| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
Laboratory of Neuroendocrinology,3 and Cognitive and Developmental Neuroscience,4 The Babraham Institute, Babraham, Cambridge CB2 4AT United Kingdom
ABSTRACT
The endogenous opioid peptides have been implicated in mediating the actions of estrogen and progesterone on GnRH release. We used in situ hybridization histochemistry to determine whether steroid-induced changes in GnRH/LH release in the female sheep are associated with changes in the cellular mRNA content of the precursors for beta-endorphin (pro-opiomelanocortin; POMC) and met-enkephalin (pre-proenkephalin; PENK). Two specific hypotheses were tested. First, that the inhibitory actions of progesterone are associated with an increase in opioid gene expression in specific hypothalamic nuclei. Our data support this hypothesis. Thus, an increase in progesterone was associated with increased POMC gene expression in the arcuate nucleus and PENK in the paraventricular nucleus. Further, the increase in POMC was restricted to regions of the arcuate nucleus that contain steroid sensitive beta-endorphin neurons. Our second hypothesis, that gene expression for the two opioid precursors would decrease prior to the start of the estradiol-stimulated GnRH surge, was not supported. Rather, POMC (but not PENK) gene expression in the arcuate nucleus was significantly higher in estradiol-treated animals than controls at the peak of the GnRH surge. These data suggest that beta-endorphin neurons in subdivisions of the arcuate nucleus and enkephalin neurons in the paraventricular nucleus are part of the neural network by which progesterone inhibits LH release. While enkephalin neurons may not play a role in estrogen positive feedback, increases in POMC mRNA in the arcuate nucleus at the time of the GnRH peak may be important for replenishing beta-endorphin stores and terminating estrous behavior.
estradiol, gonadotropin-releasing hormone, luteinizing hormone, neurotransmitters, progesterone
The ovarian steroid hormones estradiol (E) and progesterone (P) are of central importance in the control of reproductive neuroendocrine function in female mammals. These hormones exert key regulatory actions on fertility by altering the activity of the neural circuits that control the release of gonadotropin releasing hormone (GnRH). The precise patterns of GnRH release that occur in response to elevated steroid concentrations have been well studied in the ewe and demonstrate that the increasing circulating concentrations of E during the late follicular phase of the estrous cycle trigger the preovulatory GnRH surge [1–3]. During the luteal phase of the cycle, when P is the dominant ovarian steroid hormone, the frequency of GnRH pulses is markedly reduced [3–6]. In addition, P can also block the GnRH surge-induced by E [7, 8].
The mechanisms by which ovarian hormones modulate both the tonic and surge modes of GnRH release remain unclear. We do know, however, that a major mechanism by which both steroids exert their actions on GnRH release is via nuclear receptors. As ovine GnRH neurons themselves do not appear to contain receptors for either E receptor 
or P [9, 10], these steroid hormones are believed to modulate GnRH release predominantly by an indirect action via steroid receptor containing intermediate neurons [11]. The endogenous opioid peptides have been implicated in mediating the feedback actions of E that regulate the GnRH surge in several species, including the rat [12], human [13], and the ewe [14]. In this situation, opioids have been regarded as a brake that must be removed before the GnRH surge can be initiated [15]. There is also evidence that the endogenous opioid peptides are involved in the inhibitory action of P on GnRH release [16–18]. Although all three major classes of endogenous opioid peptides, the endorphins, enkephalins, and dynorphins, have been shown to influence gonadotropin secretion, the role of each specific opioid, the hypothalamic site, and mechanism by which they exert their inhibitory action are still largely unresolved. Although there is evidence that ß-endorphin is most important for estrogen positive feedback, which opioid system(s) mediates the inhibitory actions of P is less clear, although there is historic evidence for the involvement of endorphins and enkephalins as well as more recent indications of the influence of dynorphin [19, 20]. This study concentrated on the potential involvement of ß-endorphin neurons in the arcuate nucleus and enkephalin neurons in the paraventricular and ventromedial nuclei in the modulation of GnRH release. There is substantial evidence in the ewe that the arcuate and ventromedial nuclei are key hypothalamic sites involved in the relay of steroidal information to the GnRH neurons in the preoptic area [21, 22]. In addition, recent information shows that there are sex- and steroid-associated differences in enkephalin neurons in the ovine paraventricular nucleus and suggests that they may have an important role in the control of LH release [23]. In rodents, enkephalin neurons have been shown to project to GnRH neurons in the preoptic area [24], further supporting a role for these neurons in the control of the reproductive axis.
