Sexual differentiation of pituitary function: apparent difference bewteen primates and rodents.

Surges in luteinizing hormone secretion resembling those which occur spontaneously during the menstrual cycle were induced by acute elevations in circulating estrogen concentrations in both male and female rhesus monkeys gonadectomized in adulthood. These experiments demonstrate that in primates, in contrast to rodents, exposure of the hypothalamohypophyseal unit to androgens throughout fetal and postnatal development does not prevent the differentiation of the control system that governs cyclic gonadotropin secretion.

Estradiol enables cortisol to act directly upon the pituitary to suppress pituitary responsiveness to GnRH in sheep.

We have shown that cortisol infusion reduced the luteinizing hormone (LH) response to fixed hourly GnRH injections in ovariectomized ewes treated with estradiol during the non-breeding season (pituitary-clamp model). In contrast, cortisol did not affect the response to 2 hourly invariant GnRH injections in hypothalamo-pituitary disconnected ovariectomized ewes during the breeding season. To understand the differing results in these animal models and to determine if cortisol can act directly at the pituitary to suppress responsiveness to GnRH, we investigated the importance of the frequency of GnRH stimulus, the presence of estradiol and stage of the circannual breeding season. In experiment 1, during the non-breeding season, ovariectomized ewes were treated with estradiol, and pulsatile LH secretion was restored with i.v. GnRH injections either hourly or 2 hourly in the presence or absence of exogenous cortisol. Experiments 2 and 3 were conducted in hypothalamo-pituitary disconnected ovariectomized ewes in which GnRH was injected i.v. every 2 h. Experiment 2 was conducted during the non-breeding season and saline or cortisol was infused for 30 h in a cross-over design. Experiment 3 was conducted during the non-breeding and breeding seasons and saline or cortisol was infused for 30 h in the absence and presence of estradiol using a cross-over design. Samples were taken from all animals to measure plasma LH. LH pulse amplitude was reduced by cortisol in the pituitary clamp model with no difference between the hourly and 2-hourly GnRH pulse mode. In the absence of estradiol, there was no effect of cortisol on LH pulse amplitude in GnRH-replaced ovariectomized hypothalamo-pituitary disconnected ewes in either season. The LH pulse amplitude was reduced in both seasons in experiment 3 when cortisol was infused during estradiol treatment. We conclude that the ability of cortisol to reduce LH secretion does not depend upon the frequency of GnRH stimulus and that estradiol enables cortisol to act directly on the pituitary of ovariectomized hypothalamo-pituitary disconnected ewes to suppress the responsiveness to GnRH; this effect occurs in the breeding and non-breeding seasons.

Psychosocial stress suppresses attractivity, proceptivity and pulsatile LH secretion in the ewe.

Various stressors suppress pulsatile secretion of luteinizing hormone (LH) in ewes and cortisol has been shown to be a mediator of this effect under various conditions. In contrast, little is known about the impact of stress and cortisol on sexual behavior in the ewe. Therefore, we tested the hypothesis that both psychosocial stress and stress-like levels of cortisol will reduce the level of attractivity, proceptivity and receptivity in addition to suppressing LH secretion in the ewe. In Experiment 1, a layered stress paradigm of psychosocial stress was used, consisting of isolation for 4 h with the addition of restraint, blindfold and noise of a barking dog (predator stress) at hourly intervals. This stress paradigm reduced LH pulse amplitude in ovariectomized ewes. In Experiment 2, ovariectomized ewes were artificially induced into estrus with progesterone and estradiol benzoate treatment and the layered stress paradigm was applied. LH was measured and sexual behavior was assessed using T-mazes and mating tests. Stress reduced pulsatile LH secretion, and also reduced attractivity and proceptivity of ewes but had no effect on receptivity. In Experiment 3, ewes artificially induced into estrus were infused with cortisol for 30 h. Cortisol elevated circulating plasma concentrations of cortisol, delayed the onset of estrus and resulted in increased circling behavior of ewes (i.e. moderate avoidance) during estrus and increased investigation and courtship from rams. There was no effect of cortisol on attractivity, proceptivity or receptivity during estrus. We conclude that psychosocial stress inhibits LH secretion, the ability of ewes to attract rams (attractivity) and the motivation of ewes to seek rams and initiate mating (proceptivity), but cortisol is unlikely to be the principal mediator of these effects.

