Page images
PDF
EPUB

iron from the ferritin stores in the liver. This free iron catalyzes the formation of the highly reactive hydroxyl radical, which, together with superoxide radicals, initiates peroxidation of polyunsaturated fatty acids (Hultcrantz et al.

1991).

Cellular organelles called peroxisomes partially metabolize lipids to form hydrogen peroxide. During fasting, when levels of fatty acids are increased in the liver and circulation, alcohol is oxidized predominantly via catalase (Handler and Thurman 1990; Thurman and Handler 1989), thereby producing hydrogen peroxide, which may enhance the liver lipid peroxidation described above.

Acetaldehyde, the first product of alcohol metabolism, can be oxidized by aldehyde dehydrogenase (ALDH) and by two other enzymes, aldehyde oxidase and xanthine oxidase (XO). Both these enzymes are active only at very high acetaldehyde concentrations. Metabolism of acetaldehyde by XO generates the superoxide O2; acute alcohol administration has been reported to increase the generation of XO (Kato et al. 1990; Sultatos 1988). That XO contributes to alcohol-induced hepatic lipid peroxidation is supported by the finding that such peroxidation is significantly inhibited by pretreatment with allopurinol, an XO inhibitor (Park et al. 1988). In addition to XO, aldehyde oxidase could be involved in the alcohol-induced hepatic oxidative

stress.

Another possible mechanism for the pathogenesis of alcohol-induced liver damage could be an imbalance between the increased production of free radicals and decreased availability of antioxidants, a condition known as oxidative stress.

Disturbances in antioxidant defense

Another possible mechanism for the pathogenesis of alcohol-induced liver damage could be an imbalance between the increased production of free radicals and decreased availability of antioxidants, a condition known as oxidative stress (Sies and Cadenas 1985). In alcoholics, impairment of antioxidant protective mechanisms provided by glutathione (GSH), vitamin E, ascorbic acid, and selenium have been reported.

GSH, a tripeptide composed of glycine, glutamic acid, and cysteine, exists in both reduced (95 percent) and oxidized (5 percent) forms and is localized predominantly in the cytosol with about 10 percent present in the mitochondria. It plays an important role in scavenging toxic free radicals (by serving as a cofactor for the GSH S-transferase enzymes) produced by the cytochrome P450 system. GSH is an important cofactor for glutathione peroxidase, which catalyzes the reduction of hydrogen peroxide and organic hydroperoxides. Depletion of GSH can promote lipid peroxidation. Thus, GSH is important in protecting hepatocytes against electrophilic metabolites in general and against reactive oxygen species in particular.

Acute in vivo exposure to alcohol elicits a decrease in hepatic GSH levels within several hours (Gonzalez et al. 1988; Speisky et al. 1985, 1988). Mechanisms of depletion of GSH by a single large dose of alcohol include a reduction in hepatic GSH synthesis and increased efflux of GSH from the liver (Lauterberg et al. 1984; Speisky et al. 1985). In vitro incubation of isolated hepatocytes with alcohol did not decrease GSH.

Effects of chronic alcohol administration on hepatic GSH content are controversial. Most studies in rats have reported no change (Callans et al. 1987) or slight increase (Kawase et al. 1989; Yang and Carlson 1991) in GSH levels after 4 to 8 weeks of the rats' receiving liquid diets containing alcohol. Other investigators reported a decrease in GSH levels (Boyer and Peterson 1990; Fernandez-Checa et al. 1987). Although GSH levels were practically unchanged after chronic alcohol feeding, the hepatic turnover of GSH was increased (Callans et al. 1987; Morton and Mitchell 1985).

Although chronic alcohol administration did not cause appreciable changes in total hepatic GSH, there was a significant decrease in the mitochondrial content of GSH as a result of decreased transport of GSH from the cytosol into the mitochondria. Some of the alcohol-induced mitochondrial damage may be a consequence of the loss of mitochondrial GSH, which protects against oxidant stress occurring during electron transport.

In baboons and humans, GSH levels were decreased after chronic alcohol ingestion. In laboratory experiments, the addition of a precursor to GSH called S-adenosyl-L-methionine to the diets of baboons corrected the observed decrease in

GSH after chronic alcohol feeding (Lieber et al. 1990).

