A voluntary oral ethanol-feeding rat model associated with necroinflammatory liver injury

The intragastric (IG) ethanol infusion model results in fatty liver, necrosis, inflammation and fibrosis. This model was utilized to study the pathogenesis of alcoholic liver disease (ALD). Disadvantages of the IG model include maintenance of the animals and equipment expense. To develop a voluntary feeding model for ALD, we took advantage of two important observations in the IG model: (i) female rats demonstrate greater severity of alcohol-induced liver injury than males and (ii) rats fed fish oil as a source of fatty acids develop more severe alcoholic liver injury than rats fed other fatty acids with ethanol.

Female Wistar rats (205 to 220 g) were fed for 8 weeks a diet containing 8% ethanol, fish oil (30% of calories), protein, and dextrose. Pair-fed controls (FD) received dextrose in amounts isocaloric to ethanol. The following measurements were made: liver pathology [fatty liver (0 to 4), necrosis, inflammation and fibrosis by Sirius Red], endotoxin and alanine aminotransferase (ALT) in plasma, urine ethanol, lipid peroxidation, nuclear factor kappa-B (NF-kappaB) and mRNA levels for tumor necrosis factor-alpha (TNF-alpha), cyclooxygenase-2 (COX-2), and inducible nitric oxide synthase (iNOS). Protein levels for iNOS and nitrotyrosine were evaluated by immunohistochemistry and Western Blot analysis. Liver proteasome and cytochrome P450 2E1 activity and protein levels of asialoglycoprotein receptor (ASGPR) were also evaluated. In addition, mRNA levels of fibrogenic markers were assessed.

All animals lost weight for the initial 2 to 3 weeks but then gained weight until killing at 8 weeks. There was, however, a significant difference (p < 0.05) in weight between the ethanol-fed (Etoh) and (FD) groups at the end of the experiment. The mean urine ethanol levels ranged between 190 and 240 mg/dl. The severity of pathological changes was greater (p < 0.01) in Etoh vs. FD: fatty liver, 3.0 +/- 1.2 vs. 1.2 +/- 0.4; necrosis (foci/mm(2)), 3.9 +/- 2.3 vs. 0.4 +/- 0.3; inflammation (cells/mm(2)), 19.0 +/- 6.3 vs. 1.8 +/- 0.6. Centrilobular collagen deposition (% area), assessed by Sirius Red staining, was greater in Etoh vs. FD. Levels of endotoxin, ALT, CYP2E1 and lipid peroxidation markers were also higher (p < 0.01) in Etoh vs. FD. Levels of NF-kappaB and mRNA of pro-inflammatory mediators (TNF-alpha, COX-2, iNOS) and procollagen-I were increased (p < 0.05) in ethanol-fed rats. Immunohistochemical analysis showed more intense staining for both iNOS and nitrotyrosine in the centrilobular areas in the Etoh vs. FD groups. The greater area of positive staining for iNOS and nitrotyrosine in Etoh vs. FD was confirmed by Western Blot analysis. An increase in the expression of mRNA for profibrogenic genes (p < 0.05) was seen in ethanol-fed rats.

A voluntary feeding regimen consisting of fish oil and ethanol in female rats is technically less demanding yet produces pathological and biochemical changes similar to those observed with the IG model. Pathological changes include fatty liver, necrosis and inflammation. Increased NF-kappaB and mRNA and protein levels of the pro-inflammatory mediators TNF-alpha, COX-2 and iNOS, coincided with the presence of necroinflammatory changes. The voluntary feeding regimen is proposed as an alternative to the IG model in the study of alcoholic liver injury.