| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and ERβ in the Modulation of Luteinizing Hormone Responses to Kisspeptin in the Female Rat: Implications for the Generation of the Preovulatory SurgeDepartment of Cell Biology, Physiology, and Immunology (J.R., E.V., J.M.C., F.G., V.M.N., E.A., L.P., M.T.-S.), University of Córdoba, and CIBER (CB06/03) Fisiopatología de la Obesidad y Nutrición, Instituto Salud Carlos III, 14004 Córdoba, Spain; and Organon NV (F.A.D., A.G.H.E., P.I.v.N.), 5340 BH Oss, The Netherlands
Address all correspondence and requests for reprints to: Manuel Tena-Sempere, Physiology Section, Department of Cell Biology, Physiology, and Immunology, Faculty of Medicine, University of Córdoba, Avda. Menéndez Pidal s/n, 14004 Córdoba, Spain. E-mail: fi1tesem{at}uco.es.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERβ in the modulation of LH responses to kisspeptin and the generation of the preovulatory surge. Selective blockade of ER in cyclic females blunted LH responses to kisspeptin, eliminated the endogenous preovulatory rise of LH, and blocked ovulation. In contrast, antagonism of ERβ failed to cause major changes in terms of LH surge and ovulatory rate but significantly augmented acute LH responses to kisspeptin. Notably, defective LH secretion and ovulation after ER blockade were not observed after GnRH stimulation, which elicited maximal acute (<2 h) LH responses regardless of ER/ERβ signaling. In addition, net LH secretion in response to kisspeptin was decreased by ovariectomy and increased after selective activation of ER but not ERβ. Altogether, our data document the prominent positive role of ER in the regulation of GnRH/LH responsiveness to kisspeptin and, thereby, ovulation. In addition, our results disclose the putative function of ERβ as negative modifier of GnRH/LH response to kisspeptin, a phenomenon that might contribute to partially restraining LH secretion at certain physiological states. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our knowledge on the neuroendocrine systems responsible for the control of gonadotropin secretion in general, and the preovulatory surge in particular, was recently revolutionized by the recognition of the fundamental reproductive roles of the ligand-receptor KiSS-1/G protein-coupled receptor (GPR) 54 system, originally identified as metastasis-suppressor (3, 4, 5). Indeed, such unsuspected facet of the system was first evidenced when inactivating mutations in the gene encoding the GPR54 were shown to induce hypogonadotropic hypogonadism in humans and mice (6, 7). Those original findings underscored the putative reproductive function of the ligands of GPR54, a family of structurally related peptides, globally termed kisspeptins, which derive from the differential proteolytic processing of a common precursor, encoded by the metastasis-suppressor KiSS-1 gene (8). The physiological relevance of kisspeptins and GPR54 in the neuroendocrine control of reproductive function has been substantiated during the last 3 yr by an ever-growing number of molecular and pharmacological studies, conducted mostly in laboratory rodents but also in the sheep and primates, including humans. Altogether, these studies have demonstrated that kisspeptins are very potent elicitors of gonadotropin secretion in mammals, mainly via direct activation of GnRH neurons, and have underscored a central role of KiSS-1 neurons at discrete hypothalamic areas as integrators and conduits for the regulatory effects of key signals in the control of gonadotropin secretion, including metabolic factors and, importantly, sex steroids (3, 4, 5, 9).
