Chronic administration of eicosapentaenoic acid and docosahexaenoic acid as ethyl esters reduced plasma cholesterol and changed the fatty acid composition in rat blood and organs.

Fish oils rich in n-3 fatty acids have been shown to decrease plasma lipid levels, but the underlying mechanism has not yet been elucidated. This investigation was performed in order to further clarify the effects of purified ethyl esters of eicosapentaenoic acid (EPA-EE) and docosahexaenoic acid (DHA-EE) on lipid metabolism in rats. The animals were fed EPA-EE, DHA-EE, palmitic acid, or corn oil (1 g/kg/d) by orogastric intubation along with a chow background diet for three months. At the end the animals were sacrificed. Plasma and liver lipids were measured, as well as lipid-related enzyme activities and mRNA levels. The fatty acid composition of plasma and different tissues was also determined. This study shows that, compared to the corn oil control, EPA-EE and DHA-EE lowered plasma cholesterol level, whereas only EPA-EE lowered the amount of plasma triacylglycerol. In liver peroxisomes, both EE preparations increased fatty acyl-CoA oxidase FAO activities, and neither altered 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase activities. In liver microsomes, EPA-EE raised HMG-CoA reductase and acyl-CoAicholesterol acyltransferase activities, whereas DHA-EE lowered the former and did not affect the latter. Neither product altered mRNA levels for HMG-CoA reductase, low density lipoprotein-receptor, or low density lipoprotein-receptor related protein. EPA-EE lowered plasma triacylglycerol, reflecting lowered very low density lipoprotein secretion, thus the cholesterol lowering effect in EPA-EE-treated rats may be secondary to the hypotriacylglycerolemic effect. An inhibition of HMG-CoA reductase activity in DHA-EE treated rats may contribute to the hypocholesterolemic effect. The present study reports that 20:5n-3, and not 22:6n-3, is the fatty acid primarily responsible for the triacylglycerol lowering effect of fish oil. Finally, 20:5n-3 was not converted to 22:6n-3, whereas retroconversion of 22:6n-3 to 20:5n-3 was observed.