Early-Phase Alcoholic Liver Disease: An Update on Animal Models,Pathology,and Pathogenesis

Abstract

Alcoholic liver disease (ALD) remains to be one of the most common etiology of liver disease and is a major cause of morbidity and mortality worldwide. The pathologic stages of ALD comprises of steatosis, steatohepatitis, and fibrosis/cirrhosis. Steatosis and steatohepatitis represents the early phase of ALD and are precursor stages for fibrosis/cirrhosis. Numerous research efforts have been directed at recognizing cofactors interacting with alcohol in the pathogenesis of steatosis and steatohepatitis. This review will elucidate the constellation of complex pathogenesis, available animal models, and microscopic pathologic findings mostly in the early-phase of ALD. The role of endotoxin, reactive oxygen species, alcohol metabolism, and cytokines are discussed. Understanding the mechanisms of early-phase ALD should provide insight into the development of therapeutic strategies and thereby decrease the morbidity and mortality associated with ALD.

Alcohol abuse is one of the main causes of morbidity and mortality throughout the Western world (National Institute on Alcohol Abuse and Alcoholism [NIAAA] 2001) Although alcohol has toxic effects on numerous tissues, one of the most common target organs is liver. The pathogenesis of alcoholic liver disease (ALD) has been increasingly delineated in a number of studies (Arteel 2003; Arteel et al. 2003; Brunt 2002; Day 2000; Diehl 2002; Ishak, Zimmerman, and Ray 1991). However, relatively little is known about the actual mechanisms for the development and progression of ALD. It is a well-known fact that a complex relationship exists between the degree of alcohol consumption and risk of developing ALD (Sorensen et al. 1984). Because there is a considerable variation in individual susceptibility to the development of liver injury, factors other than alcohol abuse per se must clearly be involved in the hepatic susceptibility to alcohol. Alcoholic steatosis (fatty liver) is thought to progress gradually to steatohepatitis (steatosis coupled with inflammation), fibrosis (deposition of extracellular matrix), and cirrhosis (fibrosis with regenerating nodules) (Diehl 2002). Although fatty liver is considered to be a rapid and direct result of alcohol ingestion, the evidence showing that steatohepatitis and necrosis can develop as a result of alcohol intoxication alone is less convincing thus far.

Efforts to identify involvement of coexistent factors in the progression of ALD from steatosis to steatohepatitis are complex. Examples of coexistent factors interacting with alcohol include immunological, genetic, hormonal, or nutritional conditions, which may be involved in the progression of hepatic pathology beyond steatosis (Becker et al. 1996). It has become increasingly evident that endotoxemia, alcohol metabolism, oxidative stress, and cytokines play a key role in alcohol-mediated liver injury (Arteel 2003; Diehl 2002; Lieber 2000a, 2000b; Thurman et al. 1998). Based on the lack of answers for prevention or treatment (or both) of alcoholism, it is important to devise therapies that diminish the impact of ethanol on hepatic function. To achieve this goal, better understanding of the pathogenesis by which ethanol ingestion leads to liver disease is critical. Accomplishing this goal is possible via experiments in animal models aimed at dissecting biochemical mechanisms and relating to natural history of ethanol-induced pathology in patients. Hence, the purpose of this review is to summarize and update the findings of ALD as it relates to early pathologic events such as steatosis and steatohepatitis. Early stages of ALD and associated histologic alterations, current animal models, and pathogenetic mechanisms will be reviewed.

PATHOLOGIC STAGES OF ALCOHOLIC LIVER DISEASE

The histopathological features of ALD are very complex and variable. Virtually all patterns of microscopic histopathologic alterations may be encountered with patients with a history of chronic ethanol intake. The pathologic spectrum of human ALD (Figure 1) includes fatty liver, steatohepatitis, fibrosis, and cirrhosis. Chronic ethanol administration to rodents has been demonstrated to lead to a variety of microscopic hepatic changes, including steatosis, hepatocellular oncotic necrosis, apoptosis, inflammatory cell infiltration, terminal hepatic venular sclerosis, proliferation of the smooth endoplasmic reticulum, and mitochondrial aberrations (Figure 2) (Iseri, Lieber, and Gottlieb 1966; Lieber and DeCarli 1976; Maddrey 1995; Tsukamoto et al. 1984a, 1984b). Steatosis and steatohepatitis represent the early phase of alcoholic liver disease. Although comparable pathologic alterations such as steatosis, oncotic necrosis, apoptosis, and hepatocellular neutrophilic infiltration also occur in the early phase of human ALD (Diehl 2002; MacSween and Burt 1986), several other histopathologic changes noted in human alcoholics, such as Mallory bodies and advanced liver cirrhosis, are rarely seen in experimental animal models. As in the case with humans, it appears that alcohol, although necessary, is not sufficient to cause liver damage and other mitigating factors contribute to disease progression. The readers are strongly encouraged to refer to the section on animal models for a detailed explanation.

Alcoholic Steatosis (Fatty Liver)

