High-Level Expression of Rat Class I Alcohol Dehydrogenase Is Sufficient for Ethanol-Induced Fat Accumulation in Transduced HeLa Cells

Transcription

1 High-Level Expression of Rat Class I Alcohol Dehydrogenase Is Sufficient for Ethanol-Induced Fat Accumulation in Transduced HeLa Cells ANDREA GALLI, 1 DONNA PRICE, 2 AND DAVID CRABB 2 The mechanisms by which ethanol causes fatty liver are complex. Reducing equivalents generated during ethanol oxidation inhibit tricarboxylic acid cycle activity and fatty acid oxidation. In addition, ethanol inhibits lipoprotein export and increases fatty acid uptake and lipid peroxidation. To test the role that alcohol metabolism by alcohol dehydrogenase (ADH) has on cellular lipid metabolism, a cell line expressing rat ADH was generated by transducing HeLa cells with an ADH-expressing retrovirus. The cells expressed high levels of ADH protein and had ADH activity similar to that of liver. Exposure of the cells to 20 mmol/l ethanol for 24 hours led to substantial accumulation of free fatty acids and triacylglycerol in the transduced, but not wild-type, HeLa cells. The rate of synthesis of saponifiable lipid was increased significantly by ethanol under these conditions. Ethanol exposure also promoted triacylglycerol accumulation when the cells were incubated with linoleic acid. This was associated with a decrease in the rate at which the cells oxidized 1-[ 14 -C]-linoleic acid. Fat accumulation was not prevented by including -tocopherol in the medium, arguing against a role for lipid peroxidation. However, the presence of methylene blue completely prevented the fat accumulation. This was associated with a return of the elevated lactate/pyruvate ratio toward normal. These data suggest that generation of reducing equivalents by ADH was sufficient to cause fat accumulation in this cell model. (HEPATOLOGY 1999;29: ) Abbreviations: NAD, nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide; ADH, alcohol dehydrogenase; DPPD, N,N -diphenyl-pphenylenediamine; CYP2E1; cytochrome P4502E1. From the 1 Department of Clinical Pathophysiology, University of Florence, Florence, Italy; and 2 Departments of Medicine and Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN. Received May 11, 1998; accepted January 6, Supported by the NIAAA (AA06434, AA to D.W.C.), and the Indiana Alcohol Research Center (P ). Address reprint requests to: David Crabb, Emerson Hall 312, 545 Barnhill Drive, Indianapolis, IN dcrabb@iupui.edu; fax: (317) Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/0 Accumulation of fat in the liver is one of the earliest manifestations of heavy alcohol consumption. This occurs in well-nourished individuals and does not necessarily lead to more severe forms of liver injury. Short-term studies on isolated hepatocytes or perfused liver have shown that ethanol reduced the rate of oxidation by 25% to 35%, 1 and by unknown mechanisms stimulated fatty acid uptake. Inhibition of oxidation correlated reasonably well with the ratio of -hydroxybutyrate to acetoacetate in the cell medium. This ratio reflects the intramitochondrial ratio of reduced nicotinamide adenine dinucleotide to nicotinamide adenine dinucleotide (NADH/NAD ). Thus, ethanol oxidation by alcohol dehydrogenase (ADH) is believed to cause a shift in the cytosolic NADH/NAD ratio, which in turn increases the NADH/NAD ratio in the mitochondria. Because many of the enzymes of fatty acid oxidation and the tricarboxylic acid cycle are pyridine nucleotide dependent, their activities are inhibited by NADH, resulting in reduced ability to oxidize fatty acids and acetyl-coa. 2 Longer-term studies using cultured rat hepatocytes have confirmed the reduction in oxidation, stimulation of fatty acid uptake, and also demonstrated a reduction in very-low-density lipoprotein secretion by ethanol-treated hepatocytes. 3 An alternative hypothesis was supported by the work of DiLuzio, who showed that rats administered alcohol as a single large dose developed triacylglycerol accumulation in the liver that could be reduced by prior treatment with -tocopherol or the antioxidant, N,N -diphenyl-p-phenylenediamine (DPPD). 