens of patients at different stages of ALD.15 When compared with livers of nonalcoholic individuals, the prevalence of etheno-DNA adducts (εdA) was approximately 16 times higher in fibrotic and cirrhotic livers because of chronic alcohol consumption. The extent of adduct formation was comparable to that found in livers of patients with hereditary hemochromatosis, a condition in which oxidative stress attributable to iron overload leads to an extremely high risk for HCC. In contrast, analysis of healthy human and rodent liver samples by immunoaffinity-P32-postlabeling methods revealed the presence of only low background levels of these etheno-DNA adducts.24, 25 Furthermore, using a highly sensitive immunoenrichment-high-pressure liquid chromatography method for quantifying etheno-DNA adducts excreted in urine, a significantly higher amount of εdA was found in ALD patients compared with healthy subjects (H. Bartsch, J. Nair, unpublished observation).
Ethanol-induced oxidative DNA damage, as determined by the accumulation of apurinic/apyrimidinic sites in DNA correlated with an increase in the expression and activity of CYP2E1.14 Hepatic injury was found strikingly increased in transgenic mice overexpressing CYP2E1.26 In line with these data, there were fewer oxidized DNA products in CYP2E1 knockout compared with wild-type mice after challenge with ethanol, and CYP2E1-null or 1-aminobenzotriazole–treated mice were completely protected from ethanol-induced DNA damage.14 The fact that CYP2E1-null mice also developed liver injury was attributed to compensatory mechanisms, including the induction of CYP4A isozymes.27, 28 However, the role of CYP4A10 and CYP4A14 in the generation of LPO products and ROS has been controversially discussed29; Bradford et al.14 did not find increased oxidative DNA damage in CYP2E1-null mice despite elevated CYP4A enzyme levels.
Neither CYP2E1 expression nor etheno-adduct formation correlates with the amount of ethanol consumed. As already pointed out,12 the intensity of CYP2E1 expression varies substantially between alcohol consumers for yet unknown reasons. However, the low inducibility of CYP2E1 in some individuals may represent a protective factor in relation to ethanol-mediated carcinogenesis.
In addition to the data in alcoholic humans, a significant correlation between hepatic CYP2E1 and etheno-DNA-adduct formation is demonstrated in a rat model of nonalcoholic fatty liver disease, particularly when ethanol is administered chronically. This animal model of nonalcoholic fatty liver disease was chosen because nonalcoholic fatty liver disease by itself is associated with CYP2E1 induction30 and oxidative stress.31 Thus, obese Zucker rats reveal elevated hepatic CYP2E1 expression at the outset, whereas corresponding lean littermates show only faint CYP2E1 protein levels before alcohol feeding. Importantly, chronic ethanol ingestion increased hepatic CYP2E1 protein and consecutive etheno-DNA lesions further in both obese and lean animals.
CMZ, a specific CYP2E1 inhibitor, is used to treat alcohol withdrawal in patients undergoing alcohol detoxification. Gouillon and coworkers13 demonstrated that ethanol-induced liver disease in the intragastric feeding rat model was inhibited by concomitant administration of CMZ. Furthermore, a significant decrease of alanine aminotransferase, CYP2E1 activity, and protein level, as well as an alleviation of alcohol-specific histological features of liver damage, were noted in mice treated with pyrazole (a CYP2E1 inducer) when CMZ was co-administered.32 Our current data extend these findings by demonstrating that blocking human CYP2E1 activity in HepG2 cells by CMZ prevents ethanol-induced DNA damage.
It has to be emphasized that the exocyclic etheno DNA-adducts studied here are distinct from the apurinic/apyramidinic sites described by Bradford et al. Exocyclic-DNA adducts derive from the reaction of LPO-products with DNA and are believed to derive from inflammatory-driven oxidative stress reactions.33, 34 Increased accumulation of LPO byproducts such as 4-HNE and oxidized proteins are critical features in ethanol-induced liver injury. Although malondialdehyde forms a pyrimidopurinone with the N-1 and N2- position of guanine, 4-HNE induces DNA damage through its putative epoxidation product, 2,3-epoxy-4-hydroxynonanal.8, 35 Substituted 1,N2-ethenoguanine, 1,N2-ethenoguanine, 3,N4-ethenocytosine, and 1,N6-ethenoadenine have been characterized as reaction products, the most prevalent adduct being1,N2-etheno-2′-deoxyguanosine.36–38
Excess levels of etheno-adducts have been detected in chronically inflamed or infected organs of cancer-prone patients.39 Exocyclic nucleobase adducts exhibit pro-mutagenic properties8, 40 and apparently are poorly repaired in some tissues and cells,41, 42 indicating that they may be of considerable biological importance. Etheno-DNA adducts probably play a causal role in the initiation and progression of liver carcinogenesis. because they produce base pair substitution mutations; εdA can lead to AT → CG, AT → TA, and AT → CG transversions43, 44 εdC can cause CG → AT transversions and CG → TA transitions,45, 46 and N2,3-ethenodeoxyguanosine, which is also formed in vivo from LPO products, can lead to GC → AT transitions.46 These miscoding properties of etheno-bases strongly implicate them in the initiation of carcinogenesis by vinyl chloride.47 Incorporation of a single εdA in either DNA strand of HeLa cells showed a similar miscoding frequency and was more mutagenic than 8-oxo-deoxyguanosine.48 The biological impact of etheno-DNA adducts is further stressed by the observation that they are preferentially formed at the codon 249 of TP53 (which encodes p53), leading to a mutation that renders cells more resistant to apoptosis and gives them some growth advantage.49
In conclusion, we here demonstrate a causal relationship between cellular expression of CYP2E1 and the formation of carcinogenic etheno-DNA adducts in cultured cells, rat livers, and human livers exposed to ethanol. Thus, hepatic CYP2E1 induction by chronic ethanol consumption not only mediates hepatic inflammation and fibrosis but seems to be an important factor in ethanol-mediated hepatocarcinogenesis.
The authors thank all patients who participated in this study; A. I. Cederbaum, New York, NY, for providing the HepG2 cell lines; and Dr. Zili Dan for technical assistance. This article is dedicated to J. Nair, who died unexpectedly in 2007.