Of the latter, compelling evidence in rodents, sheep, and, recently, humans and nonhuman primates demonstrates that hypothalamic KiSS-1 neurons actively participate in mediating the negative feedback effects of sex steroids on GnRH/gonadotropin secretion (10, 11, 12, 13, 14, 15). Detailed studies in the mouse have shown that such a population of KiSS-1 neurons is located at the arcuate nucleus, in which KiSS-1 mRNA levels increase after gonadectomy and decrease after sex steroid replacement both in males and females, thus evidencing an inhibitory action of estrogen and testosterone on KiSS-1 gene expression at this hypothalamic site (11, 12). The complexity of the system, though, is demonstrated by the fact that in rodents another population of hypothalamic KiSS-1 neurons, located at the AVPV, has been shown to respond to estrogen in a diametrically opposite manner: KiSS-1 mRNA levels decrease after ovariectomy and significantly increase after estradiol replacement, via estrogen receptor (ER)-, but not ERβ, activation (12, 16). This observation provided the first hint for the potential involvement of central kisspeptins (arising from the AVPV) in the mediation of positive feedback effects of estrogen on gonadotropin secretion in the female, a contention that has been recently substantiated by the following findings: 1) the preovulatory surge of LH can be blocked by central immunoneutralization of kisspeptins (17); 2) the population of KiSS-1 neurons at the AVPV is clearly sexually dimorphic, being far more abundant in the female (in whom positive feedback takes place) than the male (16, 18); and 3) KiSS-1 neurons at the AVPV express ER (and to a lower extent ERβ) and are selectively activated at the preovulatory period (19, 20). Moreover, experimental manipulations known to abrogate the ability of estrogen to induce LH secretion in the female, such as neonatal androgenization, have been recently demonstrated to persistently reduce the number of KiSS-1 neurons at the AVPV and their ability to increase KiSS-1 gene expression in response to estrogen (16).
The above observations collectively point out to a discernible neuronal network, composed by KiSS-1 neurons at the AVPV that are transcriptionally activated by estrogen, selectively via ER, which is ideally placed for conveying the positive feedback effects of estrogen onto GnRH neurons (19). Of note, GnRH neurons do not physiologically express ER (21). As further support for this model, it has been recently demonstrated that axonal projections of KiSS-1 neurons from the AVPV physically contact GnRH neuronal cell bodies at the preoptic area in mice (18). Moreover, functional genomic studies have shown that selective neuronal ablation of ER (but not ERβ) is sufficient to abrogate the ability of estradiol to induce LH secretion, which is in line with the proposed ER-kisspeptin-GnRH pathway for the generation of the preovulatory surge of gonadotropins (22).
Despite the proven ability of estradiol (via ER) to transcriptionally activate KiSS-1 gene, the potential involvement of additional regulatory steps in the proposed estrogen/kisspeptin interaction, as trigger for the preovulatory LH surge, cannot be excluded. In this context, the ability of estrogen to modulate LH responsiveness to kisspeptin, its site(s) of action and receptor subtypes putatively involved, remain virtually unexplored. Similarly, whether ERβ might play a posttranscriptional role in the regulation of kisspeptin effects on GnRH/LH secretion has not been experimentally tested, and the potential involvement of progesterone receptors (PRs), which are crucially involved in the generation of the preovulatory surge (23), in such a phenomenon also has not been addressed. To cover these unsolved issues, we report herein detailed pharmacological analyses of LH responses to kisspeptin in the cyclic female rat at the periovulatory period after selective blockade of ER, ERβ, and PR in vivo. When relevant, these studies were complemented with analogous analyses in ovariectomized animals after replacement with selective ER, ERβ, and PR agonists, alone or in combination. Altogether, our data underscore a complex mode of action of estrogen signaling in the control of GnRH/LH secretion, which is likely to involve opposite roles of ER (stimulatory) and ERβ (inhibitory) in the modulation of central responsiveness to kisspeptin.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(ER agonist: ERA-45; ER antagonist: ERA-90) and ERβ (ERβ agonist: ERB-79; ERβ antagonist: ERB-88), as well as the antagonist of PR (ORG 31710), were selected and provided by Organon (Oss, The Netherlands). Progesterone (P) and GnRH were purchased from Sigma Chemical Co. (St. Louis, MO).
