N-[(1,5,9)-trimethyldecyl]-4 alpha,10-dimethyl-8-aza-trans-decal-3 beta-ol (8-azadecalin 1), a high-energy intermediate analogue for the 2,3-oxidosqualene-lanosterol cyclase, was found to be a powerful (IC50 approximately 0.1 microM) inhibitor of cholesterol biosynthesis in human hepatoma HepG2 cells. In analogy with other mammalian cells grown in the presence of cyclase inhibitors, the decrease in C27-sterol formation was accompanied by an accumulation of 2,3-oxidosqualene, 2,3:22, 23-dioxidosqualene, and by the formation of a compound characterized as 24,25-epoxycholesterol, a repressor of HMG-CoA (3-hydroxy-3-methylglutaryl coenzyme A) reductase activity. In order to assess the cyclase as a potential pharmacological target for the design of hypocholesterolemic drugs, it is important to test whether inhibitors of this enzyme are able to act synergistically on the biosynthesis of cholesterol, i.e. by decreasing the amount of lanosterol formed and by repressing the regulatory HMG-CoA reductase via the formation of regulatory oxysterols. The accumulation of 24,25-epoxycholesterol in relationship to the decrease of C27-sterol biosynthesis and of HMG-CoA reductase activity showed only a partial correlation: e.g. at [1] = 100 x IC50 only a 50% reduction in enzyme activity could be attained. In contrast, when HepG2 cells were treated with 2,3:22,23-dioxidosqualene or 24,25-epoxycholesterol, excellent correlations were found between the inhibition of C27-sterol biosynthesis and the repression of HMG-CoA reductase activity, which was almost complete at the highest concentrations of these epoxides (10(-5) M). Altogether, our results suggest that treatment of HepG2 cells with a cyclase inhibitor such as 8-azadecalin (1) does not lead to an intracellular accumulation of repressor molecules high enough to fully trigger a regulatory pathway resulting in a complete down-regulation of HMG-CoA reductase. At intermediary concentrations of cyclase inhibitors (IC50), however, a synergistic mode of action of these inhibitors seems plausible.