Cortisol stimulates secretion of dehydroepiandrosterone in human adrenocortical cells through inhibition of 3betaHSD2

Abstract

Context: Initiating factors leading to production of adrenal androgens are poorly defined. Cortisol is present in high concentrations within the adrenal gland, and its production rises with growth during childhood.

Objective: Our aim was to characterize the effect of cortisol and other glucocorticoids on androgen secretion from a human adrenocortical cell line and from nonadrenal cells transfected with CYP17A1 or HSD3B2.

Design/setting: This study was performed in cultured cells, at an academic medical center.

Methods: The effects of cortisol upon steroid production in human adrenal NCI-H295R cells were measured by immunoassay, tandem mass spectrometry, and thin-layer chromatography. The effects of cortisol upon the activities of 17, 20 lyase and 3βHSD2 were measured in NCI-H295R cells and in transfected COS-7 cells.

Results: Cortisol markedly and rapidly stimulated dehydroepiandrosterone (DHEA) in a dose-dependent manner at cortisol concentrations ≥50 μM. Cortisone and 11-deoxycortisol were also potent stimulators of DHEA secretion, whereas prednisolone and dexamethasone were not. Treatment with cortisol did not affect expression of CYP17A1 or HSD3B2 mRNAs. Stimulation of DHEA secretion by cortisol was associated with competitive inhibition of 3βHSD2 activity.

Conclusions: Cortisol inhibits 3βHSD2 activity in adrenal cells and in COS-7 cells transfected with HSD3B2. Thus, it is possible that intraadrenal cortisol may participate in the regulation of adrenal DHEA secretion through inhibition of 3βHSD2. We hypothesize that a rise in intraadrenal cortisol during childhood growth may lead to inhibition of 3βHSD2 activity and contribute to the initiation of adrenarche.

Figure 1

Effect of cortisol upon DHEA secretion from human adrenal cells. Dose-response of DHEA secretion from NCI-H295R cells, measured by RIA (A) or by liquid chromatography tandem mass spectrometry (LC/MS-MS) (B), after exposure to cortisol for 8 h. In both panels, compared with no steroid treatment, cortisol ≥50 μm caused a significant increase in DHEA secretion. Absolute DHEA measurements, in ng/mg protein, for the no steroid treatments were 52.9 (A) and 14.3 (B), respectively. C, Dose-response of DHEAS secretion from NCI-H295R cells, measured by ELISA. D, Time course of cortisol (500 μm) stimulation of DHEA secretion, measured by RIA. The inset highlights the first 2 h of treatment, with a significant increase in DHEA at and beyond 30 min. E, DHEA secretion was not stimulated after exposure for 8 h to a wide range of dexamethasone concentrations and was significantly inhibited at the highest concentration. F, Compared with no steroid treatment, exposure to cortisone or 11-deoxycortisol (11-Deoxy) stimulated DHEA secretion, whereas exposure to prednisolone was no different than untreated cells. Treatment with 8-bromo-cAMP also showed stimulation of DHEA production. F, Cells were exposed to steroids (250 μm) or 8-bromo-cAMP (100 μm) for 24 h. DHEA measurements in E and F were done by ELISA. Each value is expressed as the mean of three replicates per group. Each experiment, except the time course (D), was performed 2–4 times with similar results. The time course (D) was repeated once, though with fewer time points, and yielded similar results. *, P < 0.005; **, P < 0.001.

Figure 2

Effect of cortisol on HSD3B2 and CYP17A1 mRNA in human adrenal cells. HSD3B2 and CYP17A1 mRNA expression did not change after treatment with cortisol (500 μm) for 8 h. Each value is expressed as the mean of 5 or 6 replicates.

Figure 3

Effect of cortisol upon steroidogenic enzyme activities in human adrenal cells. Representative TLC and quantification of specific metabolic products of [³H]-pregnenolone (A) or [³H]-DHEA (B) in cells exposed to cortisol (500 μm), expressed as the percentage of total radiolabeled steroids. Media were collected 6 h after adding [³H]-pregnenolone or 2.5 h after adding [³H]-DHEA. In both panels, tables display the ratios of product to precursor steroids and the corresponding changes in enzyme activities. Using either steroid precursor, cortisol markedly inhibited 3βHSD2 activity. C, 3βHSD2 activity, calculated from the ratio of secreted unlabeled androstenedione to unlabeled DHEA, measured by tandem mass spectrometry. 3βHSD2 activity showed dose-dependent inhibition at cortisol concentrations ≥10 μm. 17αOHPreg, 17αOH-pregnenolone; Preg, pregnenolone; 17αOHProg, 17αOH-progesterone; Prog, progesterone; DHEA, dehydroepiandrosterone; A’dione, androstenedione. *, P < 0.01; **, P < 0.001.

Figure 4

Effect of cortisol on steroidogenic enzyme activities in nonadrenal cells engineered to express HSD3B2 or CYP17A1. Representative TLC and quantification of metabolic products of [³H]-DHEA in cells transiently transfected with HSD3B2 (A and B) or of [³H]-pregnenolone in cells transiently transfected with CYP17A1 (C). In the absence of transfection of either gene, no metabolism of the labeled steroid was observed. Substrate-velocity (A) and Lineweaver-Burk (B) plots were used to obtain Michaelis-Menten kinetic constants for 3βHSD2 metabolism of DHEA in the absence and presence of cortisol (50 μm). There was little change in relative maximal velocity, V_max (P = 0.28), and a 3-fold increase in the apparent Michaelis constant, K_m (P < 0.001), with the addition of cortisol, indicating that cortisol is a competitive inhibitor of 3βHSD2. In the representative TLC chromatogram in A, the DHEA substrate concentration was 0.5 μm. C, Increasing amounts of cortisol were associated with small but significant increases in 17αOH-pregnenolone and decreases in DHEA (P = 0.006 and P = 0.003, respectively for effect of cortisol), indicating a modest decrease in 17, 20 lyase activity.