NADPH oxidase-derived free radicals are key oxidants in alcohol-induced liver disease.

In North America, liver disease due to alcohol consumption is an important cause of death in adults, although its pathogenesis remains obscure. Despite the fact that resident hepatic macrophages are known to contribute to early alcohol-induced liver injury via oxidative stress, the exact source of free radicals has remained a mystery. To test the hypothesis that NADPH oxidase is the major source of oxidants due to ethanol, we used p47(phox) knockout mice, which lack a critical subunit of this major source of reactive oxygen species in activated phagocytes. Mice were treated with ethanol chronically, using a Tsukamoto-French protocol, for 4 weeks. In wild-type mice, ethanol caused severe liver injury via a mechanism involving gut-derived endotoxin, CD14 receptor, production of electron spin resonance-detectable free radicals, activation of the transcription factor NF-kappaB, and release of cytotoxic TNF-alpha from activated Kupffer cells. In NADPH oxidase-deficient mice, neither an increase in free radical production, activation of NF-kappaB, an increase in TNF-alpha mRNA, nor liver pathology was observed. These data strongly support the hypothesis that free radicals from NADPH oxidase in hepatic Kupffer cells play a predominant role in the pathogenesis of early alcohol-induced hepatitis by activating NF-kappaB, which activates production of cytotoxic TNF-alpha.

The role of Kupffer cell oxidant production in early ethanol-induced liver disease.

Considerable evidence for a role of Kupffer cells in alcoholic liver disease has accumulated and they have recently been shown to be a predominant source of free radicals. Several approaches including pharmacological agents, knockout mice, and viral gene transfer have been used to fill critical gaps in understanding key mechanisms by which Kupffer cell activation, oxidant formation, and cytokine production lead to liver damage and subsequent pathogenesis. This review highlights new data in support of the hypothesis that Kupffer cells play a pivotal role in hepatotoxicity due to ethanol by producing oxidants via NADPH oxidase.

Mechanisms of alcohol-induced hepatotoxicity: studies in rats.

Alcohol treatment results in increases in the release of endotoxin from gut bacteria and membrane permeability of the gut to endotoxin, or both. Females are more sensitive to these changes. Elevated levels of endotoxin activate Kupffer cells to release substances such as eicosanoids, TNF-alpha and free radicals. Prostaglandins increase oxygen uptake and most likely are responsible for the hypermetabolic state in the liver. The increase in oxygen demand leads to hypoxia in the liver, and on reperfusion, alpha-hydroxyethyl free radicals are formed which lead to tissue damage in oxygen-poor pericentral regions of the liver lobule.

Catalase-dependent ethanol metabolism in vivo in deermice lacking alcohol dehydrogenase.

Pathways of ethanol elimination in alcohol dehydrogenase (ADH)-positive and -negative deermice were studied using the catalase inhibitor, 3-amino-1,2,4-triazole. To verify that aminotriazole inhibited catalase effectively, the characteristic decrease in catalase-H2O2 which occurs in saline-treated controls when ethanol is peroxidized was monitored at 660-640 nm in perfused deermouse livers. Following 1.5 hr of pretreatment with aminotriazole (1.5 g/kg), the peroxidatic activity of catalase measured in vitro was inhibited by greater than 99%. Under these conditions, ethanol did not decrease catalase-H2O2 in perfused livers, indicating that catalase was inhibited. Ethanol and aniline oxidation by microsomes were also inhibited by about 67-90% after 1.5 hr of pretreatment with aminotriazole. In ADH-positive deermice, pretreatment with aminotriazole for 1.5 hr prior to injection of ethanol (2.0 g/kg) decreased rates of ethanol elimination in vivo from 13.2 +/- 0.8 to 10.2 +/- 0.4 mmoles/kg/hr. In ADH-negative deermice, similar treatment decreased rates of ethanol elimination in vivo from 4.5 +/- 0.4 to 1.1 +/- 0.6 mmoles/kg/hr. Following pretreatment with aminotriazole (1.0 g/kg) for 6 hr, rates of ethanol elimination in ADH-negative deermice returned to near basal values. Under these conditions, the peroxidatic activity of catalase measured in vitro and the ethanol-dependent decrease in catalase-H2O2 in perfused livers also returned to near basal levels; however, the oxidation of ethanol by cytochrome P-450 was inhibited completely. It is concluded, therefore, that time of pretreatment with aminotriazole is an important variable which must be controlled carefully to inhibit catalase completely. Since catalase was active while cytochrome P-450 was not following 6 hr of pretreatment with aminotriazole, it is concluded that ethanol elimination occurs predominantly via catalase-H2O2 in ADH-negative deermice under these conditions.