The most common predicted histological change in ALD is steatosis, a rapid metabolic response to excessive alcohol consumption (Figure 2). It is a common occurrence in alcoholics: the reported incidence ranges from 10% to 90%. Fatty liver is thought to be a result of metabolic disturbances, such as decreased fatty acid oxidation, increased triglyceride synthesis, reduced fat export, and mobilization of extrahepatic fat stores (Lieber 1993, 1994; Maher 2002; Zhao et al. 2004). Fatty infiltration is both macrovesicular (one large fat droplet per hepatocyte and lateral displacement of the nucleus) (Ishak, Zimmerman, and Ray 1991) and microvesicular (several small fat droplets per hepatocyte). The term “alcoholic foamy degeneration” is synonymous with microvesicular steatosis. The hepatocytes are filled with many small fat droplets (less than 1 μm) surrounding centrally placed nuclei (Figure 2). There is no regiospecificity with regards to the fat accumulation. Variations in fat content and regiospecificity can be attributed to different animal models employed to investigate ALD. In rats, either perivenous, periportal, midzonal, or panlobular fat accumulation is observed, depending upon animal models, dietary factors, and also upon gender (Apte, McRee, and Ramaiah 2004; Thurman et al. 1998; Tsukamoto et al. 1985a). It was long considered that steatosis was a rather benign condition, due to its common occurrence and rapid disappearance upon ethanol withdrawal. Even in the most severe degree of fatty livers, the fat disappeared after 3 to 4 weeks of abstinence (Desmet 1985; MacSween and Burt 1986). In addition, the metabolic abnormalities in fatty liver were suggested to be insufficient to lead to inflammation (Lieber 1994). However, later studies have indicated that the metabolic changes taking place during the fatty changes may sensitize the cells to further injury (Bathgate and Simpson 2002; Bykov et al. 2003; Fisher et al. 2003; Galli et al. 2001; Teli et al. 1995). Thus, it is now suggested that the more fat in the liver, the higher the susceptibility to severe damage (Day and James 1998). The mechanisms for enhanced susceptibility of fatty liver to ethanol-induced liver damage have been investigated (Bathgate and Simpson 2002; Bykov et al. 2003; Maher 2002; Sorenson et al. 1984; Teli et al. 1995). Studies suggest that cellular changes taking place during the fatty metamorphosis may sensitize hepatocytes to further injury (Bathgate and Simpson 2002; Teli et al. 1995). It is known that fatty liver is highly vulnerable to oxidative stress or the injury mediated by endotoxin or cytokine action (Bykov et al. 2003; Colell et al. 1998; Yang et al. 1997). Recent investigations have suggested that decreased expression of peroxisome proliferator-activated receptors may play a role in increased sensitization of fatty hepatocytes to liver injury (Everett, Galli, and Crabb 2000; Fisher et al. 2003; Galli et al. 2001). Although these studies reflect enhanced susceptibility of steatotic hepatocytes, a recent study has shown that fatty liver resulting from ethanol (EtOH) administration for 5 weeks demonstrated enhanced capacity to regenerate and consequent decline in hepatic injury (Apte et al. 2003a, 2004). The decrease in liver injury is not due to alterations in cytochrome P450 (CYP) CYP2E1-mediated mechanisms because CYP2E1 is implicated in EtOH-mediated liver injury (see the section on ethanol metabolism). The mechanism by which EtOH enhances signal transduction for hepatocyte proliferation is possibly through nuclear factor (NF)-κB (Apte et al. 2003a, 2004). Such increased hepatocyte proliferation as an early response to hepatic injury due to chemicals is well understood (Ramaiah et al. 1998, 2000; Soni and Mehendale 1998). Additional studies are needed to identify if alcoholic steatosis represents a benign condition or a precursor to subsequent pathologic stages of ALD.

Alcoholic Steatohepatitis (ASH)

It is well known that continued ingestion of alcohol in humans results in neutrophilic steatohepatitis (Bautista 2002; Jaeschke 2002). In fact, pathological evaluations of liver tissue have consistently demonstrated the presence of neutrophils in the liver parenchyma. Neutrophil infiltration in the liver is known to significantly contribute to the pathologic findings noted in ALD (Bautista 2002; French 2002, 2003). Alcoholic steatohepatitis seldom (<10% cases) reverts to normal hepatic histology, even when the precipitating condition is removed (French 2002). Rather, patients with steatohepatitis often develop increased hepatic fibrosis and, with time, cirrhosis occurs in a substantial fraction (i.e., in almost 50%) of these individuals (Figure 1) (Galambos 1972). Thus, liver-related morbidity and mortality occur in patients with steatohepatitis and this stage appears to represent a rate-limiting step in the progression to cirrhosis and clinical liver disease in patients with ALD (Diehl 2002; Galambos 1972; Nanji 2002). The histomorphological pattern of alcoholic steatohepatitis in human alcoholics consists of the infiltration of polymorphonuclear neutrophils, hepatocyte degeneration, ballooning, and oncotic necrosis. Similar pathologic changes are noted in rodent models (Figure 2). Apoptosis is also a hallmark of alcoholic steatohepatitis. There is no direct evidence that ethanol per se causes inflammation. The presence of fat in the liver seems to be a prerequisite to the development of inflammation, possibly because a fatty liver is more vulnerable to various factors that trigger inflammation (Day and James 1998). For instance, there is evidence for an involvement of bacterial endotoxins and viral hepatitis (Apte, McRee, and Ramaiah 2003b; Sata et al. 1996; Thurman et al. 1998). Furthermore, oxidative stress induced either by dietary polyunsaturated fatty acids or by iron supplementation may aggravate the inflammation in experimental models (Nanji, Khettry, and Sadrzadeh 1994; Tsukamoto et al. 1995). A more pronounced inflammatory response is seen in livers from female rats than males, suggesting an immunomodulatory effect of estrogen (Thurman et al. 1998; Yin et al. 2000). Moreover, studies with rodents suggest that binge-type alcohol treatment enhances the inflammatory response in the liver (Enomoto et al. 1998, 1999).

Substantial progress has been made in the general understanding of neutrophil-infiltrative mechanisms in the liver (Bautista 2002; French 2002; Jaeschke 2002; Lieber 2000b). However, the precise neutrophil migration events from the hepatic vasculature to final attachment and attack on hepatocytes are not known. The role of inflammatory mediators TNF (tumor necrosis factor) α, complement, PAF (platelet-activating factor), vasoconstrictors (endothelin-1), adhesion molecules (selectins, LFA-1/Mac-1, intercellular cell adhesion molecule [ICAM]-1, vascular cell adhesion molecule [VCAM]-1), chemokines, and cytokines have been reported in literature as possible mechanisms (Bautista 2002) for neutrophil sequestration (first step in neutrophil extravasation) in hepatic vasculature. Transendothelial migration of neutrophils (second step) following sequestration is a prerequisite for hepatic cytotoxicity that involves adhesion molecules (Jaeschke and Smith 1997; Jaeschke, Farhood, and Smith 1991). The interaction between ICAM-1 on endothelial cells and its counter receptor CD11/CD18 (β₂ integrins) on neutrophils and interaction between vascular VCAM-1 and neutrophil β₂ integrins are reported to be possible mechanisms for neutrophil extravasation (Jaeschke 2002). In addition to the role of adhesion molecules, CXC chemokines and apoptotic cell death are also known to be involved in transmigration of neutrophils leading to hepatocyte cytotoxicity (Ziol et al. 2001). Finally, adherence of neutrophils (third step) to parenchymal cells involves interaction through LFA-1 (CD11a/CD18) or Mac-1 (CD11b/CD18) on neutrophils with ICAM-1, a counterreceptor on hepatocytes (Jaeschke 2002). Following attachment of neutrophils to hepatocytes, the death of the hepatocytes occurs through generation of reactive oxygen species (NADPH oxidase–derived superoxides and myeloperoxidase-derived hypochlorous acid) and proteases (cathepsin-G and elastase) by neutrophils (Nanji 2002).