Vitamin E (α-tocopherol) is a major antioxidant against membrane lipid peroxidation (McCay 1985; Gotoh and Niki 1992). Although acute alcohol administration did not significantly alter the total hepatic a-tocopherol concentration (Bjorneboe et al. 1987), it reduced a-tocopherol levels in liver mitochondria (Rouach et al. 1988). This effect may be attributed to an increase in mitochondrial a-tocopherol demand secondary to free radical generation. Chronic alcohol consumption reduced hepatic a-tocopherol content in rats (Bjorneboe et al. 1987) and in human alcoholics (Bjorneboe et al. 1988). Furthermore, Kawase et al. (1989) reported that chronic alcohol feeding significantly increased hepatic lipid peroxidation in rats maintained on a diet low in vitamin E. Thus, chronic alcohol consumption and low dietary intake of vitamin E render liver cells more vulnerable to free radical attack.

Cytokine production

Cytokines are hormone-like proteins associated with changes such as fever, leukocytosis (an increase in white blood cells), thrombocytosis (an increase in blood platelets), and the hepatic acute phase response that occurs in patients with inflammatory disorders (Andus et al. 1991). Cytokines also serve as a signal linking a peripheral inflammatory site with the liver, which itself is a potential site for cytokine synthesis. For instance, Kupffer cells (resident macrophages of the liver that are responsible for phagocytosis and degradation of endotoxins such as lipopolysaccharides) are potent producers of the cytokine TNF-α, interleukin-1 (IL-1), and interleukin-6 (IL-6) (Busam, Bauer et al. 1990; Busam, Homfield et al. 1990). In humans, cytokines may be involved in the fibrotic and cirrhotic transformation of the liver and may contribute to the pathogenesis of alcoholic hepatitis. Activation of protein kinase C in Kupffer cells leads to the production of cytokines, which promote the differentiation of Ito cells into collagenproducing myofibroblasts. The lipid peroxides may then stimulate collagen synthesis by differentiated Ito cells, resulting in fibrosis and, ultimately, cirrhosis (Matsuoka et al. 1990; Weiner et al. 1990).

The cytokine TNF-α mediates many of the biologic effects of endotoxins (toxins present in the bacterial cell wall as a lipopolysaccharide complex). Patients with alcoholic hepatitis frequently have endotoxemia (endotoxins in the blood)

and manifest clinical symptoms similar to the actions of TNF-α. TNF-α, which is strongly suspected of playing an etiologic role in liver injury (Lehmann et al. 1987), is found in measurable amounts in the plasma of patients with severe alcoholic hepatitis, and TNF-a's presence correlates with mortality from this condition (Felver et al. 1990). Monocytes (white blood cells) from patients with alcoholic hepatitis produced higher amounts of TNF-α in vitro (McClain and Cohen 1989; Weiner et al. 1990). Furthermore, serum levels of IL-6 and TNF-α were significantly elevated in patients with liver cirrhosis (Bird et al. 1990; Deviere et al. 1990). Increased production of TNF-α stimulates collagen deposition, resulting in fibrosis, as explained below.

Hepatocellular necrosis (death of liver cells) is followed by fibrosis and cirrhosis caused by an increase in collagen synthesis and deposition. Other components of extracellular matrix, such as fibronectin (a protein that is produced by Ito cells and is key in the wound-healing process), fibroblasts, and hepatocytes, contribute to the process of fibrosis and cirrhosis. Cytokines such as TNF-α and transforming growth factor B-1 (TGF-B-1) increase hepatocellular collagen synthesis (Czaja et al. 1989). Proliferation of Ito cells is controlled by TGF-α, IL-1, and TNF-α (Matsuoka et al. 1989, 1990; Pinzani et al. 1989). Under the influence of TGF-B-1, Ito cells differentiate to form myofibroblasts.

Collagen synthesis by Ito cells is also stimulated by acetaldehyde (even in the small amounts formed during alcohol metabolism) and 4-hydroxynonenal (an aldehyde produced during lipid peroxidation), which activate the transcription of genes for collagen synthesis (Brenner and Chojkier 1987). Several transcription factors that bind to the promoter region of collagen genes have been identified; one is modified by acetaldehyde, thereby resulting in alteration of its function.