Induction of peroxisomes by treatment with perfluorooctanoate does not increase rates of H2O2 production in intact liver.

Increases in acyl coenzyme A (CoA) oxidase activity due to peroxisome proliferation are postulated to cause oxidative stress via elevated production of H2O2, leading to DNA damage. These changes are suspected to be responsible for tumor formation caused by non-genotoxic carcinogens which do not bind to DNA but cause proliferation of peroxisomes. However, the activity of the peroxisomal enzyme acyl CoA oxidase assayed in vitro in the presence of excess fatty acyl CoA substrate may not reflect rates of H2O2 generation in intact liver where fatty acid supply is carefully controlled in part by delivery of substrate. The purpose of this work was to determine if rates of hepatic H2O2 generation were altered in perfused liver and in vivo following induction of H2O2-generating acyl CoA oxidase activity. Injection of the potent peroxisome proliferating agent perfluorooctanoate into rats 5 days prior to sacrifice caused an expected 4-fold increase of H2O2-generating acyl CoA oxidase activity measured in hepatic homogenates. In contrast, rates of H2O2 generation in perfused liver measured spectrophotometrically (660-640 nm) through a lobe of the liver were not altered by perfluorooctanoate treatment (7.3 +/- 1.5 vs. 7.8 +/- 0.5 mumol/g/h in livers from untreated control rats). Similar treatment with perfluorooctanoate also increased in vitro acyl CoA oxidase activity 9-fold in livers from deermice; however, rates of elimination of methanol, a selective substrate for catalase in rodents whose oxidation is limited by the supply of H2O2, were not altered significantly in vivo (control, 110 +/- 11 mumol/g/h vs. perfluorooctanoate, 112 +/- 32 mumol/g/h). Taken together, these data demonstrate that elevation of H2O2 formation by acyl CoA oxidase activity measured in vitro is not necessarily associated with increases in rates of H2O2 generation in intact perfused liver or in vivo, most likely due to rate-limitation in intact cells by fatty acid supply. These data do not support the hypothesis that the induction of peroxisomes leads to excessive H2O2 production and oxidative stress. It follows that alternative hypotheses to explain carcinogenesis caused by peroxisome-proliferating agents need to be considered.

Role of hormones in the mechanism of the swift increase in alcohol metabolism in the rat.

Gastric intubation of female Sprague-Dawley rats (80--150 g) with one large dose (5 g/kg) of ethanol doubled both hepatic oxygen uptake and ethanol metabolism within 2.5 hr in the perfused rat liver (Swift Increase in Alcohol Metabolism--SIAM). Hepatic oxygen uptake could also be elevated by direct infusion of epinephrine and glucagon into the perfused liver. Alcohol treatment produced significant increases in circulating epinephrine, norepinephrine and glucose but did not effect levels of plasma immunoreactive insulin. Administration of alpha- and beta-adrenergic blocking agents, adrenalectomy and hypophysectomy prevented the increase in oxygen uptake due to ethanol treatment. These data suggest that catecholamines and possibly other hormones play an important role in the mechanism of the Swift Increase in Alcohol Metabolism (SIAM).

The swift increase in alcohol metabolism (SIAM) in four inbred strains of mice.

Ethanol metabolism increases two to three hours after the administration of ethanol. This phenomenon, called the Swift Increase in Alcohol Metabolism (SIAM), has been compared in four inbred strains of mice (DBA/2J; C3H/HeJ; AKR/J; C57BL/6J). Basal rates of ethanol elimination were determined in individual mice after intraperitoneal injections of ethanol. Little variability in this basal rate of ethanol elimination was observed within each strain. Mice were then exposed to ethanol vapor for 4 hours. In both injected and treated mice the dose of ethanol was varied to produce blood ethanol levels ranging from 50 to 250 mg%. Ethanol elimination increased maximally 1.5 to 4-fold in all four strains following 4 hours of vapor treatment at the same blood ethanol level; however, the dose at which the maximal increase occurred differed among the strains. DBA/2J mice exhibited a maximal increase in the rate of ethanol elimination when ethanol concentrations were in the range of 30 to 50 mg%; the increase was smaller as the dose was increased. In contrast, AKR/J and C57BL/6J mice required 100 to 150 mg% ethanol to activate SIAM. These data indicate clearly that the SIAM effect is a common phenomenon, and that dose-response relations differ in various inbred strains of mice.