In situ hybridization histochemistry was employed to determine whether changes in the cellular content of mRNA for the precursors for ß–endorphin (pro-opiomelanocortin; POMC) and met-enkephalin (pre-proenkephalin; PENK) occur in association with E positive and P negative feedback on GnRH release. The present study had two main aims. Specifically, we tested the hypotheses that POMC or PENK gene expression decreases prior to the start of the E stimulated GnRH surge, while luteal phase concentrations of P that inhibit episodic GnRH release are associated with an increase in POMC or PENK gene expression. The studies were performed in animal models of the follicular and luteal phases respectively [5] that have been well characterized, in which ovariectomized ewes were treated with steroids using constant release hormone implants.
Two studies were carried out using reproductively mature ewes (2–6 years old) of the Clun Forest (Experiment 1) or Poll Dorset breed (Experiment 2) that were maintained in indoor accommodation, under natural lighting conditions at the Babraham Institute (Cambridge, UK, 52° 21'N). Animals were fed a diet of sheep concentrates with free access to hay and water.
Ewes were ovariectomized (OVX) and immediately given a 10-mm Silastic implant of crystalline E to maintain low physiological concentrations of this hormone (1–2 pg/ml; 5). Procedures were carried out with approval from the Welfare and Ethics Committee of the Babraham Institute and under Home Office Project License PPL 80/1037.
Experiment 1: Does the Cellular Content of POMC or PENK mRNA Decrease Prior to the Start of the E-Induced Surge of GnRH?
The 46 ewes in the study were OVX and subjected to two consecutive ‘artificial’ estrous cycles using a modification of the method described by Goodman et al. [25]. Specifically, an artificial luteal phase was created by giving the OVX ewes an intravaginal P implant for 11 days (CIDR-G, InterAg, Hamilton, New Zealand). Following removal of the CIDR (Controlled Internal Drug Release device), an artificial follicular phase was initiated 24 h later with 4 x 30mm subcutaneous implants of E to raise peak follicular phase concentrations to between 8 and 12 pg/ml. Blood samples for LH measurements were collected hourly from 8 until 40 h after E administration. These data were used to predict the time of the GnRH surge in the second artificial cycle so that ewes could be killed at precise points along the time course. The time course of the E-induced GnRH surge has been demonstrated to be highly repeatable for any individual ewe [2]. During the second artificial follicular phase the animals were randomly assigned to two groups. Twenty were given E treatment (as in cycle 1) to stimulate a GnRH/LH surge, while the remainder received sham implantation and acted as controls. One group of ewes (n = 6) was killed just before E administration (Fig. 1; pre-E). The E-treated ewes were used at one of four time points (see Fig. 1) relative to the predicted phase of the GnRH surge as follows; i) Presurge (n = 5), 8 h after E administration but some 8–10 h before the onset of the LH/GnRH surge; ii) Ascending limb (n = 5), 22.4 ± 1.5 h after the insertion of E implants and 2–6 h after the beginning of the LH/GnRH surge; iii) Midpeak (n = 5) 26.2 ± 0.7 h after E, just after the peak of the LH surge but at the expected time of the GnRH peak; or iv) Postpeak (n = 5), 40 h after E administration, when both LH and GnRH concentrations have returned to baseline. Groups (n = 5) of control ewes were killed at the same time as each group of E-treated ewes. During this artificial cycle, samples of jugular blood were collected at hourly intervals from 6 h before E administration. Additional samples were obtained from the Ascending limb and Mid-peak groups at half hourly intervals from 2 h before death to obtain a more detailed profile of LH release. This study was carried out in October/November, at the beginning of the breeding season.
|
Experiment 2: Are the Inhibitory Actions of Luteal Phase Concentrations of P on Episodic GnRH Release Associated With an Increase in the Cellular Content of POMC or PENK mRNA?