Role of acetaldehyde

About 90 percent of acetaldehyde formed in the liver by ADH is oxidized to acetic acid by the mitochondrial ALDH (Lieber 1991). Chronic alcohol consumption results in a reduction in the capacity of mitochondria to metabolize acetaldehyde, thus leading to higher acetaldehyde levels in alcoholics (Baraona et al. 1987). Acetaldehyde is biochemically more reactive and less stable than its precursor, alcohol. Because of its reactivity, acetaldehyde can promote lipid peroxidation and bind covalently to cellular proteins.

Consequently, acetaldehyde has been suspected of damaging cellular microstructure, inducing fibrosis, and affecting energy metabolism. Acetaldehyde also appears to be a key generator of free radicals in the liver (French 1991; Kennedy and Tipton 1990; Lieber 1991; Tuma et al. 1990, 1991).

Acetaldehyde has been reported to bind to microsomal liver proteins (Lieber and Nomura 1981), to other hepatic macromolecules (Lin et al. 1988), to serum albumin (Donohue et al. 1983), and to tubulin, the constituent protein of microtubules (Savolainen et al. 1987; Smith et al. 1989; Tuma et al. 1991). Acetaldehyde binding to tubulin prevents its polymerization to form microtubules, the disruption of which has been implicated in the inhibition of protein secretion; the result is the retention of secretory proteins within hepatocytes (Smith et al. 1989). This protein retention may contribute to swelling and hypertrophy (enlargement) of liver cells observed in alcoholic liver disease, thus resulting in an increase in resistance to blood flow through the liver and ultimately leading to the development of portal hypertension (Lieber 1991). Acetaldehyde binding to other proteins such as hemoglobin and cytochrome P450 IIE1 leads to interference with enzymatic and other critical hepatic functions, thereby resulting in the perpetuation of liver damage (Lin et al. 1988). For instance, these acetaldehyde adducts have been reported to serve as neoantigens in mice (Israel et al. 1988) and in humans (Hoerner et al. 1986). GSH can prevent binding of acetaldehyde to tubulin and other proteins (Mitchell et al. 1991).

One of the few studies that has examined genetic susceptibility to alcoholic liver cirrhosis has identified differences in the type I collagen gene, which was associated with the development of alcoholic cirrhosis.

Fibrosis is a distinguishing feature of serious alcoholic liver damage. The pathogenesis of fibrosis in alcoholic liver damage may involve direct deposition of collagen induced by acetaldehyde (Brenner and Chojkier 1987). In cultured fibroblasts and myofibroblasts (derived from Ito cells), acetaldehyde was found to stimulate both synthesis of collagen and transcription of its cor

responding gene (Brenner and Chojkier 1987). Whether acetaldehyde binds to a specific site on the gene or whether its effects are indirectly mediated through the promotion of lipid peroxidation remains to be elucidated.

Collagen is a protein that is deposited in the central hepatic venules during fibrosis. One of the few studies that has examined genetic susceptibility to alcoholic liver cirrhosis has identified differences in the type I collagen gene, which was associated with the development of alcoholic cirrhosis (Moshage et al. 1990). This finding needs to be confirmed by additional studies before conclusions regarding the importance of collagen formation in pathogenesis of alcoholic cirrhosis can be established.

Immunologic mechanisms

It is well documented that alcohol depresses cell-mediated immune responses (see next section) and that alcoholics with liver disease generally have fewer circulating T-cells than nonalcoholic patients (MacGregor 1986). Furthermore, antibodies against acetaldehyde-protein adducts are frequently detected in the serum of patients with alcoholic liver disease (Paronetto 1986), although these antibodies are found also in alcoholics without significant liver damage and in patients with nonalcoholic cirrhosis. Therefore, the relative significance of immunemediated injury in the etiology of alcoholic liver disease remains to be elucidated. Nonetheless, the finding of a correlation between the histocompatibility complex human lymphocyte antigen (HLA) B8 (sometimes associated with autoimmune diseases) (Saunders et al. 1982) and HLA B35 (Marbet et al. 1988) with increased progression of liver disease suggests that individuals with a genetic predisposition for immune dysregulation may be at higher risk for liver disease (Fleisher et al. 1988; Israel et al. 1988).