Experimental designs
To dissect out the potential roles of ER and ERβ, as well as PR, in the modulation of LH responses to kisspeptin in the cyclic female rat at proestrus (i.e. when the preovulatory surge of gonadotropins takes place), in experiment 1 the effects of an effective dose of kisspeptin-10 on LH secretion were tested after selective blockade of ER, ERβ, or PR in vivo, acutely at the preovulatory period. Adult virgin female rats, weighing 245 ± 15 g, were monitored for estrous cyclicity by daily vaginal cytology. Only rats with at least two consecutive regular 4-d estrous cycles were used for subsequent pharmacological studies. Groups of cyclic females (n = 20–24) were injected twice (at 2100 h of diestrus-2 and 0900 h of proestrus), via sc route, with one of the following compounds/dose regimens: ER antagonist ERA-90 (1.5 mg/kg); ERβ antagonist ERA-88 (1.5 mg/kg); or PR antagonist Org 31710 (8 mg/kg). Rats injected with vehicle (5% mannitol + 0.5% gelatin in sterile deionized water) served as controls. Subsequently each group was split in two halves (n = 10–12) that received, at 1200 h of proestrus, a single intracerebroventricular (i.c.v.) injection of either 1 nmol kisspeptin-10 or sterile saline (10 µl). The experiment was repeated twice. The protocol of i.c.v. administration was as described in detail elsewhere (10, 24). To allow delivery of kisspeptin into the lateral cerebral ventricle, the animals were implanted with i.c.v. cannulae lowered to a depth of 4 mm beneath the surface of the skull; the insert point was 1 mm posterior and 1.2 mm lateral to bregma. The dose of 1 nmol kisspeptin in 10 µl per rat was selected on the basis of our recent data on the ability of this dose to potently elicit LH secretion in adult female rats, at different stages of the estrous cycle (24). Blood samples (250 µl) were obtained by jugular venipuncture before (0 min), and at 15, 60, 120 , 210, 300, 390, and 480 min after i.c.v. injections, following previously published protocols (24). Additional blood samples were taken from each animal between 0900 and 1000 h of the following estrus, when the rats were decapitated and ovary and oviduct samples were routinely collected for assessment of ovulation, as described in detail elsewhere (25).
In experiment 2, the effects of an effective dose of GnRH on LH secretion were tested at proestrus in adult female rats after selective blockade of ER, ERβ, or PR in vivo acutely at the preovulatory period. This experiment was conducted in parallel to experiment 1, thus allowing direct comparison between the LH-releasing effects of kisspeptin-10 and GnRH in the above models of ER/PR antagonism. Protocols of acute ER, ERβ, and PR antagonism at the preovulatory period, as well as serial blood sampling (except at 300 and 480 min), were similar to those of experiment 1. An effective dose of GnRH (1 µg/rat in 100 µl sterile saline), delivered via the ip route, was selected based on previously published data (26). Independent control groups (injected sc with 5% mannitol/0.5% gelatin; injected ip with sterile saline) were included in this experiment. Because experiments 1 and 2 were conducted in parallel and hormonal data from control groups from both experiments were remarkably similar, these results were combined for joint presentation of results, as described in detail below. Monitoring of the ovulatory rate was included in all groups, as conducted in experiment 1.
In experiment 3, the effects of an effective dose of kisspeptin-10 on LH secretion were tested at proestrus, in adult female rats after a protracted protocol of blockade of ER or ERβ in vivo. Experimental procedures were similar to those of experiment 1, except for the fact that sc injections of the antagonists of ER or ERβ were continued every 12 h for a complete cycle, starting at 2100 h diestrus 2 and finishing at 0900 h proestrus of the next cycle (n = 10 injections over 4.5 d). Intracerebral injections of kisspeptin-10 (1 nmol) were conducted at 1200 h proestrus, as described for experiment 1. Blood samples (250 µl) were obtained by jugular venipuncture before (0 min), and at 15, 60, 120, 210, 390, and 480 min after i.c.v. injections. Additional blood samples were taken from each animal between 0900 and 1000 h of the after estrus, when the rats were decapitated and ovary and oviduct samples were routinely collected for assessment of ovulation.