4 DiLuzio extended that work to show that ethanol administration increased the formation of thiobarbituric acid reactive substances and malondialdehyde, both of which changes were reversed by treatment with DPPD. Finally, they showed that parenteral administration of DPPD prevented the fatty liver caused by daily gavage alcohol feeding. 5 The mechanism for the ethanol effect on lipid peroxidation was not defined, but DiLuzio showed that pretreatment of the animals with the ADH inhibitor, pyrazole, prevented the development of fatty liver after a single large dose of ethanol, implicating ethanol metabolism via ADH in the genesis of fatty liver. 6 Moreover, Muller and Sies reported that ethanol stimulated the production of ethane and pentane (end-products of lipid peroxidation) in perfused rat liver. 7 This effect was seen with low levels of ethanol (the halfmaximal effect was seen at 0.5 mmol/l ethanol, near the K m for rat ADH 8 ) and was blocked by 4-methylpyrazole, further implicating the ADH pathway. They went on to show that acetaldehyde at low concentrations ( µmol/l) also stimulated ethane and pentane formation. 7 Furthermore, recent experiments indicate that metabolism of ethanol via ADH increases the level of reactive oxygen species in isolated hepatocytes or perfused liver. 9,10 These experiments are difficult to interpret because of the presence of alternative pathways of ethanol metabolism in the liver and the lack of absolutely specific inhibitors. Cytochrome P4502E1 (CYP2E1) is known to generate reactive oxygen species such as hydroxyethyl radical, 11 which could 1164

2 HEPATOLOGY Vol. 29, No. 4, 1999 GALLI, PRICE, AND CRABB 1165 initiate lipid peroxidation, and can be inhibited by 4-methylpyrazole. 12 The peroxisomal oxidation pathway, a source of hydrogen peroxide, is also inhibited by 4-methylpyrazole at the fatty acyl-coa synthetase step. 13 One approach to avoiding these problems is to create cells expressing only a single alcohol-metabolizing enzyme, and test the effect of ethanol on that cell line. Cederbaum et al. created a cell line derived from HepG2 cells that expressed high levels of CYP2E1. The presence of this enzyme activity conferred sensitivity of the cells to cytotoxic effects of ethanol 12 and acetaminophen, 14 but the effect of ethanol on fat metabolism in the cells has not been reported. ADH-expressing cell lines have also been developed. 15,16 A CHO cell line expressing mouse ADH 16 was used to examine long-term effects of ethanol exposure. A sixfold increase in cellular triacylglycerol was observed when the cells were cultured in 20 mmol/l ethanol for 8 days Methylpyrazole blocked this effect of ethanol. The authors concluded that this was the result of acetaldehyde accumulation, and, indeed, acetaldehyde levels in the medium were over 400 µmol/l. They did not attempt to differentiate the effect of acetaldehyde, which itself is known to inhibit fatty acid oxidation 18 from effects of the redox state. We have now constructed a HeLa cell line that expresses rat ADH at high levels and tested the effect of ethanol on lipid metabolism in these cells. Our results suggest that ADH-mediated ethanol metabolism is sufficient to cause the retention of lipid in the short-term because of altered redox state. MATERIALS AND METHODS Materials. Most chemicals were from Sigma Chemical Corp. (St. Louis, MO). Tissue culture reagents were from GIBCO/BRL (Gaithersburg, MD). Silica gel H thin-layer chromatography plates were from Analtech, Newark, DE. Western blots were developing using the Amersham (Arlington Heights, IL) ECL chemiluminescence kit. 1-[ 14 C]-linoleic acid (55 mci/mmol) was from Amersham. 3 H 2 O (100 mci/ml) was from ICN (Costa Mesa, CA). Generation of the ADH-Expressing Cell Line. HeLa cells were obtained from the American Type Culture Collection. Rat ADH cdna was cloned into the vector plncx, a retroviral vector driving expression of the cloned cdna from the cytomegalovirus promoter and conferring resistance to G Transducing retrovirus was generated in the packaging cell line, PA317, and used to transduce HeLa cells. 