Twenty-four OVX ewes were subjected to an artificial estrous cycle immediately prior to this study (mid-June; anestrous season) and following this all steroid implants (including the 10 mm E implant) were removed. The animals were randomly assigned to three groups (n = 8 each), two of which received P treatments for 10 days while the control group remained steroid-free. Specifically, one group received two intravaginal CIDRs to raise P to early luteal phase concentrations, while the other received two CIDRs plus two Silastic packets containing P to produce mid-luteal phase concentrations of the hormone. On the day before the study (Day 9 after P administration) blood samples were collected at 15-min intervals for 12 h to determine the pattern of episodic LH release. In addition, P concentrations were determined in two samples per ewe, one selected at random from this series and one collected on day 10.
Hormone concentrations. Concentrations of LH and P (Experiment 2) were determined by radioimmunoassay in samples of plasma that had been stored at –20°C. Plasma LH was measured in duplicate 100-µl aliquots using antiserum CSU204 (kindly provided by GD Niswender) and the NIDDK S11 standard [26]. The inter- and intraassay coefficients of variation for Experiment 1 (6 assays) were 13.2% and 8.7%, respectively, and 13.0% and 8.5% (2 assays) for Experiment 2. The limit of detection was 0.2 ng/ml. Concentrations of P were determined in a single assay using a commercially available kit (Coat-a-Count, Diagnostic Products Corporation, Los Angeles, CA) in which the limit of detection was 0.1 ng/ml.
In situ hybridization. Animals were killed with an overdose of barbiturate (Lethobarb; 20mg/kg BW i.v., Duphar Veterinary, Southampton, UK), the brain removed and a block of tissue containing the hypothalamus dissected within 3 minutes of death and immediately frozen on dry ice. The hypothalamic block was stored at –70°C. Sections through the hypothalamus (15 µm) were cut in a coronal plane on a cryostat and thaw mounted (3/slide) onto Vectabond-subbed slides (Vector Laboratories Inc., Peterborough, UK). Sections were stored at –70°C until hybridized with a 35S-labeled 45-mer oligonucleotide probe complementary to the bovine cDNA sequence (PENK) or a 46-mer oligonucleotide probe complementary to the porcine cDNA sequence (POMC). Some sections from these blocks of tissue have been used in other studies that investigated whether changes in GnRH mRNA expression were associated with E-positive [27] and P-negative feedback [28].
The probes were labeled at the 3' end with (35S) deoxy-ATP (1000–15 000-Ci/mmol; New England Nuclear-Dupont, Boston, MA) using terminal deoxynucleotidyl transferase (50 U; Pharmacia, Upsalla, Sweden) and purified by filtration on a Sephadex G50 column. On the day of hybridization, slides were quickly warmed to room temperature, fixed in 4% phosphate-buffered paraformaldehyde (pH 7.4 for 20 min), washed in phosphate-buffered saline, and dried in an ascending series of alcohols. The 35S-labeled probes were diluted in hybridization buffer (20 x saline sodium citrate [SSC], 50% deionized formamide, 10% dextran sulphate, 1 x Denhardt's solution, 250 µg/ml sheared salmon testis DNA, and 0.3% ß-mercaptoethanol) to give a final concentration of about 1.2 x 103 cpm/µl; 250 µl was added to each slide. After an overnight incubation at 37°C, slides were washed thoroughly in SSC, rinsed in water, and placed in 300 mM ammonium acetate/70% ethanol solution for 30 sec, followed by absolute ethanol for 30 sec before being allowed to air dry. Slides were randomly assorted, coated in Ilford K-5 nuclear track emulsion and kept in light-secure boxes at 4°C for between 9 and 20 days (PENK) or approximately 25 days (POMC). Slides were developed in Ilford Phenisol developer, diluted 1:5 in distilled water (5 min at room temperature). After a light counterstaining with methylene blue, slides were dehydrated, dipped in xylene, and coverslipped with Ralmount mountant. Controls included slides in which the 35S-labeled probe was applied in the presence of 50-fold excess of unlabeled probe.
