Role of estradiol in cortisol-induced reduction of luteinizing hormone pulse frequency

Abstract

Precise control of pulsatile GnRH and LH release is imperative to ovarian cyclicity but is vulnerable to environmental perturbations, like stress. In sheep, a sustained (29 h) increase in plasma cortisol to a level observed during stress profoundly reduces GnRH pulse frequency in ovariectomized ewes treated with ovarian steroids, whereas shorter infusion (6 h) is ineffective in the absence of ovarian hormones. This study first determined whether the ovarian steroid milieu or duration of exposure is the relevant factor in determining whether cortisol reduces LH pulse frequency. Prolonged (29 h) cortisol infusion did not lower LH pulse frequency in ovariectomized ewes deprived of ovarian hormones, but it did so in ovariectomized ewes treated with estradiol and progesterone to create an artificial estrous cycle, implicating ovarian steroids as the critical factor. Importantly, this effect of cortisol was more pronounced after the simulated preovulatory estradiol rise of the artificial follicular phase. The second experiment examined which component of the ovarian steroid milieu enables cortisol to reduce LH pulse frequency in the artificial follicular phase: prior exposure to progesterone in the luteal phase, low early follicular phase estradiol levels, or the preovulatory estradiol rise. Basal estradiol enabled cortisol to decrease LH pulse frequency, but the response was potentiated by the estradiol rise. These findings lead to the conclusion that ovarian steroids, particularly estradiol, enable cortisol to inhibit LH pulse frequency. Moreover, the results provide new insight into the means by which gonadal steroids, and possibly reproductive status, modulate neuroendocrine responses to stress.

Figure 1

Schematic description of the animal model and designs of experiment (Exp) 1 and 2. A, Artificial follicular phase model used in experiment 1 (full artificial follicular phase group in experiment 2). Ewes were ovariectomized (OVX) in the early luteal phase of the estrous cycle (d −12) and immediately treated with two intravaginal progesterone (P)-releasing devices (controlled internal drug release; DEC International NZ/Ltd, Hamilton, New Zealand) and a single 1-cm estradiol (E) implant sc, which maintains luteal phase plasma concentrations of estradiol (1–2 pg/ml) and progesterone (2–4 ng/ml) (15,52). After 12 d, the P devices were removed to simulate corpus luteum regression (−P, arrow) and four 3-cm estradiol implants were inserted sc 16 h later to simulate the follicular phase rise in estradiol secretion (+E). Preovulatory-like surges of GnRH and LH (open arrow, LH surge) typically begin about 24 h after onset of the estradiol rise to 6–8 pg/ml (13,15,18). B, Basal estradiol group of experiment 2, which differed from the full artificial follicular phase group in that the simulated preovulatory estradiol rise was omitted. C, No progesterone group of experiment 2, which differed from the full artificial follicular phase in that progesterone was not administered. Plasma progesterone and estradiol patterns are schematically depicted by the dashed and solid lines, respectively. Time in hours is relative to onset (0 h) of cortisol or vehicle infusion (cross-hatched horizontal bars). Frequent sampling periods are indicated by solid boxes.

Figure 2

A, Example LH profiles from vehicle-treated (top) and cortisol-treated (bottom) ovariectomized ewes in which ovarian steroid hormones were not replaced. B, LH profiles from vehicle-treated (top) and cortisol-treated (bottom) ovariectomized ewes in which ovarian steroids were replaced to mimic the follicular phase of the estrous cycle. For both A and B, the early (13–17 h) and late (25–29 h) sampling periods are shown on the left and right, respectively. Peaks of pulses are depicted with contrasting open or filled circles.

Figure 3

Example LH profiles from vehicle-treated (top) and cortisol-treated (bottom) ewes from full artificial follicular phase group (A), basal estradiol group (B), and no progesterone group (C). For A–C, the early (13–17 h) and late (25–29 h) sampling periods are shown on the left and right, respectively. Peaks of pulses are depicted with contrasting open or filled circles.

Figure 4

Change in LH pulse frequency between the early (13–17 h) and late (25–29 h) sampling periods in individual cortisol-treated ewes of each experimental group. The late period was after the estradiol (E) rise in the full artificial follicular phase and no progesterone (P) groups. Frequencies in individual ewes are depicted by points (solid or open) connected by a line. Solid points depict ewes in which frequency decreased by two or more pulses between the early and late periods; open points depict ewes in which frequency did not decrease to this extent.