Neutrophil chemoattraction

Acute alcoholic hepatitis in humans is characterized by infiltration of the liver by polymorphonuclear leukocytes (PMNs), or neutrophils (Zakim et al. 1990). Metabolism of alcohol by human hepatocytes in vitro produced a lowmolecular-weight polar lipid that exhibited a strong chemotactic activity for PMNS (Roll et al. 1986). The formation of this polar lipid is blocked by inhibiting alcohol metabolism, by oxygen radical scavengers, or by iron chelation (Roll et al. 1991) but is stimulated by increasing

intracellular iron, thus suggesting that oxygen radicals may be responsible for the generation of this chemoattractant lipid molecule. Once formed, this chemoattractant molecule could further exacerbate the inflammatory process in alcoholic liver injury by stimulating PMNs to generate their own oxygen-derived free radicals (Williams and Barry 1987).

Oxygen deficiency

Chronic feeding of alcohol to animals predominantly affects the zone of the liver surrounding the central vein (perivenular zone) where oxygen tension is low compared with that in the periportal region of the liver around the portal vein (Orrego et al. 1988). French et al. (1984) reported that increasing hepatic oxygen delivery corrected the reductions in adenine nucleotide synthesis (an indication of the energy status) in chronic-alcohol-fed rats. In baboons with alcoholrelated liver damage, hepatic venous oxygen tension and hepatic blood flow were decreased (Lieber 1990). Thus, lack of oxygen caused by either a decrease in hepatic blood flow or a reduction in oxygen tension may play a role in liver damage. This hypoxia theory led to the suggestion that propylthiouracil (PTU) could be used to treat alcoholic liver disease (Orrego et al. 1987). A study by Lieber (1989) demonstrated that impaired oxygen consumption, rather than lack of oxygen supply, may explain alcohol-induced liver injury.

Treatment of Alcohol-Induced
Liver Injury

PTU, an antithyroid drug, was found experimentally to attenuate the hypermetabolic state of rat livers induced by chronic alcohol consumption (Israel et al. 1975). It also reduced mortality from alcoholic liver disease in a controlled clinical trial of 310 alcohol abusers at various stages of liver injury (Orrego et al. 1987). The therapeutic benefit of PTU in the treatment of alcoholic hepatitis, cirrhosis, or both needs to be confirmed in future investigations.

Because the mechanisms involved in injury to the hepatocytes are under genetic control, it is probable that genetically determined differences in individuals might explain why some individuals develop cirrhosis or progress faster than others to an end-stage condition (Johnson and Williams 1985; Williams and Saunders 1983). Abstinence from alcohol will prevent further liver damage in individuals with early stages of

liver disease. Treatment modalities, including liver transplantation, are used for individuals. with end-stage liver disease. Liver transplantation has been performed in alcoholics and has demonstrated success and survival rates equal to that of transplantation for nonalcoholic cirrhotic patients (Stevens et al. 1991; Van Thiel et al. 1991). It is important to note that reported recidivism rates (return to drinking) have been very low (Kumar et al. 1990). Other treatment modalities of alcoholic liver disease include corticosteroid therapy (Maddrey 1990). The administration of S-adenosyl-L-methionine attenuated alcohol-induced liver injury in baboons (Lieber et al. 1990).

Alcohol and the Pancreas

The effects of alcohol on the pancreas have long been recognized. Approximately 65 percent of all cases of pancreatitis are alcohol related (Balart and Ferrante 1982). Epidemiologic studies have demonstrated a linear relationship between mean daily alcohol consumption and risk for pancreatitis (Singh 1990).

The acinar cells of the pancreas secrete enzymes necessary for digestion. Alcohol-induced injury to the acinar cells may lead to interstitial leaking of these enzymes. As a result of the proteolytic and lysosomal activity of these enzymes, "autodigestion" of the tissue may occur. Acute pancreatitis is characterized by destruction of the acinar cells' membranes and inflammatory damage to blood vessels (Singh 1990). Chronic relapsing pancreatitis is more commonly associated with alcoholism than acute pancreatitis (Korsten 1989).

An increase in protein and trypsinogen secretion was reported in alcoholics as compared with nonalcoholic controls (Renner 1980). This finding suggests that hypersecretion of these pancreatic enzymes may be the cause of acute pancreatic tissue injury.

Relapsing pancreatitis in the chronic alcoholic may be due to an increase in the protein content of pancreatic secretions, which induces the formation of viscous protein plugs in the pancreatic ducts (Korsten 1989). Mezey et al. (1988) reported malnutrition to be a contributing factor to pancreatitis in a U.S. sample. Similarly, increased dietary fat may potentiate ethanol-induced pancreatic injury (Tsukamoto et al. 1988).