Of the three steps in neutrophilic migration events, hepatic transendothelial migration (second step) of neutrophils represents a highly complex process (French 2002). Transendothelial neutrophil migration requires migration of neutrophils across the hepatic space of Disse and interstitial matrix. Neutrophils migrate directionally across interstitial tissue encountering extracellular matrix (ECM) components. Digestion of the space of Disse and interstitial matrix components by matrix metalloproteinases (MMPs) facilitates neutrophil migration (French 2002; Nanji 2002). In addition, neutrophils form sites of close contact with the ECM substrate that involves integrin adhesion to an ECM substrate. These events are currently unknown in alcoholic hepatitis. Few studies that have investigated transendothelial migration of neutrophils in alcoholic liver disease have focused mostly on adhesion molecules (Jaeschke 2002; Kono et al. 2001b). Very little is known about the actual role of ECM during alcoholic hepatitis. A novel approach in the primary author’s laboratory is to address the role of osteopontin (a matricellular protein), major hepatic ECM proteins (collagen IV, heparin sulfate proteoglycan, fibronectin), and MMPs 2, 3, and 4 in hepatic neutrophil infiltration (Apte, McRee, and Ramaiah 2003b). In fact, preliminary recent studies have shown that osteopontin may play a central role in the neutrophilic ASH (Apte, McRee, and Ramaiah 2003b). Studies such as these are important to identify the key mechanisms involved in ASH and thereby provide insights into the development of therapeutic strategies in ALD.

Alcoholic Fibrosis and Cirrhosis

Perivenular fibrosis is considered as the first irreversible step in ALD progressing into severe fibrotic changes and eventually to cirrhosis (MacSween and Burt 1986; Teli et al. 1995). The fibrotic process is characterized by a proliferation of stellate cells (HSCs) and their transformation into myofibroblasts. The pathogenesis of alcohol-associated fibrosis is complex and remains speculative. Although several molecular events are implicated in the pathogenesis of fibrosis, these mechanisms are beyond the scope of this review and will not be discussed. The readers are referred to excellent reviews in the literature (French 1993, 2003; Friedman 1993; Maher 1990; Nanji 2002; Teli et al. 1995).

ANIMAL MODELS OF ETHANOL-INDUCED LIVER INJURY

The significant variation in animal models used in the study of ALD reflects the difficult task of developing a suitable paradigm that exactly replicates the human prototype. Each model emphasizes one or more features seen in humans and in this way each model makes a contribution to some aspect of our understanding of the pathogenesis of ALD. Well-established rat models to investigate ALD include the chronic feeding model of Lieber and DeCarli (1982), the Tsukamoto and French enteral model (Tsukamoto and French 1985a, 1985b), the total enteral nutrition (TEN) model for dietary manipulations (Badger et al. 1993), and the simple rat model based on lipopolysaccharide (LPS) sensitization (Enomoto et al. 1999). Animals have been administered ethanol chronically by various methods in attempts to develop liver lesions resembling those seen in human ALD. Simple inclusion of ethanol in the drinking fluid seldom causes high and sustained elevation of blood ethanol levels and only a moderate rise in liver triglycerides is observed (Lieber and DeCarli 1989). If the concentration of ethanol in the drinking fluid is increased above a certain level, intake decreases sharply, leading to systemic dehydration, reduced food intake, and ceased or reduced growth rate. To overcome the failure of most animals otherwise to consume higher amounts of ethanol voluntarily, it has been administered in a nutritionally adequate liquid diet that provided a maximum of 35% to 40% of total calories from alcohol. This situation resembles that of many alcoholics, who often receive more than 50% of their total energy as ethanol (Patek et al. 1975; Salaspuro and Lieber 1980). Although there are several models with rodents, primates, pigs, and other small animals, this review will focus on important rodent models as it they relate to early phase alcoholic liver disease.

Oral Liquid Diets: Lieber-DeCarli Diets

Because rats have aversion to drinking alcohol in drinking water, the incorporation of ethanol in a specialized liquid diet forces rats to consume high amounts of ethanol in a balanced liquid diet that contains sufficient water and all necessary nutrients. It was developed almost four decades ago (Lieber et al. 1963) and proved to be very useful and practical in studies of the pathogenesis of the early phase of ethanol-induced hepatic changes. Controls were pair fed an equicaloric amount of diet with ethanol replaced by carbohydrates (Lieber and Decarli 1989; Table 1). The improved formula consisted of casein (providing 18% of calories) supplemented with methionine and cysteine, a mixture of dextrin and maltose (providing 11% and 47% of calories for ethanol and control diets, respectively), and fat (35% of calories, mainly olive oil, corn oil, and safflower oil). All essential vitamins (A, D, E, K, Bs), minerals, and fiber were present (Lieber and DeCarli 1986, 1989). The amount of ethanol in the diet was gradually increased during the first week to provide 36% of total calories.

This so called “Lieber-DeCarli” formula has been extensively used in rodent studies (Table 1). The average daily ethanol intake resulted in fatty liver and in metabolic tolerance (i.e., ethanol elimination rate was increased) (Lieber and DeCarli 1970b). A sixfold increase in hepatic triglycerides was observed after 1 month of feeding, an effect that persisted for 22 weeks (Lieber and DeCarli 1970a). For proper fatty liver to develop, at least 21% of the calories had to be derived from fat (Lieber and DeCarli 1970a), although even a low-fat (13%) ethanol diet causes some steatosis (Di Luzio and Hartma 1969). The incorporation pattern of dietary fatty acids in liver triglycerides indicated that most fat comes from the diet (Lieber, Spritz, and DeCarli 1966), and much less from hepatic lipogenesis (Tsukamoto et al. 1984a). Lesions beyond steatosis are rare in this model. For example, rats fed for up to 9 months on ad libitum feeding in Lieber-DeCarli liquid diet had no fibrotic changes (Leo and Lieber 1983). This probably is a consequence of the rather modest blood ethanol levels achieved with this regimen (Lieber and DeCarli 1989). Sustained high blood alcohol concentrations indeed seems to be a prerequisite for the progression of ALD process beyond steatosis (French, Morimoto, and Tsukamoto 1995; Lieber and DeCarli 1976). Although a distinct improvement over alcohol exposure in drinking water, the Lieber-DeCarli and related models of ad libitum feeding of ethanol-containing liquid diet still have some limitations. For example, despite partially overcoming their dislike of ethanol-containing diet, rodents on ethanol diet still consume less than animals on carbohydrate control diet. To address this issue, animals that are fed control diet are restricted calorically to match the alcohol-consuming group (Lieber and DeCarli 1989). Therefore, although this model employs more physiologically relevant alcohol exposure, there is an inherent concern in comparing “pair-fed” rodents that typically consume their daily caloric allowance in a short time period (e.g., 3 to 6 h) with ethanol-fed rodents that consume the same amount of calories over the course of a day.