20 After selection of G418-resistant cells, the clones were expanded and screened for ADH expression by Western blotting. Cell Culture. The cells were grown in Dulbecco-modified minimal essential medium supplemented with 5% charcoal-stripped fetal bovine serum, 63 µg/ml penicillin, and 100 µg/ml streptomycin. The cells were cultured in the presence of 20 mmol/l ethanol (with other medium additions as noted) in an incubator containing a large reservoir of 20 mmol/l ethanol in water to prevent evaporative loss. After 24 hours of treatment, the cells were scraped and lysed by sonication. Cellular lipids were extracted with chloroform:methanol (2:1, vol/vol), dried under nitrogen, and redissolved in toluene for thin-layer chromatography. 21 Analytical Methods. ADH activity was measured spectrophotometrically at 37 C in buffer consisting of 0.25 mol/l Tris Cl, 10 mmol/l ethanol, and 2 mmol/l NAD. 8 Catalase was assayed according to Van Remmen. 22 Lactate and pyruvate were determined enzymatically in the cell medium after a heat-treatment step to inactivate any lactate dehydrogenase present in the medium. 23,24 Methylene blue, which can non-enzymatically oxidize NADH in the assays, was removed by treatment of the medium with Florisil. Western blotting was performed by previously described methods, 20 using rabbit anti-serum generated against recombinant human ADH (a generous gift of Dr. William Bosron and the Indiana Alcohol Research Center). The filters were developed using the Amersham ECL chemiluminescence kit. Thin-layer chromatography used a moving phase of chloroform:methanol:acetic acid (80:20:1, vol/vol/vol). The amount of lipid applied to the plate was normalized to the amount of protein in the cell homogenate from which the lipid was extracted. The amount loaded was that equivalent to 1 mg of cell extract protein. The lipids were visualized by charring. 21 Medium ethanol and acetaldehyde concentrations were determined by gas chromatography using headspace sampling of the medium. The rate of lipid synthesis was estimated from the fate of incorporation of 3 H 2 O, which labels the NADPH pool and thus newly synthesized lipids. 25 Cells were trypsinized, counted, and then cultured in 2 ml of medium in a sealed Erlenmeyer flask (containing 95% oxygen:5% CO 2 ) for 4 hours in the presence of 3 H 2 O (added at 150 µci/ml). The cells were killed with perchloric acid, and the lipids extracted as described. 25 Both saponifiable (mainly triacylglycerol) and nonsaponifiable (mainly cholesterol) fractions were saved and counted in a scintillation counter. There was no observable quenching. Rates of synthesis are expressed as cpm per million cells per hour. The rate of fatty acid oxidation was measured by incubating the cells in 2 ml of medium as described above in the presence of 100 µmol/l linoleic acid (183 cpm/nmol) for 4 hours. The cells were killed by injection of perchloric acid through a serum cap. 14 CO 2 released from the medium during 1 hour of shaking at 37 C was collected in a hanging cup containing Hyamine hydroxide. The Hyamine was then counted in a scintillation counter. Linoleic acid oxidation was expressed as cpm per million cells per hour. RESULTS Creation of ADH-Expressing HeLa Cells. HeLa cells were chosen for these experiments because they lack low K m ADH and aldehyde dehydrogenase. 20 Although these cells were not tested for the presence of CYP2E1, this cytochrome is not known to be expressed in any cell line except for the Fao hepatoma line. 26 The catalase activity of the cells was estimated to be µmol/min/mg protein. This is considerably lower than the activity in rat liver homogenates ( 8,000 µmol/min/mg protein). 22 Moreover, because HeLa FIG. 1. Western blot of HeLa cells transduced with ADH-expressing retrovirus. The cell extracts were fractionated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and blotted to nitrocellulose. The proteins were detected with antibody raised against human liver ADH. Lane 1 represents 30 µg of cellular protein from untransduced, parental HeLa cells; lane 2 is 5 µg of rat liver homogenate protein; lanes 3 through 9 represent 30 µg of cellular protein extracted from G418-resistant HeLa cell clones.