Finally, in experiment 4, the potential roles of ER, ERβ, and PR signaling in the modulation of LH responses to kisspeptin in the adult female rat were further explored in a model of ovariectomy and selective replacement of ER, ERβ, or PR agonists. Regularly cycling adult virgin female rats were subjected to bilateral ovariectomy (OVX), under ether anesthesia, at random stages of the estrous cycle. One week after surgery, the animals (n = 10–12/group) were subjected to a protocol of 4-d replacement with via sc route, with one of the following compounds/dose regimens: ER agonist ERA-45 (0.1 mg/kg every 12 h); ERβ agonist ERA-79 (0.5 mg/kg every 12 h); or combination of both (ER+ERβ agonists). OVX rats (n = 12) injected with vehicle (5% mannitol/0.5% gelatin in sterile deionized water) served as controls. In addition, groups of OVX females (n = 10–12 per group) were implanted with SILASTIC brand silicon tubing (Dow Corning, Midland, MI) elastomers (20 mm length; inner diameter, 0.062 cm; exterior diameter, 0.125 cm) containing P and were subjected to a protocol of 4-d replacement with vehicle, ER agonist, or ERβ agonist, as described above. Selection of capsule length for P implants was based on previous physiological studies in the OVX female rat (24). Finally, a group of gonadal-intact, cyclic female rats at diestrus 1 (n = 10) was included in the study. Procedures of i.c.v. injection of 1 nmol kisspeptin-10 were as described for experiments 1 and 3. Blood samples (250 µl) were obtained by jugular venipuncture before (at –30 and 0 min; for determination of basal preinjection levels), and at 15, 60, 120, and 240 min after i.c.v. administration of kisspeptin.
The compounds used for selective activation (agonists) or blockade (antagonists) of ER, ERβ, and PR were chosen on the basis of their biological profiles in different heterologous cell systems and in vivo tests (see Appendix A; all appendices appear in the supplemental data published on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org.). Selection of the protocols (doses/regimens) of administration agonists and antagonists of ER, ERβ, and PR was based on pharmacokinetic data provided by the supplier (Organon; see Appendix A).
LH measurement by specific RIA
Serum LH levels were determined in a volume of 25–50 µl using a double-antibody method and RIA kits kindly supplied by the National Institutes of Health (Dr. A. F. Parlow, National Institute of Diabetes and Digestive and Kidney Diseases, National Hormone and Peptide Program, Torrance, CA). Rat LH-I-9 was labeled with 125I using Iodo-gen tubes, following the instructions of the manufacturer (Pierce, Rockford, IL). Hormone concentrations were expressed using reference preparations LH-RP-3 as standard. Intra- and interassay coefficients of variation were less than 8 and 10%, respectively. The sensitivity of the assay was 5 pg/tube. Accuracy of determinations was confirmed by assessment of rat serum samples of known hormone concentrations used as external controls.
Presentation of data and statistics
Hormonal determinations were conducted in duplicate, with a minimal total number of 10–12 samples/group. When appropriate, in addition to individual time point measurements, integrated LH secretory responses were calculated as the area under the curve (AUC), following the trapezoidal rule. Detailed description of the procedure for estimation of net LH secretion in response to kisspeptin or during the preovulatory period is provided in Results. Hormonal data are presented as mean ± SEM. Results were analyzed for statistically significant differences using single or repeated ANOVA followed by Student-Newman-Keuls multiple range test (SigmaStat 2.0; Jandel Corp., San Rafael, CA). P 
0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, ERβ or PR: effect of kisspeptin or GnRH, ERβ, or PR during the preovulatory phase on the magnitude of LH responses to exogenous kisspeptin at proestrus, the generation of the LH surge and the ovulatory rate were explored in cyclic female rats. Treatment with a selective antagonist for ER at late diestrus 2 (2100 h) and early proestrus (0900 h) resulted in a modest but significant increase in basal LH levels (1200 proestrus) and decrease in uterus weight at early estrus. In contrast, a similar treatment protocol with a selective ERβ antagonist failed to modify basal serum levels of LH or uterus weight at estrus, whereas administration of a PR antagonist at diestrus-proestrus significantly increased basal circulating LH levels but did not change uterus weight at estrus (Appendix B).