The greater susceptibility of women to alcohol-related tissue damage may not be limited to the liver. In one study, the duration of excessive

use was shorter in women with alcoholic pancreatitis than in men, although no differences in daily alcohol consumption were detected.

Alcohol-Induced
Cardiovascular Injury

While there is considerable evidence that moder-
ate alcohol consumption decreases the risk of
death from coronary artery disease (CAD), nu-
merous studies have shown that chronic heavy
drinking is associated with other cardiovascular
diseases, such as cardiomyopathy (heart muscle
disease), hypertension (high blood pressure), ar-
rhythmias (disturbances in heart rhythm), and
stroke (cerebrovascular hemorrhage) (Lands and
Zakhari 1990). Increased public awareness of
risk factors associated with lifestyles have re-
sulted in the observed decline in mortality from
these diseases. Nonetheless, cardiovascular disor-
ders remain the leading cause of death in the
United States.

While there is considerable evidence that moderate alcohol consumption decreases the risk of death from coronary artery disease, numerous studies have shown that chronic heavy drinking is associated with other cardiovascular diseases, such as cardiomyopathy, hypertension, arrhythmias, and stroke.

The following discussion focuses on recent studies examining the relationship between possible cardiovascular benefits associated with light to moderate drinking and on deleterious cardiovascular effects, such as hypertension, arrhythmias, cardiomyopathy, and stroke, identified with chronic heavy drinking. Possible mechanisms underlying alcohol's effects on the heart and vascular system are also explored.

Previous epidemiologic studies (reviewed by Moore and Pearson 1986) demonstrated a lower risk for CAD in light and moderate drinkers compared with abstainers. Thus, a U-shaped relationship has been reported in which nondrinkers have a slightly higher rate of coronary heart disease than light or moderate drinkers, and heavy drinkers have an elevated risk (Rosenberg 1981; Stampfer et al. 1988). However, other investiga

tors have suggested that moderate drinking does not protect against CAD and have argued that higher mortality among abstainers was due to the inclusion in this category of people who stopped drinking because of ill health (Marmot and Brunner 1991; Shaper 1990; Shaper et al. 1988). Furthermore, the Stampfer et al. (1988) study that showed decreased risk for CAD in light-drinking women was criticized because a higher percentage of obese and diabetic women was included among the nondrinkers than among the light and moderate drinkers. However, recent studies investigating these factors did not support that conclusion and noted that including "sick quitters" in the abstinent category cannot completely explain the apparent protective effect of moderate drinking against CAD (Boffetta and Garfinkel 1990; Jackson et al. 1991; Klatsky et al. 1990; Miller et al. 1990; Rimm et al. 1991). Even so, some investigators have remained skeptical (Thelle et al. 1991) or have not found significant beneficial effects associated with moderate alcohol consumption (Farchi et al. 1992).

Studies that showed evidence for a protective effect of alcohol reported an average alcohol consumption of fewer than three drinks per day in men (Boffetta and Garfinkel 1990). Similar results have been obtained with women. Stampfer et al. (1988) determined that consumption of approximately one drink per day decreases the risk of coronary heart disease in women. Razay et al. (1992) found that consumption of up to two drinks per day is associated with lower levels of cardiovascular risk in women.

Determining the mechanisms underlying the apparent protective effect of moderate drinking has been an area of much interest to investigators. Narrowing of the coronary arteries by atherosclerosis is a principal mechanism leading to coronary insufficiency (Fuster et al. 1992). Since increased low-density lipoprotein (LDL) cholesterol is a major risk factor for CAD, and increased high-density lipoprotein (HDL) cholesterol is widely regarded to be protective, it was natural to speculate that moderate alcohol consumption may increase HDL cholesterol (Leaf 1988; Steinberg et al. 1991). Various subfractions of HDL exist; the subfraction believed to protect against CAD, the HDL-2 subfraction, was increased with alcohol consumption (Razay et al. 1992), as was the HDL-3 subfraction, which is not associated with protection against CAD (Puchois 1990; Razay et al. 1992). Further, Langer et al. (1992) estimated that elevation of

« PreviousContinue »