3 1166 GALLI, PRICE, AND CRABB HEPATOLOGY April 1999 FIG. 2. Lipid content of HeLa and HeLa- ADH cells. The cells were cultured for 24 hours in the absence ( ) or presence ( ) of 20 mmol/l ethanol before the cells were harvested and the lipids extracted for thinlayer chromatography. The lane labeled standards represents a mixture of the indicated lipids that were used to identify the mobility of the different classes of lipids, not the identity of the lipids extracted from the cells. cells were derived from a cervical carcinoma, they lack the ability to synthesize apolipoproteins or ketone bodies. Several ADH-expressing clones were isolated from the HeLa cells transduced with retroviruses carrying the rat ADH cdna (Fig. 1). No ADH protein was detected in the parental HeLa cells (lane 1). Rat liver homogenate (lane 2) served as positive control for ADH protein (note that one sixth as much protein was loaded in this lane as in the others). Lanes 3 through 9 represented G418-resistant clones. The cell line represented in lane 9 was used for subsequent experiments and is hereafter referred to as HeLa-ADH cells. ADH activity in the parental HeLa cells was not detectable, while that in the HeLa-ADH cells was 20 µmol/min/g cell protein, similar to that observed in rat liver (21 µmol/min/g protein). 27 There was no apparent effect of high-level expression of ADH in the HeLa cells on cell growth rate. Effect of Ethanol on Cellular Metabolism in HeLa-ADH Cells. When the cells were cultured with 10 mmol/l ethanol, ethanol disappeared from the medium at a rate of approximately 11 µmol/h per flask (using 75-cm 2 tissue culture flasks containing cells/flask) over the first 6 hours. Medium acetaldehyde under these conditions was below 40 µmol/l at 2, 4, and 6 hours. Acetate accumulated in the medium to mmol/l at the end of 6 hours. The HeLa cells have very low background activity of low K m aldehyde dehydrogenases, 20 so it is not known by which metabolic pathway the acetaldehyde generated from ethanol is converted to acetate. The effect of ethanol metabolism on cellular redox state was assessed. When the medium was supplemented with 1 mmol/l pyruvate and 10 mmol/l lactate to approximate the lactate/pyruvate ratio in normal liver, 28 the addition of ethanol increased the ratio from 10 to 44 (n 3) after 6 hours. Thus, the cells metabolized ethanol to acetate, with low steady-state acetaldehyde levels and increased lactate/pyruvate ratio, very similar to what occurs with isolated hepatocytes. 29 Effects of Ethanol on Lipid Accumulation and Fatty Acid Synthesis by HeLa-ADH Cells. Exposure of HeLa and HeLa-ADH cells to 20 mmol/l ethanol for 24 hours was not toxic as assessed by microscopy. The effect of ethanol treatment on lipid metabolism was monitored by thin-layer chromatography of total cellular lipids. Similar amounts of total neutral lipids were applied in each lane as indicated by similar amounts of TABLE 1. Effect of Ethanol on Lipid Synthesis, Fatty Acid Oxidation, and Cytosolic Redox State of HeLa and HeLa-ADH Cells Saponifiable HeLa Cells HeLa-ADH Cells Saponifiable Nonsaponifiable Nonsaponifiable Lipid synthesis Control 1, , Ethanol-treated (20 mmol/l) 1, , * HeLa cells HeLa-ADH cells Linoleic acid oxidation Control Ethanol-treated (20 mmol/l) * HeLa ADH cells Redox state Control Ethanol-treated (20 mmol/l) * Methylene blue (10 µmol/l) * Ethanol methylene blue * NOTE. Rates of lipid synthesis were determined by incorporaton of 3 H 2 O into saponifiable and nonsaponifiable fractions over 4 hours, and are expressed as cpm/million cells/hour. Linoleic acid oxidation was measured by the release of 14 CO 2 from 1-[ 14 C]-linoleic acid over 4 hours with a starting concentration of 100 µmol/l in the medium and is expressed as cpm/million cells/hour. Lactate and pyruvate were determined in the medium after a 4-hour incubation. *Statistically significant difference from control HeLa-ADH cells (P.05). Difference from the HeLa parent cells (P.05).