Serial blood sampling along the afternoon of proestrus in cyclic female rats injected with vehicle demonstrated the occurrence of the expected preovulatory surge, with low circulating levels of LH before 1400 h and a progressive increase thereafter that peaked between 1700 and 2000 h and declined subsequently. In this group, i.c.v. injection of 1 nmol kisspeptin-10 at 1200 h proestrus evoked a robust LH secretory response between 1200 and 1400 h (peak levels at 1300 h) that was followed by an augmentation and forward shift of the endogenous preovulatory surge of LH that reach maximal levels at 1700 h and declined thereafter. Of note, injection of an effective dose of GnRH to control females at 1200 h proestrus induced maximal LH secretory responses, whose magnitude exceeded that elicited by kisspeptin administration at the acute phase. Indeed, maximal LH levels were obtained from 15 min after GnRH injection onward, with a plateau for the effects of GnRH on LH secretion, which was detected up to 390 min after administration of GnRH (Fig. 1A
).
|
during late diestrus 2 and early proestrus, the generation of the preovulatory surge of LH was totally prevented, as demonstrated by serial blood sampling along the afternoon/evening of proestrus. Moreover, antagonism of ER resulted in a significant attenuation of the magnitude of acute LH responses to i.c.v. injection of 1 nmol kisspeptin-10 at proestrus; a treatment that failed also to rescue the endogenous surge of LH at late proestrus. In clear contrast, administration of GnRH at 1200 h proestrus after blockade of ER resulted in maximal LH responses, which were similar at the acute phase to those obtained in control cyclic females. Yet in this group, LH levels after GnRH administration peaked at 15 min, remained at high (maximal) levels up to 120 min, and abruptly decreased thereafter. Nonetheless, after GnRH injection, LH concentrations at 1830 h proestrus remained over those of females treated with ER antagonist and i.c.v. injected with vehicle or kisspeptin-10 (Fig. 1B
).
Treatment of female rats with a selective antagonist for ERβ during late diestrus 2 and early proestrus failed to induce major significant changes in the profile of the preovulatory LH surge, which rose from 1400 h proestrus onward, peaked at 1830 h and declined thereafter. In this model of acute antagonism of ERβ, i.c.v. injection of 1 nmol kisspeptin-10 at 1200 h proestrus evoked maximal LH secretory responses between 1200 and 1400 h (peak levels at 1300 h) that was followed by a surge of LH that closely paralleled the circulating levels of the endogenous preovulatory peak in this group. Worth noting, injection of an effective dose of GnRH at 1200 h proestrus to females pretreated with the ERβ antagonist induced maximal LH secretory responses, which equaled those elicited by kisspeptin administration in this group (Fig. 1C).
Finally, administration of a selective PR antagonist to cyclic females at late diestrus 2 and early proestrus induced an unambiguous elevation of basal LH levels but significantly attenuated the subsequent rise in circulating LH at the time of the preovulatory surge (1830 h proestrus). In this model of acute PR antagonism, i.c.v. injection of 1 nmol kisspeptin-10 at 1200 h proestrus evoked consistent LH secretory responses between 1200 and 1400 h, which peaked at 15 min and declined thereafter and were followed by a severely attenuated endogenous preovulatory surge of LH. In addition, administration of GnRH at 1200 h proestrus after blockade of PR resulted in maximal LH responses, which clearly exceeded at the acute phase those evoked by kisspeptin-10. Nonetheless, in contrast to the pattern of response to GnRH in cyclic control females, LH levels peaked at 1300 h and sharply declined thereafter, with LH concentrations similar to those of corresponding controls at 1530 h proestrus (Fig. 1D).
