Fibrate hypolipidemic drugs regulate the concentrations of plasma high density lipoproteins (HDL), which are inversely correlated to the development of atherosclerosis. In rodents fibrates lower HDL levels due to a decreased transcription of its major apolipoprotein, apo A-I, in liver, whereas in man fibrates increase plasma levels of HDL via an induction of human apo A-I gene expression. The fibrate effect on human apo A-I is mediated by the transcription factor PPAR-alpha (peroxisome proliferator-activated receptor) which interacts with a positive PPAR-response element (PPRE) in its promoter. The lack of induction of apo A-I expression by fibrates in rodents is due to three nucleotide differences in the rodent apo A-I promoter eliminating binding of PPAR and activation by fibrates. These in vitro observations were extended in vivo in transgenic mice and rabbits overexpressing the human apo A-I gene under control of its homologous promoter containing the human apo A-I PPRE. Whereas the endogenous mouse apo A-I gene is repressed, treatment with fibrates results in the transcriptional induction of human apo A-I gene expression. This induction is accompanied by increased plasma concentrations of human apo A-I and HDL. To determine whether fibrates increase HDL and apo A-I concentrations without inducing hepatomegaly and peroxisome proliferation, their effects were tested in rabbits, an animal model more resistant to peroxisome proliferation. In contrast to normal rabbits, in which plasma lipoprotein levels remain unchanged, fibrate treatment of transgenic apo A-I rabbits results in increased plasma HDL and human apo A-I concentrations due to the induction of human apo A-I gene expression in liver, without affecting liver weight or peroxisomal acyl-CoA oxidase activity. In conclusion; (1) fibrates regulate plasma HDL concentrations, at least partly, due to their effects on apo A-I gene transcription; (2) the opposite effects of fibrates on apo A-I gene expression in rodents and humans are due to sequence differences in regulatory elements in their respective genes; (3) solely the presence of the human apo A-I gene is sufficient to confer fibrate-responsiveness on HDL; and (4) the beneficial effects of fibrates on lipoprotein metabolism are independent of any undesirable proliferation of peroxisomes.