4 HEPATOLOGY Vol. 29, No. 4, 1999 GALLI, PRICE, AND CRABB 1167 cholesterol in each lane in each figure. The standards were included to show the expected mobility for each class of lipid, but this system does not identify individual lipid species. HeLa cells contained very low levels of triacylglycerol or free fatty acid (Fig. 2). After 24 hours of growth in ethanolcontaining medium, there was no change in the amount of cholesterol, cholesteryl ester, free fatty acid, or triacylglycerol in the HeLa cells. HeLa-ADH cells contained higher levels of triacylglycerol than HeLa cells in the absence of ethanol. This suggested that culture medium might be contaminated with a substrate for ADH, but ethanol was not detected in the culture medium by gas chromatographic analysis, nor was a substrate for yeast ADH (e.g., longer-chain alcohols) detected in an enzymatic assay. This effect of expression of ADH by the cells, although consistently seen, was unexplained. There was a substantial increase in the amount of fatty acid and triacylglycerol in the HeLa-ADH cells after exposure to ethanol (Fig. 2). Table 1 shows the effect of ethanol on the rates of lipid synthesis by the HeLa and HeLa-ADH cells. The rates were similar between HeLa and HeLa-ADH cells. Ethanol increased the rate of synthesis of saponifiable lipid (mainly triacylglycerol) by about 30% in the HeLa ADH cells, but had no effect on lipid synthesis in the control HeLa cells. There was a trend toward increased nonsaponifiable lipid synthesis in ethanol-treated HeLa-ADH cells, but this did not reach statistical significance. Effect of Ethanol on Fatty Acid Disposition by HeLa-ADH Cells. It was also of interest to determine the cells ability to handle exogenous fatty acids. HeLa cells accumulated fatty acids and triacylglycerol only when linoleic acid was present at an initial concentration of 100 µmol/l, and this was not influenced by the presence of ethanol (Fig. 3A). In HeLa-ADH A B FIG. 3. Lipid content of HeLa (A) and HeLa-ADH cells (B) cultured in the presence of linoleic acid and ethanol. The cells were cultured for 24 hours in the presence of the noted concentrations of linoleic acid with ( ) or without ( )20mmol/L ethanol before harvesting and lipid analysis.

5 1168 GALLI, PRICE, AND CRABB HEPATOLOGY April 1999 cells, it was difficult to tell if the presence of linoleic acid in the medium increased the accumulation of lipid, but both free fatty acids and triacylglycerol appeared to be further increased when ethanol was present (Fig. 3B). Table 1 shows the effect of ethanol on the oxidation of 1-[ 14 C]-linoleic acid by the cells. The HeLa cells oxidized linoleic acid more rapidly than did the HeLa-ADH cells. Ethanol had no effect on this rate in the HeLa cells, but suppressed the rate in HeLa-ADH cells. Thus, the accumulation of lipid in the HeLa-ADH cells may in part be the result of a reduced rate of fatty acid oxidation. Mechanism for the Increased Levels of Lipid in HeLa-ADH Cells. To ensure that the effect of ethanol on HeLa-ADH cells was caused by enzymatic activity of ADH, we examined the effect of 4-methylpyrazole on lipid levels in these cells (Fig. 4). Not only did 4-methylpyrazole prevent the effect of ethanol on fatty acid and triacylglycerol accumulation, but it reduced the amount of lipid present in the cells in the absence of ethanol. We then tested the hypotheses that increased fatty acid and triacylglycerol levels in ethanol-treated HeLa-ADH cells were the result of oxidative stress and lipid peroxidation or altered redox state of the cells. There was no effect of supplementing the medium with 100 µmol/l -tocopherol on lipid levels in control cells or cells exposed to ethanol (Fig. 5). The presence of methylene blue at 10 µmol/l had no effect on lipid levels in control cells, but completely blocked the effect of ethanol on fatty acid and triacylglycerol accumulation in the HeLa-ADH cells (Fig. 6). This suggested that the effects of ethanol resulted largely from the production of a more reduced redox state in the HeLa-ADH cells. To confirm this, the lactate/pyruvate ratio in the medium was measured (Table 1). At the end of a 4-hour incubation, the lactate/pyruvate ratio was 9.8 in the absence of ethanol, and rose fivefold with ethanol treatment. Methylene blue alone reduced the lactate/ pyruvate ratio below that seen in the control cells, and lowered the lactate/pyruvate ratio induced by ethanol, although not to the control level. As a final test, the HeLa-ADH cells were incubated for 24 hours with 100 mmol/l isopropanol, a substrate of ADH that can contribute electrons to the NADH pool, but cannot be incorporated into fatty acids well FIG. 4. Effect of 4-methylpyrazole on lipid accumulation in HeLa-ADH cells exposed to ethanol. HeLa-ADH cells were exposed to ethanol (20 mmol/l), with or without 4-methylpyrazole (100 µmol/l) for 24 hours before harvesting for lipid analysis. because it is oxidized to acetone. Isopropanol increased triacylglycerol levels in HeLa-ADH cells, but not HeLa cells (not shown). DISCUSSION The notion that alcoholic fatty liver is the result of generation of large amounts of NADH, with subsequent inhibition of tricarboxylic acid cycle and -oxidation activity, 2 has been challenged by the finding that antioxidants such as -tocopherol or DPPD could reduce fat accumulation in livers of rats given a single large dose of ethanol, 4 and prevent it in rats chronically fed alcohol. 