Does seasonal reproductive state affect the neuroendocrine response of the ewe to a long-day pattern of melatonin?

This study examined whether or not the reproductive response of female sheep to photoperiod varies with seasonal reproductive state. The specific objective was to test the hypothesis that the reproductive response to a long-day pattern of melatonin varies with the reproductive state of the ewe. The response examined was the synchronization of reproductive neuroendocrine induction (rise in serum luteinizing hormone, or LH) following nocturnal infusion of melatonin into pinealectomized ewes for 35 consecutive nights. This infusion restored a pattern of circulating melatonin similar to that in pineal-intact ewes maintained in a long photoperiod (LD 16:8). The ewes had been pinealectomized and without melatonin replacement for 16-25 months prior to the study. They were in differing reproductive states at the start of the infusion, as their endogenous reproductive rhythm had become desynchronized among individuals and with respect to time of year. Noninfused pinealectomized ewes served as controls. Regardless of the reproductive state at the start of the 35-day infusion of the long-day pattern of melatonin, all treated ewes exhibited the same reproductive neuroendocrine response after the infusion was ended. This consisted of a synchronized rise in LH some 6-8 weeks after the infusion was terminated, the maintenance of a high level of serum LH for some 15 weeks, and a subsequent precipitous fall in LH to a very low level. These results provide evidence that a long-day pattern of melatonin can synchronize reproductive neuroendocrine induction in the ewe, regardless of reproductive condition, and thus do not support the hypothesis that this response differs with seasonal reproductive state.

Circannual alterations in the circadian rhythm of melatonin secretion.

To determine if a circadian rhythm known to be functionally related to the reproductive axis varies on a circannual basis, we monitored the circadian secretion of melatonin at monthly intervals for 2 years in four ovariectomized, estradiol-implanted ewes held in a constant short-day photoperiod. Prior to the study, ewes had been housed in a short-day (8L:16D) photoperiod for 4 years and were exhibiting circannual reproductive rhythms as assessed by serum luteinizing hormone (LH) levels. Three of the four sheep showed unambiguous deviations from the expected nocturnal melatonin secretion at two different times approximately 1 year apart. Nocturnal rises in melatonin, which usually last the duration of the dark phase, were delayed by 3-14 h or were missing. Altogether, five of the seven melatonin alterations observed in these three ewes occurred during the nadir of the circannual LH cycle. In the remaining ewe, we did not observe an altered melatonin secretory pattern during this period, and this ewe also failed to show a high amplitude circannual cycle of LH. The results provide evidence for a circannual change in the circadian rhythm of melatonin secretion. This alteration in melatonin secretion may serve as a "functional" change in daylength, and thereby may influence the expression of the circannual reproductive rhythm of sheep held in a fixed photoperiod for an extended time.

Pulsatile secretion of luteinizing hormone: differential suppression by ovarian steroids.

In sheep, physiological levels of estradiol and progesterone each suppress the pulses of LH characteristics of tonic LH secretion, but do so by completely different mechanisms. Estradiol treatment decreases LH pulse amplitude but not frequency and also inhibits the height of the LH peak resulting from the administration of gonadotropin-releasing hormone (GnRH). In contrast, progesterone decreases the frequency of LH pulses without reducing their amplitude or the response to exogenous GnRH. This suggests that progesterone suppresses tonic LH secretion by acting in the brain to decrease the frequency of GnRH pulses, while estradiol may suppress the response of the pituitary to GnRH and thereby decrease LH pulse amplitude.

Do gonadotropin-releasing hormone, tyrosine hydroxylase-, and beta-endorphin-immunoreactive neurons contain estrogen receptors? A double-label immunocytochemical study in the Suffolk ewe.