Intragastric Ethanol Feeding Models

In both these models ethanol liquid diet is fed through a permanent indwelling intragastric catheter. By regular monitoring of blood ethanol levels, the ethanol infusion rate and the nutrient composition of the diet can be titrated (to an average of 12 to 15 g/kg/day in rats) and high blood ethanol levels are achieved. As in the oral feeding model, controls are infused with isocaloric amounts of ethanol-free diet, with carbohydrates replacing ethanol. The gastrostomy tube is usually implanted on adult rats weighing 300 to 400 g (Tsukamoto et al. 1985b). The gastrotomy tube is connected to spring coils and swivels to protect the cannulas and permit free movement of the infused animal (Tsukamoto et al. 1985b). Daily monitoring of alcohol intoxication is necessary, because the rate of ethanol infusion needs to be adjusted to achieve consistently high, yet tolerable ethanol levels. Monitoring of ethanol inebriation is by jugular blood or urine sampling and also by visual inspection of the animals (Badger et al. 1993; Yin et al. 1999). Approximately 30% to 50% of the rats on ethanol diet developed macrovesicular and microvesicular steatosis, focal necrosis, and mononuclear inflammation (French et al. 1988; Tsukamoto et al. 1985a). Early perivenous fibrogenesis starts to develop in 3 to 6 months, provided that a high-fat diet with 42% to 49% of total energy as ethanol is infused (Kamimura et al. 1992). One interesting finding with the intragastric feeding procedure is that the blood alcohol levels (BALs) fluctuate over a 5-day cycle despite a constant ethanol dose. The extent of ethanol fluctuation has been found to correlate to the degree of hepatic damage (Tsukamoto et al. 1985b). The intragastric feeding technique can also be applied to mice (Yin et al. 1999). This modification has led to numerous advances by the use of genetically altered mice (e.g., “knockout” mice) (Kono et al. 2001a, 2001b). The average daily dose in the mice peaks at ∼25 g/kg/day; the difference between these dosing regimens in rats (see above) and mice is to achieve similar urine alcohol profiles between the species. The clear advantage of the enteral feeding model to ad libitum feeding in rodents is the increase in circulating ethanol levels attainable. Specifically, with ad libitum feeding (in liquid diet), levels of alcohol higher than 100 mg/dl are rarely achieved. In contrast, circulating levels in the enteral feeding model can reach as high as 500 mg/dl, with average concentrations of ∼150 mg/dl. Although there are distinct advantages of enteral alcohol feeding, especially the ability to deliver controlled high doses of ethanol to increase pathology, disadvantages of intragastric ethanol feeding include the need for surgical manipulation, significant animal husbandry, and the relative cost of the model compared to ad libitum feeding.

Simple Nonsurgical LPS Sensitization Model

The simple, nonsurgical LPS sensitization model makes it possible to achieve liver pathology (Figure 2; steatosis, inflammation, and necrosis) that resemble alterations that occur in the enteral feeding model (Tsukamoto et al. 1985a). This is a modification of the Lieber-Decarli diet, which is known to enhance hepatic neutrophil infiltration. It is known that high blood concentrations of alcohol accompanied by elevated endotoxin levels result in significant neutrophilic infiltrate (Enomoto et al. 1999; Tamai et al. 2002). Moreover, increased concentrations of endotoxin and the endotoxin-inducible cytokines, TNF-α and interleukin (IL)-1, and other TNF-α–inducible cytokines, including IL-6 and IL-8, have been reported in patients with alcoholic steatohepatitis (Bautista 1997). Based on these reports, a modification to the in vivo protocol by incorporating LPS (endotoxin) was developed, which results in significantly enhanced hepatic neutrophil infiltration. The primary author’s laboratory has successfully employed this model to investigate the role of matricellular protein, osteopontin (OPN), in hepatic neutrophil infiltration (Apte, McRee, and Ramaiah 2003b).

MECHANISMS OF LIVER INJURY

Endotoxin and Alcoholic Liver Disease

Endotoxin (lipopolysaccharide) is a polymer found in the outer membrane of gram-negative bacteria. As a cell membrane component, endotoxin serves as a barrier against phagocytosis by cells of the host’s immune system (Nikaido 1979). Once released, this molecule is a potent stimulator of the inflammatory response during bacterial invasion. Endotoxin consists of three distinct structural components: the O-specific polysaccharide, oligosaccharide core, and the lipid portion. The O-specific polysaccharide is the most variable part of the endotoxin molecule. This portion consists of repeating oligosaccharide units arranged in a pattern distinct for each serotype of bacteria. The O-specific polysaccharide protrudes into the extracellular environment and elicits the production of anti-O antibodies by the host during bacterial invasion. The oligosaccharide core is highly conserved among bacterial serotypes (Nikaido 1979). This segment is made up of several sugars including heptose, which is only found in gram-negative bacteria and rare species of algae. The third segment of endotoxin, commonly referred to as lipid A, is linked to the core region by ketodeoxyoctanoate (KDO). This hydrophobic lipid region is embedded in the membrane and is made up of long-chain fatty acids bound to a glucosamine-disaccharide backbone. The bond between KDO of the oligosaccharide core and lipid A is easily hydrolyzed by acid and free lipid A is responsible for many of the pathological effects observed during endotoxin exposure (Rietschel et al. 1994).

