Dexamethasone does not exert direct intracellular feedback on steroidogenesis in human adrenal NCI-H295A cells.

Experimental therapy of fetuses affected with congenital adrenal hyperplasia (CAH) has been reported by administering dexamethasone (Dex) to pregnant women at risk for carrying a CAH fetus. Such prenatal therapy can almost wholly eliminate virilization of the external genitalia of affected female fetuses, but only when treatment is started before 9 weeks of gestation. As it is not known whether the hypothalamic-pituitary-adrenal axis is functional at this time, and as the minimal effective doses of Dex are substantially supraphysiologic for the fetus, the mechanism of action of prenatal Dex treatment has been unclear. To assess the possibility that Dex might act directly on the adrenal, we used human adrenocortical NCI-H295A cells, an established model of the human fetal adrenal. Short term (6 h) incubation of these cells with Dex decreased synthesis of 11-deoxycortisol and 17alpha-hydroxyprogesterone and increased synthesis of deoxycorticosterone (DOC), but only at very high concentrations of Dex (> or =10 microM) that affect P450c17 (17alpha-hydroxylase/17,20 lyase) as a competitive inhibitor; no effects were seen at lower concentrations corresponding to those used in prenatal treatment. When NCI-H295A cells were treated with up to 100 microM Dex for 72 h, dot-blot analysis showed no changes in the abundance of the mRNAs for P450scc (cholesterol side-chain cleavage enzyme), P450c17, or 3beta-hydroxysteroid dehydrogenase II (3betaHSDII). When NCI-H295A cells were transfected with promoter/reporter constructs for the human genes for P450scc and P450c17, 24-h treatment with Dex at doses up to 100 microM had no effect on the transcription of these constructs. Because the regulation of steroidogenic processes in NCI-H295A cells closely resembles such regulation in the adrenal, we suggest that Dex may not act by an intra-adrenal mechanism in the experimental prenatal treatment of CAH.