Involvement of hormones in the swift increase in alcohol metabolism.

The purpose of this study was to determine the time course of changes in blood levels of various hormones in C57BL/6J mice during exposure to ethanol vapor. Groups of adult male mice were given 2.0 g per kg ethanol intraperitoneally or as continuous vapor for four hours and rates of ethanol elimination were measured. In parallel, blood samples were collected at timed intervals over 5.5 hours during and following exposure to ethanol. Blood levels of epinephrine, norepinephrine, corticosterone, and glucagon were elevated two- to four-fold during ethanol treatment and declined to basal values within one hour following termination of treatment. Elevated blood levels of epinephrine, norepinephrine and corticosterone were highly correlated with higher rates of ethanol elimination (r = 0.80, 0.78, and 0.72, respectively). In contrast, thyroxine and insulin levels were not affected by ethanol. These findings are consistent with the idea that acute administration of ethanol causes the release of glycogenolytic hormones which in turn increase rates of ethanol metabolism.

4-Methylpyrazole inhibits fatty acyl coenzyme synthetase and diminishes catalase-dependent alcohol metabolism: has the contribution of alcohol dehydrogenase to alcohol metabolism been previously overestimated?

Alcohol dehydrogenase (ADH)-deficient deer mice were used as an animal model to investigate the effect of 4-methylpyrazole on alcohol metabolism. After intraperitoneal dosing of these mutant mice with 4-methylpyrazole, rates of ethanol and methanol metabolism in vivo were decreased significantly, by 41% and 35%, respectively. In perfused liver, rates of ethanol metabolism were also decreased up to 61% by 100 microM 4-methylpyrazole. Further, when livers were perfused with methanol, a selective substrate for catalase, rates of methanol metabolism were decreased by 64% by 4-methylpyrazole. It was further determined that 4-methylpyrazole administration caused negligible changes in total hepatic catalase activity and in rates of oxidation of ethanol by isolated microsomes; rather, it acts on catalase-dependent alcohol metabolism by limiting the supply of H2O2. In this study, 4-methylpyrazole inhibited fatty acyl CoA synthetase competitively in liver homogenates. Fatty acyl CoA synthetase is a key enzyme involved in the supply of substrate for peroxisomal oxidation of alcohols via catalase-H2O2. When palmitate was studied, rates of formaldehyde production from methanol were reduced competitively by 4-methylpyrazole; however, when the product palmitoyl CoA was used, the addition of 4-methylpyrazole did not alter activity. 4-Methylpyrazole also inhibited fatty acyl CoA synthetase activity measured directly from CoA disappearance. These data indicate that fatty acyl CoA synthetase is inhibited by 4-methylpyrazole, thus reducing the availability of H2O2 for catalase-dependent alcohol metabolism. Inhibition of methanol metabolism in deer mice expressing ADH indicates that this phenomenon also occurs in species with ADH. Taken together, these data support the hypothesis that the contribution of ADH to alcohol metabolism may have been previously overestimated.

In vivo formation of a free radical metabolite of ethanol.

Free radical metabolism of ethanol has been suggested as a factor in its hepatotoxicity. Although evidence of lipid radical formation due to ethanol treatment in vivo has been reported, free radicals from ethanol itself have not been detected in living animals. However, by applying the EPR spectroscopy technique of spin trapping to the study of ethanol-treated alcohol dehydrogenase-deficient deermice (Peromyscus maniculatus), we have detected the alpha-(4-pyridyl-1-oxide)-N-t-butylnitrone (POBN)/alpha-hydroxyethyl radical adduct in bile from animals administered [1-13C]ethanol and the spin trap POBN. Hyperfine coupling constants were aN = 15.48, a beta H = 2.02, and a beta 13C = 4.61 G. In addition, an ethanol-dependent but 13C-invariant radical adduct, presumably lipid derived, was detected. Hyperfine coupling constants were aN = 15.38 and a beta H = 2.5 G. This report demonstrates, for the first time, the in vivo formation of the alpha-hydroxyethyl free radical metabolite of ethanol.