Mechanism of Ethanol-Induced Endotoxemia: Role of Kupffer Cells

Gram-negative bacteria and endotoxin are found predominantly in the ileum and colon. Normally, the gut wall provides a protective barrier against the release of large amounts of endotoxin into the systemic circulation. However, chronic ethanol exposure damages intestinal villi and increases the permeability of marker compounds such as horseradish peroxidase via the paracellular route (Baraona and Lieber 1974; Worthington, Meserole, and Syrotuck 1978). Therefore, mucosal injury due to ethanol exposure could allow endotoxin to escape into the blood where it is removed by both circulating and fixed mononuclear phagocytes (Kupffer cells). Kupffer cells participate in the effects of acute as well as chronic ethanol exposure via paracrine interactions with hepatocytes. For example, Kupffer cells are the primary source of prostaglandins, which have been shown to stimulate glycogenolysis and oxygen uptake (Casteleijn et al. 1988; Kuiper et al. 1988; Qu et al. 1998). Destruction of Kupffer cells with GdCl₃ prior to the administration of ethanol completely prevented increases in oxygen uptake and ethanol metabolism due to acute ethanol treatment (Bradford, Misra, and Thurman 1993). Inactivation of Kupffer cells with the dihydropyridine-type calcium channel blocker nimodipine similarly prevented the hypermetabolic state (Iimuro et al. 1996). In chronic studies, depletion or inactivation of Kupffer cells diminished hepatic steatosis, inflammation, and necrosis, characteristic of early alcoholic hepatitis (Adachi et al. 1994; Iimuro et al. 1996), without affecting ethanol metabolism (Koop et al. 1997). Furthermore, treatment of rats with antibodies directed against TNF-α during ethanol exposure also diminished liver injury (Iimuro et al. 1997). Taken together, these findings support the hypothesis that Kupffer cells mediate alcoholic liver injury via production of toxic products such as cytokines.

CD14 Endotoxin Receptor and LPS-Binding Protein

Because blood leaving the gut empties directly into the portal vein, Kupffer cells are primarily responsible for endotoxin clearance (Wisse 1977). Although several pathways of endotoxin clearance have now been identified, Hampton and coworkers reported that scavenger receptors on macrophage are the major route of internalization and detoxification of endotoxin (Hampton et al. 1991), a process that does not trigger macrophage activation or cytokine production (Gegner, Ulevitch, and Tobias 1995; Hampton et al. 1991).

On the other hand, endotoxin clearance can also be achieved via other receptors such as CD14 (Gegner, Ulevitch, and Tobias 1995). CD14 is a glycoprotein found on the surface of myeloid cells and is the most well-characterized receptor involved in macrophage activation (Schumann et al. 1990). This receptor has a molecular weight ranging from 48 to 60 kDa and is not homologous to any other protein. It can exist as a soluble protein or a membrane-bound receptor anchored to the cell via glycosylphosphatidylinositol (Bazil et al. 1986; Wright et al. 1990). Soluble CD14 is found in serum as well as urine and is thought to be shed from macrophage following hydrolysis of phospholipids. This receptor mediates the uptake of endotoxin by cells such as endothelial cells that do not express membrane-bound CD14 (Pugin et al. 1993). Membrane-bound CD14 does not transverse the cell membrane and it is not yet clear how this receptor mediates cell activation. Instead CD14 interacts with other transmembrane proteins to mediate its effect on the cell. In macrophage, one protein that is proposed to be a binding partner with CD14 is the toll-like receptor 4 (TLR4). Lipopolysaccharide-binding protein (LBP) is another soluble protein that interacts with endotoxin in serum (Thurman et al. 1998) and is proposed to mediate the delivery of LPS to its receptor(s) on macrophages. This protein is synthesized constitutively by hepatocytes and binds to the lipid A portion of endotoxin. LBP acts as an opsonin and facilitates binding of endotoxin to CD14 receptors (Schumann et al. 1990). In support of the above hypotheses, mice deficient in CD14 receptors (Yin et al. 1999), TLR4 (Uesugi et al. 2001), and LBP (Uesugi et al. 2002), are all partially protected against liver damage due to ethanol.

Alcohol Metabolism

Toxic Byproducts of Alcohol Metabolism

In the liver, there are three enzyme systems that predominantly mediate ethanol metabolism (oxidation). Although the dominant system is clearly alcohol dehydrogenases (ADH), the cytochrome P450 systems (mostly CYP2E1) and catalases also contribute to ethanol metabolism (Figure 3). All of these pathways produce acetaldehyde as their product. Although acetaldehyde is subsequently oxidized to acetate by aldehyde dehydrogenases (ALDHs), the rate of this reaction is sufficiently slow that increases in acetaldehyde occur in humans consuming alcohol. A number of the systemic toxic effects of ethanol abuse (e.g., flushing, headaches, and nausea) are proposed to be mediated by direct or indirect effects of elevated acetaldehyde levels in the blood. At the level of the liver, acetaldehyde may also play a role in ALD (Eriksson 2001; Lieber 1988). For example, acetaldehyde can form adduct with proteins or small molecules at reactive residues (e.g., cysteines). Chemical alteration of these molecules can change and/or interfere with normal biologic processes and be directly toxic to the cell. Such modified biologic molecules may also stimulate the immune response and cause an autoimmune-like disease. Indeed, antibodies against such oxidatively modified proteins have been shown in both humans and animal models of ALD (Klassen, Tuma, and Sorrell 1995, 2000; Niemala 2001). For example, a hybrid adduct of malondialdehyde and acetaldedyde (MAA) unique to alcohol exposure has also been shown to raise an immune response both in human alcoholics and in animal models of ALD (Thiele et al. 2001).

Another indirect mechanism by which alcohol metabolism may contribute to ALD is the induction of CYP2E1. The colocalization of CYP2E1 in the hepatic lobule with regions of initial liver damage after alcohol (perivenous) led to the original hypothesis that CYP2E1 plays a causal role in ALD. Indeed, work with inhibitors of CYP2E1 have shown that they partially block hepatic pathology caused by ethanol in experimental animal models (Bardag-Gorce et al. 2000), supporting this hypothesis. Although the mechanism(s) by which CYP2E1 induction may play a role in ALD are unclear, one mechanism that has been proposed is that this enzyme is a source of reactive oxygen species and thereby contributes to oxidative stress caused by alcohol (see below and Figure 3). However, clear understanding of the contribution of CYP2E1 in vitro has been difficult because cultured hepatocytes have relatively low CYP2E1 activity. To address this problem, HepG2 cells that overexpress CYP2E1 were developed by Cederbaum and colleagues. The work to date with these cells has supported the hypothesis that CYP2E1 is involved in hepatocyte damage due to alcohol (Cederbaum et al. 2001). On the other hand, mice lacking CYP2E1 developed liver injury analogously to their wild-type counterparts (Kono et al. 1999). It has been suggested that other isoforms of cytochrome P450 (e.g., CYP4A) may play a compensatory role in the initiation of early alcohol-induced liver injury in the absence of CYP2E1 in the knockouts (Leclercq et al. 2001). However, recent work by the same group has shown that steatohepatitis in a dietary-induced model of nonalcoholic steatohepatitis (NASH) appears independent of CYP isoform activity (Ip et al. 2003). Furthermore, it was recently shown that broad spectrum inhibition of cytochrome P450 isoforms with aminobenzotriazole (ABT) conferred no protective effect on oxidative stress or injury in either rats or mice fed enteral alcohol (Isayama et al. 2003). It therefore appears that cytochrome P450 activity is not required for the initiation of oxidative stress and early experimental liver disease in rodents. This observation does not preclude a role for CYP2E1 in later stages of disease progression in mice or in other species. Much work remains to be done to address these issues.