We used double label immunocytochemistry to examine the brains of ovariectomized ewes and determine whether GnRH, tyrosine hydroxylase-(TH), and beta-endorphin-immunoreactive (IR) neurons contain IR-estrogen receptors (ER). Because of their possible importance as a target for the feedback actions of estradiol, we also examined the presence of nuclear ER in LH-IR cells of the pars tuberalis of the pituitary. Although preoptic GnRH neurons were frequently in close proximity to ER-IR cells, only one out of approximately 1000 GnRH cells examined was found to coexpress ER. In contrast, in the arcuate nucleus and vicinity, 3-5% of TH cells and 15-20% of beta-endorphin cells contained ER. Virtually all LH-IR cells, seen predominantly in the ventral portion of the pars tuberalis, coexpressed ER. These results suggest that in sheep as in rodents, the influence of estradiol on the reproductive neuroendocrine system is not directly mediated by GnRH neurons, but instead is conveyed to GnRH cells via presynaptic afferents. Subsets of TH- and beta-endorphin-IR cells which coexpress ER are two candidates for relaying gonadal steroid signals to GnRH cells. At the level of the pituitary, the feedback actions of estradiol may be expressed directly upon the gonadotroph.

An intra-ovarian site for the luteolytic action of estrogen in the rhesus monkey.

To investigate whether estradiol can act within the ovary to induce luteolysis in the rhesus monkey, 100 mug estradiol-17beta were injected on one of days 2, 3, or 4 after the preovulatory LH peak into one of the following sites: the stroma of the ovary containing the developing corpus luteum, the stroma of the contralateral ovary, or sc. When estradiol was injected into the ovary containing the corpus luteum, the functional life span of the corpus luteum was shortened, reflected by a premature decline in circulating progesterone to levels characteristic of the follicular phase of the cycle and an early onset of menstruation. When estradiol was injected either sc or into the ovary contralateral to that containing the corpus luteum, the life span of the corpus luteum was not shortened. The differing responses could not be attributed to differing rates of efflux of estradiol from the various injection sites; patterns and levels of estradiol in peripheral serum were essentially the same regardless of injection site. Furthermore, the premature regression of the corpus luteum was not a consequence of surgical trauma associated with the intra-ovarian injection procedure, nor could it be mimicked by another ovarian hormone, progesterone. These findings lead to the conclusion that estrogens can induce functional luteolysis in the rhesus monkey by acting directly within the ovary containing the corpus luteum.

Inhibition of tonic secretion of luteinizing hormone by progesterone in immature female sheep.

The effect of progesterone on the secretion of LH was studied in intact and castrated immature female sheep between 12 and 42 weeks of age. (First spontaneous ovulation occurs between 30 and 50 weeks of age.) Progesterone was administered by means of Silastic capsules inserted SC to achieve circulating progesterone concentrations similar to those found in the adult during the mid-luteal phase of the estrous cycle (3-4 ng/ml). Prior to treatment, concentrations of circulating LH in intact lambs fluctuated widely (2 to 20 ng/ml) over a 24- to 48-h period. These high and variable levels of serum LH, reflecting a pulsatile secretion of LH characteristic of young female sheep, were not altered by the insertion of empty capsules. Implantation of progesterone-containing capsules, however, resulted in a decrease in serum LH to low levels (less than 0.5 ng/ml) within 4 h. Furthermore, the insertion of progesterone capsules into the same lambs after castration produced an immediate and sustained suppression of LH secretion similar to that which was observed when the ovaries were present. These findings led to the conclusion that in the immature female sheep, protesterone can effect a marked inhibition of LH secretion in the presence or absence of the ovaries. Although it has been postulated that progesterone plays an important role in the negative feedback control of LH secretion in adult sheep during the estrous cycle, it remains to be determined whether this steroid normally plays a physiologic role in the regulation of tonic LH secretion prior to the onset of ovarian cyclicity in the lamb.

Modulation of pituitary responsiveness to luteinizing hormone-releasing factor during the estrous cycle of the rat.