5 Since then, additional support for the lipid peroxidation hypothesis has come from the knowledge that ethanol-inducible CYP2E1 is a source of reactive oxygen species, 9 that inhibition of CYP2E1 reduces the liver histopathology scores and fatty acid and triacylglycerol accumulation resulting from ethanol administration in the Tsukamoto-French model, 30 and that ethanol administration leads to formation of hydroxyethyl radical 31 and markedly depletes hepatic mitochondrial glutathione. 32 Because earlier work with rat liver or hepatocytes could not exclude a role for non-adh pathways, the present study was performed to determine whether the metabolism of ethanol by ADH in the absence of CYP2E1 and with only low activity of catalase was sufficient to alter cellular fat metabolism. An ADH-expressing cell line was created using transducing retroviruses to introduce the rat ADH cdna under the control of the cytomegalovirus promoter into HeLa cells. These cells appear to be a good model for studies on ethanol metabolism by ADH. HeLa-ADH cells had ADH activity similar to that found in rat liver (per milligram of cell protein). This value was several-fold higher than that reported for HepG2 cells 15 or CHO cells 16 expressing mouse ADH when the activity is normalized to enzyme assays performed at ph 7.4. Moreover, HeLa-ADH cells could oxidize ethanol in the medium, generating both acetaldehyde and acetate, and developed a more reduced cytosolic redox state in the presence of ethanol, as reflected by the increased lactate-to-pyruvate ratio in the medium. Interestingly, these cells appeared to have higher levels of triacylglycerol, but not free fatty acids, in the absence of added ethanol, and did not oxidize linoleic acid as well as the parental HeLa cells. There is no explanation for this finding at present; however, 4-methylpyrazole blocked this effect, suggesting that it is a result of expression of ADH, and not an artifact of selection of this particular HeLa-ADH clone. We found that the parental HeLa cells, which lack detectable class I ADH activity or protein, did not accumulate fatty acids nor triacylglycerol after exposure to ethanol. In contrast, cellular fatty acids and triacylglycerol were both increased in HeLa-ADH cells exposed to ethanol. Because the medium contained charcoal-stripped serum to remove free fatty acids, it was likely that this increased lipid content represented de novo fatty acid synthesis from glucose, amino acids, or acetate (derived from ethanol). This conclusion was supported by the observation of an increased rate of endogenous triacylglycerol synthesis, as measured by rates of 3 H 2 O incorporation. This increase probably does not require ethanolderived acetate, because isopropanol also increased triacylglycerol accumulation. Ethanol also stimulated the retention of fatty acids and triacylglycerol when the HeLa-ADH cells were incubated in the presence of added linoleic acid. No such effect was seen in

6 HEPATOLOGY Vol. 29, No. 4, 1999 GALLI, PRICE, AND CRABB 1169 FIG. 5. Lipid content of HeLa- ADH cells cultured with ethanol and -tocopherol. Tocopherol acetate was added at 100 µmol/l immediately before the addition of ethanol (20 mmol/l) to the medium. The cells were harvested after 24 hours for lipid analysis. HeLa cells. This can be in part explained by the observed reduction in linoleic acid oxidation. Based on prior studies, the more reduced cytosolic redox state of the cells would be predicted to inhibit fatty acid oxidation, favor the formation of glycerol-3-phosphate, and thereby promote the esterification of linoleic acid. Because HeLa cells lack the ability to package and secrete very-low-density lipoproteins, one cannot invoke inhibition of this process to account for the accumulation of fat. 3 The effect of ethanol on the generation of reactive oxygen species was not directly measured. However, the antioxidant, -tocopherol acetate at 100 µmol/l, did not prevent the FIG. 6. Lipid content of HeLa- ADH cells cultured with ethanol and methylene blue. The cells were cultured with the noted additions for 24 hours. Methylene blue was added at 10 µmol/l. Ethanol was present at 20 mmol/l where noted.