Slides were selected for study from three brain regions that had been shown to exhibit either POMC or PENK gene expression [29, 30] and been implicated in the control of GnRH release: the ventromedial nucleus (VMN), the paraventricular nucleus (PVN), and the arcuate nucleus (ARC). In addition, the PVN and ARC were subdivided into rostral and caudal portions and two slides (i.e., six sections) analyzed from each nuclear subdivision. The anatomical features of sections within a region were matched as closely as possible across the groups. In Experiment 1, two separate hybridizations were carried out for PENK using sections from the earliest four time points (Fig. 1). The first contained sections from the rostral and caudal PVN and the second included slides from the VMN. Further, in Experiment 1 two separate hybridizations were performed for POMC. In the first, sections up to the midpoint of the GnRH surge were included but, because of the results of this in situ study, a further hybridization was performed that comprised sections from the ARC from the Midpeak and Postpeak groups only. All sections in Experiment 2 were hybridized together.
Analysis of cellular mRNA. Cells were judged to be positively hybridized when silver grains were clustered over a methylene blue counterstained cell and the number of grains was five times greater than that over cells in the presence of excess unlabelled probe (Figs. 2 and 3). There were no differences in the numbers of mRNA expressing cells among the groups and, thus silver grain density was determined in 60 cells/brain area (selected at random) for each probe using a Seescan Sonata II image analyser (Seescan, Cambridge, UK). An average value for cellular grain density was computed for each animal and this figure used to determine group means and SEMs.
|
|
In Experiment 2, pulses of LH were identified using the Monro algorithm [31]. Differences in P and LH concentrations between treatments and LH pulse frequency were analyzed using a one-way ANOVA. In Experiments 1 and 2, significant differences in silver grain density between the steroid-treated and nonsteroid-treated groups at the individual time points studied were also determined using one-way ANOVA followed by a Tukey-Kramer multiple comparisons test. Data were not analyzed across time. Significance was considered as P < 0.05. Data are presented as mean ± SEM based on animal means.
Distribution of mRNA for POMC and PENK
After 9 days of exposure with the PENK oligonucleotide probe, dense clusters of silver grains were located over neurons primarily in the parvocellular portion of the PVN (Fig. 2) with substantially less-dense clusters overlying cells in the VMN. Therefore, a second hybridization was performed with 20 days of exposure to produce clusters of grains over the VMN that were amenable to analysis. The distribution of cells containing PENK mRNA in the current study was similar to that described previously [30]. In addition, hybridized cells were located in areas of immunocytochemically identified met-enkephalin neurons in both the sheep [32] and the rat [33, 34]. POMC mRNA expressing cells were found throughout the rostro-caudal boundaries of the ARC and in both the medial and lateral aspects (Fig. 3B). At the level of the mamillary recess, POMC mRNA expressing cells were located ventral to the ventricle. Scattered cells were occasionally found in the VMN. The distribution of cells in the ARC that expressed POMC mRNA was similar to that reported for immunoreactive ß-endorphin neurons in this region [35]. Because ß-endorphin neurons that contain estrogen receptors (ER) are located in subdivisions of the ARC [35], our analyses were focused largely on these regions. Specifically, in the more caudal portion of the nucleus, cells double labeled for ER and ß-endorphin were reported to be mainly in the central ARC (Fig. 3A), while at the level of the mamillary recess they are located close to the midline near the margin of the third ventricle. Importantly, a high proportion of these cells would be expected to contain P receptors (PR), as approximately 95% of ER immunoreactive neurons in the ARC are also immunoreactive for PR [36].
Experiment 1: Does the Cellular Content of POMC or PENK mRNA Decrease Prior to the Start of the E-Induced Surge of GnRH?
LH results. The mean concentrations of LH in the hours immediately prior to death have been published previously [27]. Five ewes were assigned to the Ascending limb group and five to the Midpeak group based on their pattern of LH release. Specifically, ewes in the former group were killed 2–6 h after the LH surge had begun, but before the peak. Ewes in the latter group were killed after LH concentrations had begun to fall but GnRH concentrations were expected to be at their peak [2].
Effects of steroid treatments on the number of detected hybridized cells. No significant differences in the number of hybridized cells were evident between the control and steroid treated ewes in any of the three areas of the brain studied. Cells were selected at random for gene expression analysis.