This study was performed to investigate pituitary responsiveness to synthetic gonadotropic hormone-releasing hormone (Gn-RH) administered during the 5-day estrous cycle of the rat and to examine whether the observed differences can be attributed to ovarian secretion of estradiol. Gn-RH was injected intra-arterially at 4:00 PM to rats previously anesthetized with sodium pentothal to block the LH surge on proestrus and thus minimize changes in LH secretion which occur throughout the estrous cycle. Pituitary responsiveness was defined as the difference between serum LH concentrations in samples obtained immediately before and 15 min after administration of 200 ng Gn-RH, a time when maximal circulating levels of LH were observed. Administration of Gn-RH was followed by a significant increase in circulating LH on all days of the estrous cycle on which the response was tested (diestrus 2, diestrus 3, proestrus, estrus). Pituitary responsiveness was relatively low on diestrus 2 and estrus and was increased slightly on diestrus 3. On proestrus, however, pituitary response to Gn-RH increased markedly, a phenomenon abolished by ovariectomy at 8:00 AM on diestrus 3. The large increase in pituitary responsiveness observed on proestrus was not restored in such ovariectomized rats when circulating estradiol concentrations were increased and maintained at approximately 150 pg/ml by SC insertion of Silastic capsules containing estradiol-17beta immediately following ovariectomy. Nevertheless, this estradiol treatment consistently elicited an LH surge in another group of ovariectomized rats not treated with sodium pentothal or Gn-RH. Although these observations indicate that an ovarian hormone is essential for the increase in pituitary response to Gn-RH on proestrus, the identity of this hormone remains to be established.

A daily signal for the LH surge in the rat.

We have demonstrated that a brief step-like increment in circulating estradiol concentrations to approximately 100 pg/ml, achieved by SC insertion of Silastic capsules containing estradiol-17 beta and their withdrawal after 29 1/2 hr, elicited a daily LH surge on 4 consecutive afternoons in rats ovariectomized 2 weeks previously. Since the duration of this stimulus was similar to that of the preovulatory increment in serum estradiol concentrations, it was postulated that the endogenous estrogen signal in intact rats might also trigger repetitive LH surges if ovulation and formation of corpora lutea were prevented. To test this hypothesis, rats were ovariectomized at 10:00 AM on proestrus (day 1) and blood samples were obtained at 12:30 and 5:00 PM on days 1 to 4. Although an LH surge occurred on proestrus, no subsequent LH discharges were observed. The absence of an LH surge on consecutive days could not be attributed to a difference between the endogenous estradiol stimulus and the exogenous stimulus which elicited repetitive LH surges in long-term ovariectomized rats. Rather, it was determined that in recently ovariectomized rats, in contrast to long-term ovariectomized rats, a daily LH surge occurred only if elevated serum estradiol concentrations were maintained. Thus, by leaving implants in place, an LH discharge was elicited on 10 consecutive days. These results support the concept that a neural signal for the LH surge is emitted each day throughout the estrous cycle of the rat, and that prolonged maintenance of elevated circulating estradiol concentrations is essential for the expression of these signals.

Development of the mechanism regulating the preovulatory surge of luteinizing hormone in sheep.

Development of the mechanism controlling cyclic LH secretion in the sheep was studied by examining the ability of estradiol to elicit LH surges in lambs at various ages. Silastic capsules containing estradiol were implanted sc for a 96-hour period at 3, 7, 12, 20, and 27 weeks of age. (First spontaneous ovulation occursss between 30 and 50 weeks). Although administration of estradiol failed to elicit a discharge of LH at 3 weeks of age, LH surges of progressively increasing magnitude (36 +/- 15, 73 +/- 28, 107 +/- 19 ng/ml) were elicited by estradiol as the lambs became older (7, 12, 20 weeks). By 27 weeks, the maximal serum LH level attained during the induced surge (123 +/- 29 ng/ml) was similar to that of the estrogen-induced LH surge in anestrous adults (179 +/- 23 ng/ml). Ovulation, however, did not occur in response to the induced LH surges. An additional experiment was performed to determine whether, as in the adult, progesterone can block the estradiol-induced LH discharge in the immature (12-week old) female. A sustained elevation of circulating progesterone to levels characteristic of the mid-luteal phase of the estrous cycle of the adult (3--4 ng/ml), beginning 24 h prior to insertion of the estradiol capsules, blocked the induced LH surge. These results demonstrate that, in immature female sheep, the LH surge mechanism is capable of functioning long before first ovulation occur and, further, suggest that timing of the initial preovulatory LH surge is limited by the ability of the ovary to produce the estradiol stimulus rather than by the ability of the hypothalamo-hypophyseal system to respond to the positive feedback action of estradiol. Additionally, the hypothalamo-hypophyseal mechanism whereby progesterone blocks the LH surge develops long before first ovulation.