Considering the profiles of LH secretion described above (in response to kisspeptin and/or the endogenous LH surge), the secretory mass of LH peaks evoked by kisspeptin administration (namely, first peak between 1200 and 1400 h proestrus; i.e. within 2 h after i.c.v. injection of kisspeptin) and during the endogenous preovulatory surge (namely, second peak between 1400 and 2000 h proestrus) were calculated as AUC using the trapezoidal rule and comparatively evaluated. In addition, acute responses to GnRH (between 1200 and 1400 h after administration) were similarly calculated and compared with the magnitude of peaks induced by kisspeptin in the different experimental groups. As shown in Fig. 2, integrated LH secretion between 1200 and 1400 h proestrus was low in animals i.c.v. injected with vehicle regardless of the pretreatment, yet a modest but significant elevation was observed in females pretreated with PR antagonist. In clear contrast, i.c.v. injection of kisspeptin-10 evoked robust acute LH secretory responses, except for females pretreated with ER antagonist in which only a modest elevation was observed. Nonetheless, the magnitude of such LH peaks significantly differed among groups, with maximal secretion in females treated with the ERβ antagonist and similar (but lower) responses in vehicle- and PR antagonist-pretreated animals. Finally, injection of GnRH elicited similarly maximal LH responses between 1200 and 1400 h proestrus, which were significantly higher than those evoked by kisspeptin-10 in the corresponding groups, except for females pretreated with the ERβ antagonist, which showed similar responses to GnRH and kisspeptin administration.
|
, the magnitude of the endogenous LH surge (namely, second peak) was dramatically reduced in females pretreated with the ER antagonist (
10% of values in control females), whereas it remained unchanged after ERβ antagonism. In addition, blockade of PR at the preovulatory period significantly attenuated the magnitude of the LH surge (30% of values in control females). Intracerebral injection of kisspeptin-10 at 1200 h significantly augmented the net LH secretory mass at the second peak in control rats and females treated with the ERβ antagonist, whereas it did not enhance low LH secretion during the afternoon/evening of proestrus in females pretreated with the antagonist of ER. As was the case in control rats, kisspeptin administration significantly enhanced the magnitude of the endogenous surge of LH in females after blockade of PR, yet the magnitude of such a surge did not reach the values of reference controls (Fig. 3).
|
completely prevented ovulation, a failure that was not rescued by i.c.v. administration of kisspeptin at 1200 h proestrus. Yet ovulation was successfully induced in ER-treated animals by exogenous administration of GnRH, with a number of oocytes released per ovary similar to that of control females (Appendix C).
Preovulatory LH secretion after chronic blockade of ER or ERβ: effects of kisspeptin
Additional experiments were performed to explore the consequences of a protracted protocol of blockade of ER or ERβ on the magnitude of LH responses to exogenous kisspeptin at proestrus, the generation of the LH surge, and the ovulatory rate. Treatment with a selective antagonist for ER for a complete cycle, from late diestrus 2 (2100 h) to early proestrus (0900 h) of the next cycle, resulted in a significant increase in basal LH levels (1200 h proestrus; < 3 h after last injection) and a decrease in uterus weight at the end of treatment. In contrast, a similar treatment protocol with a selective ERβ antagonist did not change basal serum levels of LH or uterus weight (Appendix D).
Control cyclic female rats displayed the expected rise in serum LH levels at the afternoon/evening of proestrus, with peak levels at 1830 h. In keeping with data from acute experiments, i.c.v. injection of 1 nmol kisspeptin-10 at 1200 h proestrus elicited consistent LH secretory bursts between 1200 and 1400 h (peak levels at 1300 h), followed by the enhancement of the endogenous preovulatory LH surge that reached maximal levels at 1830 h and declined thereafter (Fig. 4A). Like acute blockade, chronic treatment with the antagonist of ER prevented the generation of the preovulatory surge of LH and significantly attenuated the magnitude of acute LH responses to kisspeptin. Moreover, i.c.v. injection of kisspeptin-10 in this model failed to rescue the endogenous surge of LH at late proestrus (Fig. 4B). Finally, antagonism of ERβ for a complete cycle did not induce major significant changes in the profile of the preovulatory LH surge, which peaked at 1830 h and declined thereafter, nor did it prevent acute LH responses to intracerebral administration of kisspeptin. On the contrary, as was the case in acute experiments, in this model of chronic blockade of ERβ, central injection of 1 nmol kisspeptin-10 at 1200 h proestrus induced maximal LH secretory responses between 1200 and 1400 h (peak levels at 1300 h), followed by a peak of LH that was similar in magnitude to the endogenous preovulatory surge in this group (Fig. 4C).
