Free radical products have previously been detected in rodents after chronic feeding with an ethanol-containing, high-fat diet. The significance of reactive free radical formation in ethanolinduced hepatotoxicity has been difficult to assess because most rodent models exhibit only fatty liver. However, serious hepatic damage resembling clinical alcoholic liver injury (eg, steatosis, inflammation, and necrosis) occurs in rats after con-tinuous intragastric administration of an ethanol-containing, high-fat diet developed by Tsukamoto and French. Accordingly, rats treated with ethanol for at least 2 weeks using this protocol were administered the spin trap a-(4-pyridyl-1-oxide)-N-tert-butylnitrone, and bile samples were collected. A six-line radical adduct spectrum was detected in the bile of ethanol-treated rats. A similar spectrum of lower intensity was detected with rats fed a high-fat diet without ethanol, but little or no radical adduct signal was detected with chow-fed animals. For both treatment groups, a-(4-pyridyl-1-oxide)-N-tert-butylni-trone and extra ethanol were given acutely. Destruction of Kupifer cells by chronic treatment with GdCl3 decreased by about 50% the radical adduct formation in rats fed the ethanol-containing, high-fat diet. This radical species was largely eth-anol derived, because addition of [13C] ethanol produced a 12-line spectrum, indicating the formation of a-hydroxyethyl radical. Ethanol treatment also caused hypoxia (detected on the liver surface in vivo with oxygen electrodes), which was reflected in a dose-dependent decrease in oxygen tension with ethanol. The effect was blocked by GdCI3. Hepatic damage detected by histology was prevalent in ethanol-treated rats but only mild fatty liver was observed in high-fat diet-fed controls. GdCl3 treatment eliminated hepatic damage due to high-fat and ethanol diets, and when all groups were compared a significant correlation between liver injury and radical adduct signal was observed. Thus, free radical formation in ethanoltreated rats has been detected for the first time in a model that exhibits injury characteristic of human alcoholic injury, and signal intensity correlates with hepatotoxicity. Moreover, the decrease in both free radical formation and hepatic damage produced by GdCl3 implicates Kupifer cells in the development of alcoholic liver injury. This important pathophysiological process may involve direct production of reactive oxygen species or indirect actions of mediators on parenchymal cells.

A number of studies have shown that ethanol administra-tion in vivo is associated with the production of free radicals and oxidative stress. Lipid peroxidation has been measured (1), and hepatic antioxidants decrease after ethanol administration(2-4), but the most direct evidence for free radical activity comes from the detection of spin-trapped free radicals from living animals. Reinke et al.(5) detected lipidderived radical adduct species in heart and liver of rats treated with high dietary fat (corn oil) and ethanol. With deer mice fed a corn oil-and ethanol-containing diet, both lipidderived and ethanol-derived radical adduct species were de-