Sexual differentiation of the mechanism controlling the preovulatory discharge of luteinizing hormone in sheep.

Silastic capsules containing estradiol-17beta were implanted subcutaneously into male and female sheep which had been gonadectomized as adults, In both sexes, this treatment resulted in a prompt decrease in serum LH concentrations reflecting a negative feedback suppression of tonic LH secretion. In females but not males, this decrease was followed, within 24 h, by an LH surge after which circulating LH declined to low levels resembling those in intact ewes during the luteal phase of the estrous cycle. In males, only a sustained reduction in circulating LH was observed. Three weeks after placement of the initial estradiol implants, while serum LH levels were low and stable, additional estradiol implants were inserted. This additional treatment effected step-like increments in circulating estradiol to approximately 20 or 55 pg/ml (depending on the number of implants) and elicited single or multiple LH discharges in females but failed to induce LH surges in males. These findings lead to the conclusion that the mechanism which governs the cyclic mode of LH secretion in sheep, as in rats, undergoes sexual differentiation. Further, the presence of multiple LH surges in some females suggests that an increment and sustained elevation in circulating estradiol can induce the occurrence, or permit the expression, of consecutive signals for the LH surge in the ovariectomized ewe.

The estradiol-induced surge of gonadotropin-releasing hormone in the ewe.

Previous studies suggest two roles for estradiol in inducing the LH surge in ewes: a neural action to evoke a sudden release of GnRH and a pituitary action to maximize response to GnRH. We tested two hypotheses: a follicular phase estradiol rise induces a GnRH surge; and the surge-inducing action of estradiol does not vary with season. In the breeding season, ewes in the midluteal phase of the estrous cycle were ovariectomized and treated with implants producing luteal phase levels of estradiol and progesterone, and an apparatus was surgically installed for later sampling of pituitary portal blood. At the normal time of luteolysis (1 week later), progesterone implants were removed, simulating luteal regression. Ewes were divided into two groups: estradiol implants also removed (n = 6) and estradiol implants added 16 h after progesterone removal to produce a rise in estradiol to levels that mimic those that circulate in the late follicular phase (n = 6). In anestrus, the estradiol rise treatment was replicated in ewes (n = 5) after an artificial luteal phase produced by sequential insertion and subsequent removal of progesterone implants. Regardless of season, the LH surge induced by estradiol was invariably accompanied by a massive GnRH surge, ranging from 73- to 394-fold over presurge values. The GnRH and LH surges began together, but the GnRH surge continued well beyond the surge of LH. There was no seasonal difference in time course or amplitude of the GnRH surge. Control ewes not treated with estradiol exhibited regular pulses of LH and GnRH every 1-2 h, but no surge of either hormone. We conclude that, regardless of season, a rise in estradiol to late follicular phase levels initiates a large and abrupt GnRH surge coincident with the onset of the LH surge. The LH surge ends despite continued elevation of GnRH.

A receptive period for estradiol-induced luteolysis in the rhesus monkey.

This study was performed to test the hypothesis that responsiveness to the luteolytic action of estradiol is acquired as the luteal phase of the menstrual cycle progresses. The luteolytic effect of fixed estradiol increment (270 +/- 12 pg/ml serum) was assessed at different stages of luteal function in rhesus monkeys. A 4-day elevation in estradiol early in the luteal phase (days 2--6 after the LH peak) caused a decrease in the concentration of serum progesterone but did not shorten luteal life span. In contrast, when provided during the midluteal phase (days 6--10), the same 4-day estradiol increment promptly induced premature luteolysis. Furthermore, during sustained exposure to the extradiol increment from days 2--10, signs of premature luteolysis were not evident until day 7 after the LH peak. Thus, the effects of estradiol early in the luteal phase do not alter luteal life span; it is the effects after day 6 that precipitate luteolysis. These observations support the existence of a receptive period for the luteolytic action of estradiol in the rhesus monkey.

Fos expression during the estradiol-induced gonadotropin-releasing hormone (GnRH) surge of the ewe: induction in GnRH and other neurons.