Biochemical Effects of Alcohol Metabolism

In addition to forming cytotoxic byproducts and induction of enzymes that may indirectly contribute to alcohol toxicity (e.g., CYP2E1), the metabolism of ethanol itself alters the cellular redox state, which can also indirectly be involved in the mechanism of ALD. Specifically, the oxidation of ethanol to acetaldehyde by ADH and subsequent oxidation to acetate by ALDH utilizes NAD⁺ as the electron acceptor, subsequently shifting the NADH:NAD⁺ ratio to a more reduced state. This shift in the pyridine nucleotide redox state has been shown to impair normal carbohydrate and lipid metabolism, which has multiple effects, including decreasing the supply of ATP to the cell (Lieber 2000b). The increased reduced state of pyridine nucleotides has also been proposed to be responsible for the accumulation of lipids during alcohol ingestion (steatosis), which, as discussed above, appears to be critical for the progression to more severe changes in the disease (Day and James 1998).

Reactive Oxygen and Nitrogen Species

Under basal conditions, reactive oxygen and nitrogen intermediates (ROS and RNS, respectively) are products of normal cellular metabolism that have generally beneficial effects (e.g., cytotoxicity against invading bacteria). However, because ROS/RNS also damage normal tissue, controlling the balance between pro-oxidants and antioxidants is critical for the survival and function of aerobic organisms. When the balance within the cell is tipped to favor overproduction of ROS/RNS, oxidative stress can occur. Of the many chronic diseases in which oxidative stress is proposed to play a role, ALD was one of the earliest identified (DiLuzio 1966; Shaw et al. 1981). Indeed, ALD is well known to correlate with increases in indices of oxidative stress (e.g., lipid peroxidation) in humans (Figure 3). Although this oxidative stress associated with ALD in the clinics most likely involves increases in pro-oxidant production (see below), alcoholics often replace up to 50% of their total daily calories with ethanol (Patek 1979), leading to nutritional deficiencies that is also further complicated by malabsorption in the gastrointestinal (GI) tract (Bujanda 2000). The combined result is that alcoholics often have lower levels of dietary antioxidant molecules (see Lieber 2000asee Lieber 2000b for reviews) and overall impaired antioxidant defenses. Therefore, oxidative stress in ALD is likely to be mediated both by an increase in production of pro-oxidants, as well as by a decrease in antioxidant defenses.

In support of the hypothesis that oxidative stress is involved in ALD, numerous antioxidants have been shown to protect against the damaging effects of ethanol in in vitro and in vivo models of ALD (Arteel 2003; Kono et al. 2000, 2001a, 2001b, 2001c). Although advances have been made in the understanding of the role of pro-oxidants in experimental alcohol-induced liver injury, this work has yet to translate into an accepted antioxidant therapy for ALD in humans. As mentioned above, a better understanding of the mechanisms by which oxidative stress leads to liver damage during alcohol exposure will likely increase the effectiveness of future therapies. For example, antioxidant therapies that are targeted to the specific cellular/molecular effects of ethanol may be applied in the clinic with potentially greater success. Some of the key ROS and RNS proposed to be involved in ALD are detailed below.

Reactive Oxygen Species: A Role for Superoxide

Superoxide (O₂ ^·−) can come from multiple sources within the cell, including enzymes that are designed specifically to produce this molecule (e.g., xanthine oxidase and NA(D)PH oxidases), as well as electron leakage from other enzyme systems (e.g., mitochondria and CYP2E1). Support for the role of O₂ ^·−in ALD comes mainly from correlations between injury and levels of superoxide dismutase (SOD), an enzyme that catalytically reduces O₂ ^·− to hydrogen peroxide (H₂O₂) (McCord and Fridovich 1969). For example, levels of SOD correlate inversely with the severity of pathologic changes in the enteral alcohol model (Polavarapu et al. 1996). Further, adenoviral gene delivery of either cytosolic Cu/Zn-SOD or mitochondrial Mn-SOD was also shown to prevent alcohol-induced liver injury in rats fed enteral alcohol (Wheeler et al. 2001a, 2001b). Recent work by Kessova et al. (2003) showing that SOD1 knockout mice develop more severe liver damage due to alcohol corroborate the work with SOD overexpression. Although the experimental evidence supports the idea that pro-oxidant formation is dependent on the production of O₂ ^·− during ALD, this radical is not a potent oxidant and cannot directly mediate the formation of a number of the oxidative stress indices observed with ALD. Instead, it is likely that O₂ ^·− reacts through catalytic pathways within the cell to form more potent oxidants, such as hydroxyl radical (OH^·; the Fenton reagent) (Fridovich 1995) and hypochlorous acid (HOCl⁻) (Klebanoff 1968). A more recently identified pathway of O₂ ^·−-dependent oxidative stress involves the reaction of O₂ ^·− with NO^· to form peroxynitrite (ONOO⁻), another potent oxidizing and nitrating species (Beckman et al. 1990). All of the above-mentioned pathways form pro-oxidants with the strength to lead to the formation of the products observed during ALD. Therefore, although O₂ ^·− appears to be important in ALD, it should most likely be considered to be a key ‘starting point’ of oxidative stress during alcohol exposure.

Reactive Nitrogen Species: A Role for Nitric Oxide

Like O₂ ^·− for ROS, it is likely that nitric oxide (NO) serves as the main ‘parent’ molecule for formation of other RNS. Unlike O₂ ^·−, NO^· has pleotropic effects that make it unclear as to whether it predominantly plays a protective or damaging role in ALD (Hon, Lee, and Khoo 2002). On one hand, it is well known that the dysregulation of vascular tone after acute alcohol and during alcoholic cirrhosis is mediated, in part, by decreased NO^· production (Oshita et al. 1994; Wiest and Groszmann 2002). It has also been demonstrated that NO^· is antiapoptotic in hepatocytes and is required for normal liver regeneration (Kim et al. 2000; Rai et al. 1998). On the other hand, high levels of NO^· can also potentially be damaging in ALD by favoring production of RNS, such as ONOO^–. Indeed, recent work has shown that mice deficient in the inducible form of NO^· synthase (iNOS) are protected against experimental ALD (McKim et al. 2003), supporting the idea that NO^· production from this enzyme is damaging under these conditions. RNS derived from NO^· can cause nitration reactions (e.g., 3-nitrotyrosine formation) and nitrosation reactions (e.g., nitrosothiol formation), as well as oxidation reactions during alcohol exposure. Thus, NO^· appears to play a dual role in ALD, mediating both protective effects and tissue damage by overproduction of RNS; which of these events predominates in vivo may depend on the cell type, NOS isoform, and stage of disease studied.