Inactivation of Kupffer cells prevents early alcohol-induced liver injury.

It is well recognized that consumption of alcohol leads to liver disease in a dose-dependent manner; however, the exact mechanisms remain unclear. Hypoxia subsequent to a hypermetabolic state may be involved; therefore, when it was observed recently that inactivation of Kupffer cells prevented stimulation of hepatic oxygen uptake by alcohol, the idea that Kupffer cells participate in early events that ultimately lead to alcohol-induced liver disease became a real possibility. The purpose of this study was to test that hypothesis. Male Wistar rats were exposed to ethanol continuously by means of intragastric feeding for up to 4 weeks using the model developed by Tsukamoto and French. In this model, ethanol causes fatty liver, necrosis and inflammation--changes characteristic of alcohol-induced liver disease in human beings. Kupffer cells were inactivated by twice weekly treatment with gadolinium chloride (GdCl3), a selective Kupffer cell toxicant. AST levels were elevated to 192 +/- 13 and 244 +/- 56 IU/L in rats exposed to ethanol for 2 and 4 wk, respectively (control value, 88 +/- 7). This injury was prevented almost completely by GdCl3 treatment. Fatty changes, inflammation and necrosis were also all reduced dramatically by GdCl3 treatment. The average hepatic pathological score of rats treated with ethanol for 4 wk was 4.3 +/- 0.6, which was reduced significantly in ethanol- and GdCl3-treated rats to 1.8 +/- 0.5 (p < 0.05). Rates of ethanol elimination were elevated 2- to 3-fold in rats exposed to ethanol for 2 to 4 wk. This elevation was blocked by GdCl3 treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

New, simple models to evaluate zone-specific damage due to hypoxia in the perfused rat liver: time course and effect of nutritional state.

Models were developed to study zone-specific damage in periportal and pericentral regions of the liver lobule due to hypoxia produced in the perfused liver by ischemia, nitrogen or perfusion with low flow followed by reflow. Damage was assessed by lactate dehydrogenase release and trypan blue uptake in specific regions. Perfusion for up to 120 min under the conditions employed in all models failed to damage liver from well fed rats. In contrast, perfusion of livers from fasted rats for 30 min with N2-saturated buffer produced dye uptake of 37% and 66% in periportal and pericentral regions, respectively. Damage tended to be greater in this model when calcium was omitted from the perfusate (69% and 88% staining of periportal and pericentral regions, respectively). Release of lactate dehydrogenase correlated well with the percentage of cells stained with dye. In livers from fasted rats, 90 min of low flow (ca. 1 ml/g/min) followed by 30 min of reflow at normal flow rates (ca. 4 ml/g/min) produced damage exclusively to pericental regions of the liver lobule. On the average, about 40% of hepatocytes were stained with the dye under these conditions. Sixty minutes of ischemia followed by 13 min of reflow produced damage in 12% of periportal and 32% of pericentral regions of the liver lobule. When perfusion was in the retrograde direction (60 min low flow, 30 min reflow), periportal areas were damaged but pericentral regions were spared. Thus, models have been developed to study zone-specific damage due to hypoxia in the perfused liver. The data indicate that nutritional status is an important determinant of damage to hepatocytes due to hypoxia.

Chronic ethanol treatment induces H2O2 production selectively in pericentral regions of the liver lobule.

Chronic treatment with ethanol damages pericentral regions of the liver selectively, and reactive oxygen species such as H2O2 may be involved in the mechanism of hepatotoxicity. To test this idea, the effect of chronic treatment with ethanol on rates of H2O2 production was measured in tissue cylinders isolated from periportal and pericentral regions of livers from ethanol-treated rats. Rates of hydrogen peroxide production, assessed from the oxidation of methanol to formaldehyde by catalase-H2O2, were similar in tissue cylinders isolated from periportal regions in control and ethanol-treated rats. In contrast, rates of H2O2 production were over 4-fold higher in tissue isolated from pericentral regions of livers from ethanol-treated than control animals (1.7 +/- 0.5 vs. 0.4 +/- 0.3 nmol/min/mg protein, respectively). Rates of H2O2-generating acyl CoA oxidase activity were equivalent in tissue cylinders from periportal regions of livers from both groups (approximately 2 nmol/min/mg protein), but were over 2-fold higher in tissue cylinders from pericentral regions of livers from ethanol-treated rats than from controls. In contrast, catalase activity was increased nearly 2-fold in homogenates from both periportal and pericentral regions by ethanol treatment while glutathione peroxidase activity was decreased significantly in both regions. These data demonstrate that ethanol increases H2O2 generation in pericentral regions of the liver lobule in part by elevating rates of peroxisomal beta-oxidation of acyl CoA compounds and are consistent with the hypothesis that local increases in H2O2 production may be involved in the mechanism of ethanol-induced hepatotoxicity.