PENK gene expression in the PVN and VMN during the GnRH surge. There were no significant differences in the cellular content of PENK mRNA in the PVN or the VMN between the surge groups and the controls at any of the four time points studied (Table 1).
|
POMC gene expression in the ARC during the GnRH surge. The pattern of change of POMC gene expression in the rostral and caudal portions of the ARC was the same (Fig. 4), therefore the data will be discussed without reference to these separate areas. In the E treated animals there was no evidence to support our hypothesis that a fall in the cellular content of POMC mRNA in E receptive portions of the ARC occurs prior to the GnRH surge. Rather, although silver grain density remained very constant over time in the control groups, there was a tendency for grain density to be higher after E implantation in the surge group than the nonsurge controls. This reached statistical significance (P < 0.01) at the peak of the GnRH surge. Because of this unexpected finding, a second in situ hybridization was performed for POMC mRNA to confirm this result using further sections from the Midpeak and Postpeak groups only. Once again, the cellular content of POMC mRNA was significantly greater (P < 0.001) in the E treated compared with the untreated controls at the peak of the GnRH surge. However, by the end of the GnRH surge, 40 h after E administration, similar silver grain densities were observed in the control and experimental groups.
|
Experiment 2: Are the Inhibitory Actions of Luteal Phase Concentrations of P on Episodic GnRH Release Associated With an Increase in the Cellular Content of POMC or PENK mRNA?
Progesterone and LH concentrations. The P concentrations achieved by the implants are shown in Figure 5. Concentrations in the control ewes were 0.23 ± 0.05 ng/ml. Two CIDRs raised these to early luteal phase concentrations of 2.3 ± 0.2 ng/ml and an additional two progesterone ‘packets’ produced concentrations similar to those seen in the mid-luteal phase of the ovine estrous cycle (4.8 ± 0.4 ng/ml). Elevated circulating P reduced mean LH concentrations and LH pulse frequency in a dose-dependent manner (Fig. 5).
|
PENK gene expression in PVN during P negative feedback on GnRH release. After 12 days, clusters of silver grains were found over neurons in the PVN in a similar distribution to those described in Experiment 1. Specificity of the probe was confirmed by extremely strong expression in the intermediate lobe of the pituitary gland (not shown). However, silver grains over cells in the VMN were sparse and, therefore, this area was not included in further analysis. The cellular content of PENK mRNA in both the rostral and caudal PVN was significantly (P < 0.01) higher in the P treated animals than in the controls. Data from these areas have been combined for presentation (Fig. 6a). This increase in gene expression was not dependent on the concentration of P delivered by the implants and thus was similar in the groups exhibiting early and mid-luteal phase hormone concentrations.
|
POMC gene expression in the ARC during P negative feedback on GnRH release. In this study the cellular content of POMC mRNA was determined separately for the ARC subdivisions that contain either a high or a very low density of steroid receptive ß-endorphin neurons (35, Fig. 3). As similar patterns of gene expression were found in the rostral and caudal divisions of the nucleus, data from both regions have been combined for analysis and presentation. Elevated circulating P concentrations were not associated with any changes in the cellular content of POMC mRNA in the lateral portions of the nucleus that have a low density of E-receptive ß-endorphin neurons (Fig. 6c). However, POMC gene expression was significantly higher in steroid receptor–rich ARC regions in animals with early luteal phase (P < 0.001) and mid-luteal phase concentrations of P (P < 0.05) than controls (Fig. 6b). As was the case with PENK mRNA expression, the increase in the cellular content of POMC mRNA following exogenous P administration did not depend on the concentration of P achieved.
These studies show that elevated concentrations of P that inhibit GnRH and LH release are associated with an increase in mRNA expression for two opioid precursors, POMC in the ARC and PENK in the PVN. These results support one of our hypotheses, that the inhibitory actions of luteal phase concentrations of P on episodic GnRH release are associated with an increase in the cellular content of POMC or PENK mRNA. Studies with specific receptor agonists [37] and antagonists [38], and immunoneutralization [39] have shown that opioid peptides can alter GnRH release via the three main receptor subtypes, µ, 
, and 
. Although there is strong evidence that ß-endorphin pathway(s), acting largely via µ receptors, are key to the estrogenic control of the GnRH surge, the precise nature of the opioid system by which P inhibits GnRH release is more obscure. Our data support the idea that ß-endorphin neurons in the ARC and enkephalin neurons in the PVN are part of the neural circuits by which P inhibits GnRH pulse frequency. We propose that one of the mechanisms by which P inhibits GnRH release during the luteal phase of the ovine estrous cycle is to act directly on ß-endorphin neurons in the ARC to increase their content of POMC mRNA and subsequently the content and release of the inhibitory neurotransmitter. Ovine ß-endorphin neurons project directly from the ARC to the preoptic area [20] where the majority of the GnRH neurons are located [40]. A subset of these ß-endorphin neurons, located in the medial portion of the ARC, are steroid receptive [35] and these are the neurons in which P associated increases in POMC mRNA are noted. In contrast, the more lateral areas of the nucleus where there are very few, if any, steroid receptive opioid neurons did not show any change in gene expression with increased P concentrations. The link between enkephalin neurons in the PVN and GnRH neurons is less clear. However, immunoneutralization of met-enkephalin in the ovine anterior hypothalamus, that probably encompasses the PVN, has been shown to transiently increase LH release. It has been suggested that met-enkephalin neurons in this neural loci contribute to the suppressive actions of P on LH secretion [39]. Our studies would also suggest that enkephalin neurons in the PVN are part of the neural network by which P inhibits LH secretion.