The protein product of the protooncogene c-fos was used as a marker of cellular activation in an attempt to identify those neurons in the preoptic area and hypothalamus that participate in generation of the estradiol-induced surge of GnRH in the ewe. GnRH- and Fos-expressing cells were identified immunocytochemically, and the percent of coexpression was determined in three states: mid-luteal phase (low GnRH release, n = 6); short-term ovariectomy (high episodic GnRH release, n = 6); and induced GnRH surge (high sustained release, n = 8). To induce the GnRH surge, a follicular phase rise in circulating estradiol was simulated in a physiological model for the estrous cycle. Serum LH was measured as an indicator of GnRH release. In the luteal phase, LH was basal, indicating low GnRH secretion. Few cells expressed Fos; these were not GnRH cells. Despite high intermittent GnRH release in short-term ovariectomized ewes, GnRH cells did not express Fos. During the surge (sustained high GnRH release), 41 +/- 8% of GnRH cells expressed Fos; these cells were dispersed throughout the field of distribution of GnRH neurons. In addition to Fos in GnRH-positive cells, many more non-GnRH cells in the preoptic area, anterior hypothalamus, and ventrolateral hypothalamus expressed Fos during the surge than in the luteal phase or after ovariectomy. We suggest that Fos expression in GnRH cells is markedly increased by the positive feedback action of estradiol (surge), whereas short-term removal of negative feedback (ovariectomy) has little, if any, effect, despite increased GnRH release in both states. Since estradiol induces Fos expression in far more than GnRH neurons, our results also suggest that estradiol activates other cells, some of which may be part of a neuronal chain leading to GnRH surge generation, and some of which may be related to other neural actions of estradiol, such as estrous behavior.

Dynamics of gonadotropin-releasing hormone (GnRH) secretion during the GnRH surge: insights into the mechanism of GnRH surge induction.

Recent studies demonstrate unequivocally that a preovulatory surge of GnRH is secreted into pituitary portal blood during the estrous cycle of the ewe and that this surge is induced by the follicular phase rise in estradiol. These data, obtained at 10-min intervals, suggested the surge results from a continuous elevation of GnRH rather than from a sequence of discrete pulses. This study examines the dynamics of GnRH secretion in more detail to determine if the surge results from strictly episodic release of the decapeptide. Our approach was to monitor GnRH secretion into pituitary portal blood at very frequent intervals during several "windows" of the GnRH surge induced using a physiological model for the estrous cycle. Samples of portal blood were obtained at either 2-min intervals (6 ewes), or 30-sec intervals (12 ewes) at several times during the surge; at other times portal blood was sampled less often to monitor progression of the GnRH surge. All ewes had an unambiguous GnRH surge; amplitude ranged from 100- to 500-fold over pressure levels. Regardless of sampling interval, our results provide no convincing evidence to indicate the enhanced secretion of GnRH is strictly episodic; values remained continuously elevated in portal blood. Our findings are consistent with the hypothesis that the GnRH surge is not composed entirely of discrete synchronous secretory events, and they raise the possibility that one action of estradiol in inducing the GnRH surge may be to switch the pattern of GnRH secretion into portal blood from episodic to continuous.

Pattern of gonadotropin-releasing hormone (GnRH) secretion leading up to ovulation in the ewe: existence of a preovulatory GnRH surge.

We have previously shown LH surges induced by physiological estradiol levels are invariably accompanied by robust and sustained GnRH surges in the ewe. Such an increase, however, has not been observed consistently during the preovulatory LH surge. In the present study, we examined GnRH secretion in Suffolk and Ile de France ewes during the preovulatory period using a method for pituitary portal blood collection which allows simultaneous portal and jugular blood samples to be taken at frequent intervals for up to 48 h. Ewes were sampled either during the mid-late luteal phase (n = 8) or follicular phase (n = 20). During the follicular phase, a robust increase in GnRH secretion occurred at the onset of the LH surge in 11 of 12 ewes sampled during the LH surge. The GnRH increase in most ewes was a massive surge, reaching values averaging 40-fold greater than baseline and extending well beyond the end of the preovulatory LH surge. In the single ewe not exhibiting a GnRH surge during the LH surge, postmortem inspection indicated blood was probably not sampled from the pituitary portal vessels. In the early follicular phase, GnRH-pulse frequency was greater than that observed in the luteal phase and, within the follicular phase, GnRH-pulse frequency increased further and amplitude decreased as the surge approached. These data demonstrate GnRH secretion leading up to ovulation in the ewe is dynamic, beginning with slow pulses during the luteal phase, progressing to higher frequency pulses during the follicular phase and invariably culminating in a robust surge of GnRH. The LH surge, however, ends despite continued elevation of GnRH.