7 1170 GALLI, PRICE, AND CRABB HEPATOLOGY April 1999 accumulation of fat in the cells, whereas 10 to 25 µmol/l tocopherol prevented ethanol cytotoxicity in CYP2E1- expressing HepG2 cells. 12 On the other hand, methylene blue, which is able to accept electrons from pyridine nucleotides and transfer them nonenzymatically to oxygen, partially reversed the effect of ethanol on the redox state and completely prevented the accumulation of lipid. It is unclear why only a partial correction of the redox state apparently reversed the abnormalities of lipid metabolism, although the redox effects were measured after a 4-hour incubation, while the cellular lipids were analyzed after a 24-hour incubation. Thus, these experiments indicated that metabolism of ethanol by the ADH pathway, by altering the cellular redox state, may be sufficient to cause fatty liver. This does not exclude the interactions of multiple mechanisms for fat accumulation in vivo. First, more reduced redox state is proposed as a mechanism for generation of reactive oxygen species in the mitochondrion. 9,10 It may be that the tocopherol was not present at a high-enough concentration to reduce lipid peroxidation. Second, the lipid accumulation caused by ethanol metabolism via ADH may render the cells more susceptible to lipid peroxidation initiated by peroxisomal ethanol metabolism or CYP2E1, setting up a self-perpetuating process. This would be consistent with reports that pyrazole prevented development of fatty liver after acute exposure to ethanol, 4 whereas inhibition of CYP2E1 ameliorated, but did not prevent, fat accumulation or histopathological lesions in chronically ethanol-fed rats. 30 Finally, ethanol s effect on the redox state may initiate fatty liver, but the induction of CYP2E1, progressive depletion of mitochondrial glutathione, and worsening oxidative stress may perpetuate it during chronic ethanol consumption. This model is consistent with the observation that the hepatic redox state returns toward normal with chronic ethanol administration. 33 It will be interesting to determine the effect of chronic ethanol exposure on fat accumulation and reactive oxygen generation in the HeLa-ADH cells. REFERENCES 1. Grunnet N, Kondrup J. The effect of ethanol on the beta oxidation of fatty acids. Alcohol Clin Exp Res 1986;10:64S-68S. 2. Lieber CS, Lefevere A, Spritz N, Feinman L, DeCarli LM. Difference in hepatic metabolism of long- and medium-chain fatty acids: the role of fatty acid chain length in the production of the alcoholic fatty liver. J Clin Invest 1967;46: Grunnet N, Kondrup J, Dich J. Ethanol-induced accumulation of triacylglycerol in cultured hepatocytes: dependency on ethanol metabolism. Alcohol Alcohol 1987;Suppl 1: DiLuzio NR. Prevention of the acute ethanol induced fatty liver by anti-oxidants. Physiologist 1963;6: DiLuzio NR, Hartman AD. Role of lipid peroxidation in the pathogenesis of the ethanol induced fatty liver. Fed Proc 1969;26: Morgan JC, DiLuzio NR. Inhibition of the acute ethanol induced fatty liver by pyrazole. Proc Soc Exp Biol Med 1970;134: Muller A, Sies H. Role of alcohol dehydrogenase activity and the acetaldehyde in ethanol-induced ethane and pentane production by isolated perfused rat liver. Biochem J 1982;206: Crabb DW, Bosron WF, Li T-K. Steady-state kinetic properties of purified rat liver alcohol dehydrogenase: application to predicting alcohol elimination rates in vivo. Arch Biochem Biophys 1983;224: Kurose I, Higuchi H, Kato S, Miura S, Watanabe N, Kamegaya Y, Tomita K, et al. Oxidative stress on mitochondria and cell membrane of cultured hepatocytes and perfused liver exposed to ethanol. Gastroenterology 1997;112: Bailey SM, Cunningham CC. Acute and chronic ethanol increases reactive oxygen species generation and decreases viability in fresh, isolated rat hepatocytes. HEPATOLOGY 1998;28: Cederbaum