The current results support an earlier in situ study by Broad et al. [29] in ovariectomized sheep that reported that POMC mRNA concentrations in the ARC are elevated at times when P concentrations are raised. However, they did not note any significant change in PENK gene expression. The reason for this difference between the two studies is unclear, but the method and length of steroid treatment in the two studies may be responsible. Another study used northern blot analysis of blocks of tissue containing the mediobasal hypothalamus to conclude that low physiological concentrations of P that decreased the amplitude and frequency of episodic LH release increased POMC mRNA content [41]. Although all three studies are in agreement that elevated P is associated with raised POMC mRNA in the ovine hypothalamus, the current study is the most complete in precise monitoring of both P and LH profiles and in defining the neural location of the increases in POMC mRNA.
Our results did not, however, support our second hypothesis, that POMC or PENK gene expression decreases prior to the start of the E-induced GnRH surge as it has been shown to do in rodents [42]. In the case of PENK, we did not observe any changes in cellular mRNA expression over the course of the study. Walsh et al. [43] also observed no differences in PENK mRNA expression in the PVN of ovary intact luteal phase ewes compared with those in the follicular phase or during the LH surge. However, they report a significant decline in PENK mRNA expression in the VMN between the luteal and follicular phase groups. Substantial differences in the timing of the collection of tissue relative to surge onset between the two studies could explain differences in our findings. Taken together, these findings are consistent with the idea that enkephalin neurons do not play a role in E positive feedback or in the control of estrous behavior.
However, contrary to our expectations, we did observe a significant increase in POMC gene expression at the time of the peak of the GnRH surge. Because this was an unexpected finding, a further in situ hybridization was performed using sections from the Midpeak and the Postpeak groups only. Once more we observed higher POMC gene expression at the time of the peak of the GnRH surge in the E treated ewes compared with the controls. By the time the GnRH surge had returned to baseline, this difference between the groups was no longer apparent. Three interpretations of these results spring to mind. First, perhaps elevated gene expression at the peak of the GnRH surge augments hypothalamic ß-endorphin concentrations, which in turn inhibit GnRH release and terminate the surge. This neural action would have no effect on any aspect of the LH surge, because this has been shown to end well in advance of the fall in GnRH [2].
Recently, studies in the ewe have supported the view that the extended period of GnRH secretion is involved in the maintenance of receptive behavior [44]. Hence ß-endorphin neurons could have a role in terminating the period of sexual behavior so that ewes are no longer receptive to advances from the male at times when mating would not result in a successful conception. In support of this conjecture, several studies in the rat have shown that ß-endorphin inhibits lordosis behavior [45] and that this effect is mediated via an action on GnRH release [46]. The second interpretation of our findings is that the increased cellular POMC mRNA reflects a mechanism for replenishing stores of ß-endorphin in the ARC. During the hours after E administration, this hormone first exerts a negative feedback action on GnRH release (about hours 2 to 12 after E in our model) before the positive feedback actions are observed (beginning about 16–20 h after E). It has been hypothesized that enhanced concentrations of ß-endorphin during the negative feedback phase must drop in order for the GnRH surge to take place [12]. Perhaps the role of enhanced POMC mRNA at the peak of the GnRH surge is to replenish the neurotransmitter following a period of enhanced neuronal activity. A third possibility is that rising POMC gene expression prior to the surge results in increased ß-endorphin release and that this acts as an inhibitory neurotransmitter to suppress the activity of a second inhibitory neural system that impinges on the GnRH neurons. Thus, the increase in POMC gene expression would result in disinhibition of the GnRH neurons and result in the GnRH surge. One possible neurotransmitter that might fill the role of an intermediary inhibitory neuron would be GABA. Steroid negative feedback on LH release has been reported to involve an increase in GABA concentrations in the ovine preoptic area, an action which is mediated by the endogenous opioid peptides [47].
