The origin of glucocorticoid hormone oscillations

Abstract

Oscillating levels of adrenal glucocorticoid hormones are essential for optimal gene expression, and for maintaining physiological and behavioural responsiveness to stress. The biological basis for these oscillations is not known, but a neuronal "pulse generator" within the hypothalamus has remained a popular hypothesis. We demonstrate that pulsatile hypothalamic activity is not required for generating ultradian glucocorticoid oscillations. We show that a constant level of corticotrophin-releasing hormone (CRH) can activate a dynamic pituitary-adrenal peripheral network to produce ultradian adrenocorticotrophic hormone and glucocorticoid oscillations with a physiological frequency. This oscillatory response to CRH is dose dependent and becomes disrupted for higher levels of CRH. These data suggest that glucocorticoid oscillations result from a sub-hypothalamic pituitary-adrenal system, which functions as a deterministic peripheral hormone oscillator with a characteristic ultradian frequency. This constitutes a novel mechanism by which the level, rather than the pattern, of CRH determines the dynamics of glucocorticoid hormone secretion.

Figure 1. Regulation of glucocorticoid hormone secretion.

(A) Negative feedback in the HPA axis plays a key role in regulating glucocorticoid (CORT) secretion. Neurons of the hypothalamic paraventricular nucleus (PVN) secrete CRH into the portal vein, which acts on the anterior pituitary to release ACTH into the general circulation. ACTH activates cells in the adrenal cortex to synthesize and secrete CORT, which in turn feeds back directly on the anterior pituitary to inhibit ACTH secretion, as well as acting at higher centres in the brain, including the hypothalamus and hippocampus. (B) Endogenous corticosterone (the main glucocorticoid in rodents) oscillations in a freely behaving male Sprague-Dawley rat. Shaded region indicates the dark phase. (C–F) Mathematical modelling predicts that ultradian ACTH and glucocorticoid (CORT) oscillations are regulated by a systems-level negative feedback mechanism in the pituitary-adrenal network, independent of pulsatile hypothalamic activity. Numerical simulations show that the pituitary-adrenal network can oscillate under conditions of constant CRH drive to the pituitary (C–D). Oscillations in ACTH and CORT are characterized by a small phase shift (shaded region indicates phase difference between oscillation peaks). For higher levels of CRH drive, the oscillations are rapidly damped to steady-state levels of hormone (E–F). AU, arbitrary units.

Figure 2. Glucocorticoid response to constant CRH infusion.

(A–H) Individual (A, C, E, G) and mean (B, D, F, H) corticosterone responses to constant saline (A–B) or CRH infusion (0.5 µg/h, C–D; 1.0 µg/h, E–F; 2.5 µg/h, G–H). Grey bar indicates the period of infusion; error bars represent mean ± standard error of the mean (SEM) (n = 6–8 per group). (I) Dose-dependent effect of CRH on the corticosterone response. Overall effect of the CRH infusion was significant (AUC, p<0.0001; Kruskal-Wallis ANOVA on ranks). Error bars represent mean ± SEM (n = 6–8 per group); *p<0.001. (J) Synchronous corticosterone oscillations (in individual rats) in response to constant CRH infusion (0.5 µg/h; n = 6). Grey bar indicates the period of infusion.

Figure 3. Frequency comparison of CRH-induced and endogenous glucocorticoid oscillations.

(A) Individual corticosterone oscillations in response to constant CRH infusion (0.5 µg/h). Grey bar indicates the period of infusion. (B) Normalized power spectra of the corticosterone oscillations in (A). (C) Corticosterone oscillations during the circadian peak in untreated control (UC) rats. Shaded region indicates the dark phase. (D) Normalized power spectra of the corticosterone oscillations in (C). (E) Mean peak frequency (i.e., frequency corresponding to the maximum power in the spectrum) of corticosterone oscillations in response to constant CRH infusion (0.5 µg/h; CRH; n = 6), and of corticosterone oscillations during the circadian peak in untreated control rats (UC; n = 13). Error bars represent mean ± standard error of the mean (SEM). (F) Frequency evolution of the corticosterone oscillations in (A). AU, arbitrary units.

Figure 4. ACTH and glucocorticoid response to constant CRH infusion.

(A–B) Individual (A) and mean (B) ACTH and corticosterone (CORT) oscillations in response to constant CRH infusion (0.5 µg/h; n = 6). (C–D) Individual (C) and mean (D) time course of the ACTH and corticosterone (CORT) response to constant CRH infusion (0.5 µg/h) during the initial activation phase (0–25 min) of the oscillation (n = 4). There was a significant overall effect of the CRH infusion on both ACTH and corticosterone (ACTH, p<0.0001; corticosterone, p<0.005; one-way ANOVA). ACTH was significantly different from basal (time zero) by 10 min (p<0.005), whereas corticosterone was not significantly different from basal (time zero) until 20 min (p<0.05). (E) Phase-shifted ACTH and corticosterone (CORT) response to constant CRH infusion (0.5 µg/h) over the duration of the first pulse (n = 3–7 per time point). Grey bar indicates the period of infusion (starting at 0700 h); error bars represent mean ± standard error of the mean (SEM).