Toll-like receptor 4 is involved in the mechanism of early alcohol-induced liver injury in mice.

Chronic alcohol administration increases gut-derived endotoxin in the portal blood, which activates Kupffer cells and causes liver injury. Mice (C3H/HeJ) with mutations in toll-like receptor 4 (TLR4) are hyporesponsive to endotoxin. To test the hypothesis that TLR4 is involved in early alcohol-induced liver injury, the long-term intragastric ethanol feeding protocol developed by Tsukamoto and French for rats was adapted to mice. Animals with nonfunctional TLR4 and wild-type mice (C3H/HeOuJ) were compared. Two-month-old female mice were fed a high-fat liquid diet with either ethanol or isocaloric maltose-dextrin as control continuously for 4 weeks. There was no difference in mean urine alcohol concentrations between the groups. Dietary alcohol significantly increased liver-to-body weight ratios and serum alanine transaminase (ALT) levels in wild-type mice (109 +/- 18 U/L) over high-fat controls (40 +/- 3 U/L), effects that were blunted significantly in mice with a mutation of TLR4 (55 +/- 9 U/L). While no significant pathologic changes were observed in high-fat controls, dietary ethanol caused steatosis, mild inflammation, and focal necrosis in wild-type animals (pathology score = 5.2 +/- 1.2). These pathologic changes were significantly lower in TLR4-deficient mice fed ethanol (score = 2.0 +/- 1.3). Endotoxin levels in the portal vein were increased significantly after 4 weeks in both groups fed ethanol. Moreover, ethanol increased tumor necrosis factor alpha (TNF-alpha) mRNA expression in wild-type, but not in TLR4-deficient, mice. These data are consistent with the hypothesis that Kupffer cell activation by endotoxin via TLR4 is involved in early alcohol-induced liver injury.

ICAM-1 is involved in the mechanism of alcohol-induced liver injury: studies with knockout mice.

To test the hypothesis that leukocyte infiltration mediated by intercellular adhesion molecule (ICAM)-1 is involved in early alcohol-induced liver injury, male wild-type or ICAM-1 knockout mice were fed a high-fat liquid diet with either ethanol or isocaloric maltose-dextrin for 4 wk. There were no differences in mean urine alcohol concentrations between the groups fed ethanol. Alcohol administration significantly increased liver size and serum alanine aminotransferase levels in wild-type mice over high-fat controls, effects that were blunted significantly in ICAM-1 knockout mice. Dietary ethanol caused severe steatosis, mild inflammation, and focal necrosis in livers from wild-type mice. Furthermore, livers from wild-type mice fed ethanol showed significant increases in the number of infiltrating leukocytes, which were predominantly lymphocytes. These pathological changes were blunted significantly in ICAM-1 knockout mice. Tumor necrosis factor (TNF)-alpha mRNA expression was increased in wild-type mice fed ethanol but not in ICAM-1 knockout mice. These data demonstrate that ICAM-1 and infiltrating leukocytes play important roles in early alcohol-induced liver injury, most likely by mechanisms involving TNF-alpha.

Acute alcohol produces hypoxia directly in rat liver tissue in vivo: role of Kupffer cells.

Previous studies using liver slices and isolated perfused rat liver have suggested that ethanol causes hypoxia by increasing oxygen consumption. However, ethanol also increases blood flow to the liver, a phenomenon that may counteract the effects of hypermetabolism by increasing oxygen delivery. Thus whether ethanol causes hypoxia in vivo remains unclear. To clarify this important point, female Sprague-Dawley rats (100-125 g) simultaneously received pimonidazole (120 mg/kg ip), a 2-nitroimidazole hypoxia marker, and one large dose of ethanol (5 g/kg ig), which increase hepatic oxygen uptake dramatically and elevate ethanol metabolism (swift increase in alcohol metabolism) in 2-3 h. After 2 h, ethanol significantly increased the accumulation of bound pimonidazole in pericentral regions of the liver lobule. Treatment of animals with the Kupffer cell-specific toxicant, GdCl3 (10 mg/kg iv, 24 h before experiment), blocked ethanol-induced increases in pimonidazole binding. It is concluded that one large dose of ethanol causes pericentral hypoxia in rat liver tissue in vivo and that Kupffer cells are involved.