In the current study, we concentrated efforts on the period temporally close to, and after, the start of the GnRH surge. Because of this, we may have missed very early, E-induced changes in opioid gene expression that may be critical for GnRH surge generation. This fact would also apply to another study that concluded that POMC mRNA levels are not reduced in the hypothalamus at the time of the GnRH surge. In this study, POMC gene expression was investigated in ovary intact ewes at a time when the GnRH surge had probably begun [48]. The fact that we may have missed other changes in POMC gene expression are highlighted by a recent study in which a very similar ovine endocrine model to ours was used to show that POMC gene expression decreased in an area encompassing the ARC prior to the GnRH surge [49]. Importantly, gene expression was determined four hours after the start of the E treatment, while our first observation was some 4 h later than this, at 8 h after steroid implantation. It is, therefore, possible that a decrease in endorphin concentration in this region is a very early event in the chain that leads from the activation of a steroid receptive neuron to the later neural events that include the activation of the GnRH neurons themselves [50]. Further studies will be needed to resolve this issue. As our study and that of Pillon and her colleagues [49] were performed in very similar animal models, we can propose the following role for ß-endorphin neurons over the course of the ovine estrous cycle. The earliest effects of E in promoting the preovulatory GnRH surge is to act on the E-responsive ß-endorphin neurons of the ARC to decrease the cellular content of POMC mRNA and hence ß-endorphin synthesis. This reduces the negative feedback influence on the GnRH neurons and enables the influences of other neurotransmitter systems to promote the surge release of GnRH. Later in the chain of events, the low concentrations of ß-endorphin need to be replenished, hence the increase of POMC mRNA, which reaches a maximum around the peak of the GnRH surge. This increase in gene expression may also play a role in terminating the GnRH surge, thus limiting the period of reproductive behavior in the ewe to a time when successful conception is possible. Following the follicular phase, POMC gene expression is raised, leading to enhanced opioid tone during the luteal phase, which would play a key role both in suppressing tonic GnRH release and blocking E positive feedback.
ACKNOWLEDGMENTS
We thank Andrew Dady, Martin White, and Tony Jones for assistance with the animal work and surgeries and Drs. Neil Evans, Tom Harris, and Allan Herbison for input on the design and execution of the studies. Dr. John Bicknell made helpful criticism of a draft of the manuscript. We also thank NIDDK for supplying both the LH for iodination and the LH standard.
FOOTNOTES
1Supported by the Biotechnology and Biological Sciences Research Council (BBSRC). M-L.G. was supported by an EU Marie Curie Research Training Grant. 
Correspondence: 2Jane E. Robinson, Division of Cell Sciences, Faculty of Veterinary Medicine, The University of Glasgow, Bearsden Road, Glasgow G61 1QH Scotland. FAX: 0141 330 5797; e-mail: j.robinson{at}vet.gla.ac.uk
Received: 12 July 2006.
First decision: 9 August 2006.
Accepted: 29 November 2006.
REFERENCES
and -receptor agonists on plasma luteinizing hormone (LH), prolactin, and the estrogen-induced LH surge in ovariectomized ewes. Endocrinology 1996; 137:3640–3648[Abstract], and the type II glucocorticoid receptor are coexpressed in the same neurons of the ovine preoptic area and arcuate nucleus: a triple immunolabelling study. Biol Reprod 2002; 67:1605–1612This article has been cited by other articles:
|
|
 |
|
  T. A. Roepke, A. Malyala, M. A. Bosch, M. J. Kelly, and O. K. Ronnekleiv Estrogen Regulation of Genes Important for K+ Channel Signaling in the Arcuate Nucleus Endocrinology, October 1, 2007; 148(10): 4937 - 4951. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