|
antagonist, whereas persistent blockade of ERβ significantly augmented the mass of LH secretory bursts in response to kisspeptin over control values (Fig. 5). Concerning the magnitude of the endogenous LH surge (namely, second peak) this was markedly decreased in females treated for a complete cycle with the ER antagonist (10% of values in control females), whereas ERβ antagonism did not alter the mass of the preovulatory surge. As was the case for acute experiments, intracerebral injection of kisspeptin-10 at 1200 h proestrus enhanced the net LH secretory mass at the second peak in control females but not in rats pretreated with the ERβ antagonist. Of note, injection of kisspeptin failed to enhance low LH secretion during the afternoon/evening of proestrus in females chronically pretreated with the antagonist of ER (Fig. 6).
|
|
fully prevented ovulation, which did not occur even after i.c.v. administration of an effective dose of kisspetin-10 at 1200 h proestrus (Appendix E).
LH secretory responses to kisspeptin after selective activation of ER, ERβ, or PR in OVX rats
As complement to studies of pharmacological antagonism in cyclic female rats described above, experiments were performed in OVX rats subjected to replacement protocols with selective agonists of ER and ERβ, in the presence or absence of P. OVX rats without steroid replacement and cyclic females at diestrus 1 (D-1) served as controls. Gonadectomy resulted in a significant decrease in uterus weight and a robust increase in basal LH levels over control values in D-1 females. Replacement with a selective agonist of ER, but not of ERβ, fully prevented both responses, whereas administration of P marginally increased uterus weight and partially decreased serum LH concentrations over OVX values. Combined administration of the agonist of ER and P evoked responses similar to those of ER alone, except for a higher increase in uterus weight, whereas coadministration of ERβ agonist and P was more effective than ERβ alone in (modestly) decreasing serum LH concentrations. Finally, cotreatment with ER and ERβ agonists induced responses in terms of uterus weight and serum LH similar to those of ER alone (Appendix F).
LH responses to central injection of kisspeptin were assessed in the models described above. As general protocol, blood samples were taken at two times before i.c.v. injection of kisspeptin-10 to define basal preinjection LH levels, and serial blood sampling was continued up to 240 min after central kisspeptin administration. The time-course profiles of LH secretion in the different experimental groups are depicted in Fig. 7. Considering that basal LH levels dramatically varied among the experimental groups (e.g. between cyclic D-1, OVX, and OVX+ER females), and to estimate changes in net LH secretory responses to kisspeptin administration (i.e. over corresponding preinjection levels), the mass of LH secretion was calculated for each group as AUC for the 240 min period after kisspeptin injection, after subtraction of basal preinjection levels (depicted in Fig. 7, dotted lines). Only hormonal values above the preinjection concentrations were considered for analyses. As shown in Fig. 7, net LH secretory responses to kisspeptin-10 were markedly decreased after gonadectomy but significantly enhanced (over control D-1 values) after replacement with the selective ER agonist. In contrast, administration of the ERβ agonist to OVX females resulted in further decrease of integrated net LH secretory responses to kisspeptin, whereas coadministration of ER and ERβ agonist induced a significant 30% reduction of LH secretion (vs. animals injected with ER alone) during the first hour after kisspeptin administration. Finally, administration of P alone failed to change net LH secretory responses to kisspeptin in OVX rats and marginally (but significantly) enhanced LH secretion after kisspeptin injection in rats cotreated with the ERβ agonist. In clear contrast, coadministration of the ER agonist and P resulted in further increases in net LH responses to kisspeptin (vs. rats treated with ER alone). Indeed, despite effective suppression of basal LH levels, the combined administration of ER and P to OVX rats resulted in maximal net LH responses to kisspeptin, with LH secretory mass approximately 2.5-fold higher than that observed in D-1 controls and greater than 6-fold higher than in OVX animals.