Alcohol-Induced Modifications in Cellular Redox Balance Independent of Pro-Oxidant Formation

Alcohol also causes modifications to the cell that can lead to oxidative stress independent of increased pro-oxidant production per se. Lower antioxidant levels in alcoholics due to the nutritional deficiencies (see above) serve as an example. Additionally, the mobilization of free iron caused by alcohol can also lead to an increase in transition-metal catalysis to potent oxidants (e.g., the Fenton reaction). Indeed, experiments have shown that higher cellular iron levels enhance damage due to alcohol both in vitro and in vivo (Cederbaum et al. 2001; Tsukamoto et al. 1995). Recent work has also shown that iron plays a role in proinflammatory signaling in Kupffer cells via activation of the NF-κB (Xiong et al. 2003) Another example is the inhibition by alcohol of the 26S proteosome in hepatocytes by alcohol exposure (Bardag-Gorce et al. 2000). This protein complex is responsible for degrading oxidatively modified proteins. Its inhibition could therefore lead to the accumulation of oxidatively damaged proteins, possibly even in the absence of a real increase in prooxidant production (Donohue 2002) Lastly, there exists a family of proteins and systems involved in the “antioxidant network.” This group does not directly intercept pro-oxidants, but serves instead as ancillary reductants and thereby supports the catalytic activity of antioxidant proteins or small molecules. These systems utilize cellular energy to maintain these cycles. Therefore, the depletion by ethanol of both cytosolic and mitochondrial energy supplies (see above) can indirectly impair cellular antioxidant defenses. For example, the efficient reduction of hydroperoxides by glutathione peroxidases can only be maintained by active electron transfer from glutathione reductase, at the expense of NADPH; the major part of NADPH supply to the cytosol comes from the pentose phosphate shunt and is therefore directly dependent on the supply of glucose for glycolysis (see Sies 1993, for review).

How Are Oxidants Involved in Alcoholic Liver Injury?

One of the mechanisms by which oxidative stress is proposed to cause cellular injury is by direct chemical modification of biologic molecules, leading to altered and/or impaired processes within the cell. Also, as mentioned above, antibodies against oxidatively modified proteins have been shown in both humans and animal models of ALD (Klassan, Tuma, and Sorrell 1995; Thiele et al. 2001), which could stimulate an autoimmune disease in the liver. It is now apparent that pro-oxidants can also coordinate and/or increase their signals by modifying signaling cascades within the cell. Many reviews have addressed the role of signaling cascades in damage due to oxidative stress (Allen and Tresini 2000; Droge 2002; Forman and Torres 2001; Kamata and Hirata 1999) and the reader is directed to these papers for more detailed review of this point. In general, oxidant-sensitive signaling cascades include small molecules (e.g., intracellular Ca²⁺) (Ermak and Davies 2002), stress-activated protein kinases (e.g., c-Jun N-terminal kinase [JNK], extracellular regulated kinase [ERK] 1/2, and p38) (Suzuki, Forman, and Sevanian 1997), transcription factors (e.g., activator protein [AP]-1, hypoxia-inducible factor [HIF]-1, and NF-κB) (D’Angio and Finkelstein 2000), and modulators of apoptosis signaling (e.g., caspases, Bad, and Bcl-2) (Hoek and Pastorino 2002). Ethanol has been shown to alter the signal and/or magnitude of many of these cascades in vitro and/or in experimental ALD. For example, experimental ALD in rats is associated with activation of NF-κB, which is blunted by overexpression of SOD1 (Wheeler et al. 2001a, 2001b). However, whether all of these signaling cascades that are modulated by ethanol are mediated by oxidative stress per se is often not clear and requires further study.

Cytokines and Inflammatory Mediators

Priming and Sensitization of Inflammation During Alcoholic Liver Disease

A key concept in alcoholic liver injury is priming and sensitization (Figure 3). As mentioned above, the natural history of ALD is characterized by chronic inflammation in the liver. This activation of the immune response can be at least partially attributed to increased levels of activators of this response, such as LPS (Bigatello et al. 1987; Spencer, Rubin, and Lieber 1983), as discussed above. However, the levels of LPS found in alcoholics and in experimental ALD are comparably low relative to those found in endotoxemia or sepsis; furthermore, damage to liver caused by alcohol cannot be mimicked by chronic low-dose LPS in the absence of ethanol (Deaciuc et al. 1999). Instead, inflammatory cells appear to be “primed” to activation by LPS during alcohol exposure. Indeed, peripheral blood monocytes obtained from patients with alcoholic hepatitis have higher basal and stimulated levels of proinflammatory mediators (e.g., TNF-α) than do monocytes isolated from control individuals or alcoholics without active liver disease (McClain and Cohen 1989). Other cytokines and chemokines that have been shown to be elevated in ALD patients are IL-6 (Hill et al. 1992), IL-8 (Hill, Marsano, and McClain 1993), monocyte chemoattractant protein (MCP)-1 (Devalaraja et al. 1999), and macrophage inducible protein (MIP)-1 (Fisher et al. 1999). In addition to cytokine/chemokine production, there are a host of other proinflammatory mediators that are increased in ALD, including adhesion molecules, ROS/RNS (see above), and cytokine receptors (e.g., TNF receptor 1 [TNFR1]). In addition to liver damage, a number of the systemic effects of ALD (e.g., muscle wasting, increased gut permeability, and fever) are thought to be mediated by elevated levels of cytokines (McClain et al. 2002).

In addition to priming inflammatory cells, liver cells appear to be sensitized to inflammatory stimuli by alcohol administration. For example, although TNF-α is normally proproliferative in naïve hepatocytes, it is proapoptotic after ethanol treatments in vivo (Colell et al. 1998). HepG2 cells overexpressing CYP2E1 are also more sensitive to TNF-α–induced cell killing after exposure to alcohol (Pastorino and Hoek 2000; Liu et al. 2002). This sensitization to TNF-α killing appears to be mediated through the cellular “death domain” pathways (Liu et al. 2002). Other cell types (e.g., stellate cells) also appear to be primed/sensitized by alcohol exposure (Kim et al. 2000).