AI. Microsomal generation of hydroxyl radicals: its role in microsomal ethanol oxidizing system (MEOS) activity and requirement foriron.annnyacadsci1987;492: Wu DC, Cederbaum AI. Ethanol cytotoxicity to a transfected HepG2 cell line expressing human cytochrome P4502E1. J Biol Chem 1996;271: Bradford BU, Forman DT, Thurman RG. 4-Methyl pyrazole inhibits fatty acyl coenzyme A synthetase and diminishes catalase-dependent alcohol metabolism: has the contribution of alcohol dehydrogenase to alcohol metabolism been previously overestimated? Mol Pharmacol 1993;43: Dai Y, Cederbaum AI. Cytotoxicity of acetaminophen in human cytochrome P4502E1-transfected HepG2 cells. J Pharm Exp Ther 1995;273: Clemens DL, Halgard CM, Miles RR, Sorrell MF, Tuma DJ. Establishment of a recombinant hepatic cell line stably expressing alcohol dehydrogenase. Arch Biochem Biophys 1995;321: Mapoles JE, Iwahashi M, Lucas D, Zimmerman BT, Simon FR. Acetaldehyde exposure causes growth inhibition in a Chinese hamster ovary cell line that expresses alcohol dehydrogenase. Alcohol Clin Exp Res 1994;18: Meskar A, Holownia A, Bardou LG, Menez JF. Effect of acetaldehyde generated from ethanol by ADH-transfected CHO cells on their membrane fatty acid profiles. Alcohol 1996;13: Cederbaum AI, Lieber CS, Rubin E. Effect of acetaldehyde on fatty acid oxidation and ketogenesis by hepatic mitochondria. Arch Biochem Biophys 1975;169: Miller AD, Rosman GJ. Improved retroviral vectors for gene transfer and expression. Biotechniques 1989;7: Xiao Q, Weiner H, Johnston T, Crabb DW. The aldehyde dehydrogenase ALDH2*2 allele exhibits dominance over ALDH2*1 in transduced HeLa cells. J Clin Invest 1995;96: Kates M. Techniques in Lipidology. Ed 2. New York: American Elsevier, 1995: Van Remmen H, Williams MD, Yang H, Walter CA, Richardson A. Analysis of the transcriptional activity of the 5 -flanking region of the rat catalase gene in transiently transfected cells and in transgenic mice. J Cell Physiol 1998;174: Noll F. Lactate. In: Bergmeyer HU, ed. Methods of Enzymatic Analysis. 3rd Edition. Basel: Verlag Chemie, 1984, pp Lamprecht W, Heniz F. Pyruvate. In: Bergmeyer HU, ed. Methods of Enzymatic Analysis. 3rd Edition. Basel: Verlag Chemie, 1984, pp Harris RA. Studies on the inhibition of hepatic lipogenesis by N-6,O-2 dibutyryl adenosine 3,5 -monophosphate. Arch Biochem Biophys 1975; 169: Zhukov A, Ingelman-Sundberg M. Selective fast degradation of cytochrome P-450 2E1 in serum-deprived hepatoma cells by a mechanism sensitive to inhibitors of vesicular transport. Eur J Biochem 1997;247: Bosron WF, Crabb DW, Housinger TA, Li T-K. Effect of fasting on the activity and turnover of rat liver alcohol dehydrogenase. Alcohol Clin Exp Res 1984;8: Williamson DH, Lund P, Krebs HA. The redox state of free nicotinamideadenine dinucleotide in the cytoplasm and mitochondria of rat liver. Biochem J 1967;103: Crow KE, Cornell NW, Veech RL. The rate of ethanol metabolism in isolated rat hepatocytes. Alcohol Clin Exp Res 1977;1: Morimoto K, Reitz RC, Morin RJ, Koop D, Nguyen K, Ingelman- Sundberg M, French SW. CYP2E1 inhibitors partially ameliorate the changes in hepatic fatty acid composition induced in rats by chronic administration of ethanol and a high fat diet. J Nutr 1995;125: Reinke LA, Kai EK, DuBose CM, McCay PB. Reactive free radical generation in vivo in heart and liver of ethanol-fed rats: correlation with radical formation in vitro. Proc Natl Acad SciUSA1987;84: Fernandez-Checa JC, Ookhtens M, Kaplowitz N. Effects of chronic ethanol feeding on rat hepatocytic glutathione. Relationship of cytosolic glutathione to efflux and mitochondrial sequestration. J Clin Invest 1989;83: Salaspuro MP, et al. Attenuation of the ethanol-induced hepatic redox change after chronic alcohol consumption in baboons: metabolic consequences in vivo and in vitro. HEPATOLOGY 1981;1:33-38.