Nimodipine, a dihydropyridine-type calcium channel blocker, prevents alcoholic hepatitis caused by chronic intragastric ethanol exposure in the rat.

It has been shown recently that inactivation of Kupffer cells prevents free radical formation and early alcohol-induced liver injury, and that hypoxia subsequent to a hypermetabolic state caused by activated Kupffer cells is likely involved in the mechanism. Calcium is essential for the activation of Kupffer cells, which contain L-type voltage-dependent Ca2+ channels. Therefore, the purpose of this study was to determine whether a Ca2+ channel blocker, nimodipine, prevents early alcohol-induced liver injury in vivo and to evaluate its effect on intracellular calcium ([Ca2+]i) in Kupffer cells in vitro. Male Wistar rats were exposed to ethanol (10-12 g/kg/d) continuously for up to 4 weeks via intragastric feeding using an enteral model developed by Tsukamoto and French. In this model, ethanol causes steatosis, necrosis, and inflammation in only a few weeks. In the experimental group, nimodipine (10 mg/kg/d) was added to the diet and was shielded from direct light. Nimodipine had no effect on body weight over a 4-week treatment period, nor were mean ethanol concentrations or their cyclic pattern in urine affected. The mean urine ethanol values were 154 +/- 11 mg/dL in ethanol-fed and 144 +/- 38 mg/dL in ethanol + nimodipine-fed rats. After 4 weeks, serum aspartate transaminase (AST) levels were elevated in ethanol-treated rats to 183 +/- 78 U/L. In contrast, values only reached 101 +/- 9 U/L in rats given nimodipine + ethanol-values which were significantly lower. Steatosis and necrosis assessed histologically were also reduced significantly by nimodipine. Nimodipine (10 micrograms/kg) also blocked the swift increase in alcohol metabolism and elevated oxygen consumption in perfused livers from rats treated with alcohol in vivo. Further, in cultured Kupffer cells, nimodipine (1 mumol/L) largely prevented the elevation in [Ca2+]i caused by lipopolysaccharide (LPS) (LPS, 200 +/- 11 nmol/L; LPS + nimodipine, 94 +/- 31 nmol/L; P < .05). These results indicate that nimodipine prevents alcoholic hepatitis, possibly by inhibition of endotoxin-mediated Kupffer cell activation.

Antibiotics prevent liver injury in rats following long-term exposure to ethanol.

BACKGROUND/AIMS: Kupffer's cells participate in alcohol-induced liver injury, and endotoxemia is observed in human alcoholics and in a rat model. This study evaluated the effect of reducing bacterial endotoxin production by intestinal sterilization on alcohol-induced liver injury. METHODS: Male Wistar rats were exposed to ethanol continuously for up to 3 weeks via intragastric feeding. The gut was sterilized with polymyxin B and neomycin. RESULTS: Fecal culture of stool samples from ethanol-fed rats treated with antibiotics showed virtually no growth of gram-negative bacteria. Endotoxin levels of 80-90 pg/mL in plasma of ethanol-fed rats were reduced to < 25 pg/mL by antibiotics. Antibiotic treatment also completely prevented elevated aspartate aminotransferase levels and significantly reduced the average hepatic pathological score in rats exposed to ethanol. Oxygen tension on the surface of the liver measured in vivo was decreased significantly from control values of 48 +/- 1 to 39 +/- 1 mumol/L in ethanol-treated rats. This hypoxia was prevented by treatment with antibiotics. Moreover, the increase in rates of ethanol elimination due to long-term ethanol treatment was prevented by antibiotic treatment. CONCLUSIONS: Intestinal sterilization prevented alcohol-induced liver injury in the rat, supporting the idea that hypermetabolism and consequent hypoxia caused by activation of Kupffer's cells by endotoxin is involved in the mechanism.