Role of the thyroid gland in seasonal reproduction: thyroidectomy blocks seasonal suppression of reproductive neuroendocrine activity in ewes.

Seasonal reproductive transitions in ewes are generated endogenously and are synchronized by annual changes in photoperiod. Previous evidence indicates that thyroidectomy prevents the transition to anestrus in ewes maintained in a fixed day length, suggesting that the thyroid is needed for endogenously generated reproductive arrest. Here we tested the hypothesis that the thyroid is required for endogenous seasonal suppression of the neuroendocrine mechanism that regulates pulsatile LH secretion. Ewes were thyroidectomized (n = 6) in summer, 6 weeks before the onset of the breeding season, or they remained thyroid intact (n = 6). They were housed in a simulated natural photoperiod until the winter solstice; thereafter, they remained on that photoperiod (10 h of light, 14 h of darkness). To monitor pulsatile LH secretion, the ewes were ovariectomized and implanted with estradiol, and LH was measured in both frequent (every 6 min for 4 h) and infrequent (twice weekly) blood samples. In this model, high LH indicates low response to estradiol negative feedback and reproductive induction; low LH signifies high response to estradiol negative feedback and reproductive arrest. LH levels (samples twice weekly) rose some 50-fold in both groups concurrently at the start of the breeding season in September. Frequent sampling in midbreeding season (autumn) revealed that both thyroidectomized and thyroid-intact ewes exhibited frequent LH pulses, with no group difference, in either the presence or absence of the estradiol implant. A marked group difference, however, emerged at the end of the breeding season. LH fell to basal values in thyroid-intact ewes (onset of low values Feb 3 +/- 8 days), whereas levels remained elevated in thyroidectomized ewes through the end of the study (April 26). At this time, thyroidectomized ewes had more frequent LH pulses than thyroid-intact ewes both in the presence and absence of estradiol. The circadian pattern of melatonin secretion and the seasonal change in PRL were found to be unaffected by thyroidectomy. These observations support the hypothesis that the thyroid is necessary for endogenous suppression of neuroendocrine mechanisms that generate LH pulses, a suppression crucial for the transition to anestrus.

Evidence for seasonal plasticity in the gonadotropin-releasing hormone (GnRH) system of the ewe: changes in synaptic inputs onto GnRH neurons.

In the Suffolk ewe, seasonal reproductive transitions are due primarily to changes in the responsiveness of the GnRH neurosecretory system to the negative feedback influence of estradiol. As GnRH neurons in the sheep, like those in other mammals, lack estrogen receptors, the influence of estradiol on GnRH neurosecretory activity is probably conveyed via afferents. As a possible structural basis for seasonality, we examined the ultrastructure and synaptic inputs of GnRH neurons in the preoptic area of ewes during the breeding season and seasonal anestrus. GnRH neurons were examined in both ovary-intact ewes and ovariectomized ewes bearing implants that produced constant levels of estradiol to eliminate a changing hormonal milieu as a factor in any seasonal variations. We found that preoptic GnRH neurons in breeding season ewes received more than twice the mean number of synaptic inputs per unit of plasma membrane as GnRH neurons in anestrous animals. Although GnRH dendrites received more synaptic input than GnRH somas, significant seasonal differences were seen in both axodendritic and axosomatic inputs. In contrast, unidentified neurons in the preoptic area showed no significant seasonal changes in their synaptic inputs. Seasonal changes in synaptic inputs onto GnRH neurons were seen in both intact animals and ovariectomized ewes bearing estradiol implants. Consequently, these seasonal alterations are unlikely to be due to changing levels of endogenous sex steroids, but may instead reflect changes in the environmental photoperiod and/or the expression of an endogenous circannual rhythm.