Priming and sensitization also implies that there is a series of sequential events in the progression of liver injury during alcohol exposure (Figure 3). In particular, the priming of inflammatory cells by ethanol causes a more robust release of proinflammatory mediators that has an exacerbated toxic response in sensitized hepatocytes. This pattern explains why blunting the activation of Kupffer cells or employing knockouts with an impaired Kupffer cell response can protect against damage in hepatocytes. For example, knocking out the LPS serum carrier protein LBP was shown to prevent liver damage caused by alcohol, without preventing induction of CYP2E1 (Uesugi et al. 2002). Such cell-to-cell signaling is also proposed to explain the reason that some mediators appear to have both damaging and protective roles in ALD. For example, the transcription factor NF-κB is a key proinflammatory mediator in Kupffer cells by up-regulating cytokine production (e.g., TNF-α). On the other hand, NF-κB inhibition causes hepatocyte apoptosis after addition of TNF-α or Fas (Bradham et al. 1998; Hatono et al. 2000). When rats were transfected with adenoviral vectors containing the IκB superrepressor and fed alcohol enterally, liver damage was prevented (Uesugi et al. 2001). Further, the production of TNF-α from liver under these conditions was also blunted. It is therefore likely that the potential damaging/proinflammatory effects of NF-κB activation, such as TNF-α production in Kupffer cells, precede any protective/antiapoptotic effects in hepatocyte during ALD.

Effects of Proinflammatory Cytokines in Alcoholic Liver Disease Beyond Inflammation

An interesting observation during work with antioxidants and knockout mice in studies of experimental ALD is that many of these conditions blunt steatosis caused by alcohol (Arteel et al. 2003; Kono et al. 2000, 2001a, 2001b; McKim et al. 2002; Wheeler et al. 2001a, 2001b). Steatosis due to alcohol is generally considered to be the result of redox inhibition of mitochondrial β-oxidation caused by alcohol metabolism (see above). Although redox changes are likely necessary for alcohol-induced steatosis, the data obtained with antioxidants question whether redox inhibition is sufficient in and of itself (see Arteel 2003 for review). Specifically, none of the conditions mentioned above (i.e., employing knockouts with impaired pro-oxidant production or antioxidant supplementation) had any apparent effect on alcohol metabolism and therefore are unlikely to affect the shift in the pyridine nucleotide redox state caused by alcohol. A factor that these studies do have in common is that the increase in cytokine (e.g., TNF-α) production caused by alcohol was prevented. It has been shown that TNF-α (and other cytokines) can indeed influence lipid metabolism both in liver and periphery (see Pessayre, Mansouri, and Fromenty 2002 for review). For example, TNF-α increases free fatty acid release from the periphery (Hardardottir et al. 1992), increases lipogenesis in hepatocytes (Feingold and Grunfeld 1987), and also impairs β-oxidation of fatty acids (Nachiappan et al. 1994). Other cytokines induced by alcohol (e.g., IL-1 and IL-6) may also block transport and secretion of triglycerides (Navasa et al. 1998). The net consequence during alcohol exposure is that cytokines increase the delivery of fatty acids to liver while simultaneously blunting the ability of liver cells to process and remove them. In support of this idea, TNFR1 knockout mice are protected against alcohol-induced fat accumulation in experimental ALD (Yin et al. 1999). It is proposed that these effects work in tandem with alcohol-induced shifts in the NADH redox state to cause steatosis.

Genetic Polymorphisms and the Risk for Alcoholic Liver Disease

It is clear that the risk for the development of ALD increases with the time- and dose-dependent consumption of alcohol (Lelbach 1966). However, as mentioned above, even in populations that consume high levels of ethanol (e.g., >80 g/day), only a fraction of individuals develop severe forms of the disease. Even when other risk modifiers are taken into account, there is no combination of environmental risk that leads to a 100% incidence of the disease. Furthermore, some individuals develop ALD while consuming much less alcohol for a shorter period of time. For these reasons, it has been long proposed that in addition to environmental risks, there are genetic risks that increase the chance of the development of ALD (Day 2000).

It is now clear that many gene products are normally polymorphic in humans; polymorphisms in these genes can give rise to structural and functional variations in the resultant proteins, as well as the amount of protein produced. It has been proposed that polymorphisms in key genes may be the basis of the apparent genetic risk for the development of ALD. Indeed, many of the above-described pathways thought to be involved in ALD have been shown to contain polymorphisms normally within the human population. For example, ethanol-metabolizing systems (e.g., ADH, ALDH, and CYP2E1) are well known to be polymorphic (Agarwal 2001). ROS/RNS formation and damage are also candidates, but have yet to clearly shown in large case-controlled studies. As mentioned above, iron catalysis of oxidative stress may be critical in ALD and there is indeed a strong link between dysregulated iron homeostasis (e.g., hemachromatosis) and alcohol-induced liver disease (Adams and Agnew 1996). Lastly polymorphisms in proinflammatory (e.g., TNF-α) and anti-inflammatory (e.g., IL-10) cytokines have also been associated with increased risk for ALD (Grove et al. 1997, 2000). It should be mentioned that, like environmental risk factors, there is no polymorphism identified thus far that leads to a 100% incidence of ALD in individuals who consume alcohol. Instead, it is likely that a myriad of environmental and genetic risk factors combine to determine the risk of the individual. The goal of future work is to better understand these factors and how they interact.

CONCLUSION

Alcoholic liver disease encompasses a spectrum of hepatic histopathologic alterations that includes steatosis, steatohepatitis, fibrosis, and cirrhosis. Steatosis and steatohepatitis represent the early-phase histopathologic changes of ALD. Steatosis is the most common, earliest, and benign stage of ALD. Although hepatocytes in fatty liver are viable, these cells are susceptible to death from a variety of secondary stresses. The molecular bases underlying the enhanced susceptibility of fatty hepatocytes are not well understood and remain the subject of potential research. Steatohepatitis is an inflammatory response upon continued ethanol ingestion leading to a more severe pathologic change. The acute, clinical outcome of alcoholic steatohepatitis is dictated by the extent of hepatocyte loss and the severity of the associated inflammatory response. Alcoholic steatohepatitis is known to represent a rate-limiting step in the progression of ALD to fibrosis and cirrhosis. Consequently, steatohepatitis is a common cause of fibrosis and cirrhosis. A better understanding of the mechanisms that are responsible for initiation and progression of alcoholic steatohepatitis is necessary to develop effective treatments for this type of liver disease.