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERβ, as well as PR, signaling in the modulation of LH responses to kisspeptin and the generation of the LH surge in the female rat. The most salient findings of our experiments can be summarized as follows: 1) ER, but nor ERβ, is indispensable for the generation of the preovulatory surge of LH and ovulation; 2) blockade of ER severely blunted LH responsiveness to kisspeptin, whereas antagonism of ERβ significantly enhanced acute LH secretory responses to kisspeptin at the preovulatory period; 3) selective activation of ER, but not ERβ, robustly stimulated LH responses to kisspeptin in OVX rats, with maximal responses being observed in the presence of P; and 4) defective LH responses and ovulation after blockade ER could be overcome by administration of GnRH, which maximally elicited acute LH secretion regardless of ER signaling. Altogether, our present findings document a pivotal, positive regulatory role of estrogen/ER signaling in the control of GnRH/LH responsiveness to kisspeptin and thereby the preovulatory LH surge. In addition, these results are the first to disclose the putative function of ERβ as negative modifier of LH responses to kisspeptin, a finding with potential implications for the feedback control of GnRH/gonadotropin secretion.
Whereas the involvement of estrogen in the generation of preovulatory surge of gonadotropins in the cyclic female has been long recognized (1, 2), two major recent developments have significantly expanded our knowledge of the molecular and neuroendocrine basis for such a key reproductive event. First, by an elegant series of functional genomic studies, Wintermantel et al. (22) recently demonstrated that neuronal ER, but not ERβ, is mandatory for the generation of the LH surge in mice. Yet these studies did not explore the joint contribution of PR to this phenomenon and involved the use of mouse models of constitutive genetic inactivation of ERs, thus precluding the analysis of acute, rather than developmental, effects (22). Second, it has been recently demonstrated that a population of neurons expressing the KiSS-1 gene, located at the AVPV, are crucially involved in the generation of the LH surge, given that kisspeptins are potent elicitors of LH secretion in the cyclic female (24), that KiSS-1 neurons at this site are selectively activated during the preovulatory period (19), and that blockade of endogenous kisspeptins prevents the preovulatory LH surge (17). Of particular note, KiSS-1 neurons at the AVPV have been proven to express ER (and, to a lesser extent, ERβ) and are activated (c-fos expression) by protocols of estrogen priming that are able to evoke LH surges in OVX females (19). Moreover, estrogen has been demonstrated to induce KiSS-1 mRNA expression at the AVPV via ER (12). Finally, it was recently shown that KiSS-1 neurons arising from the AVPV project to and physically contact with GnRH neurons in the forebrain of the female mouse (18). Altogether, the above findings provide the basis for a neuronal circuit responsible for conveying the positive feedback effects of estrogen onto GnRH neurons, thereby contributing to the generation of the preovulatory surge of gonadotropins. In this theoretical network, estrogen would operate as potent stimulus for KiSS-1 gene transcription at the AVPV, which in turn would increase kisspeptin input on GnRH neurons. This model, however, does not consider the possibility of additional regulatory steps for such an estrogen/kisspeptin interaction in the generation of the preovulatory surge of gonadotropins.
Our current data of acute and chronic pharmacological antagonism of ER confirm the indispensable role of this receptor subtype for the expression of the surge of LH leading to ovulation and demonstrate for the first time that its short-term blockade (<12 h preceding the rise of endogenous estradiol) is sufficient to fully prevent such preovulatory LH peak and oocyte release. More intriguingly, our results also show that pharmacological inactivation of ER (even in such a short-term basis) evoked a dramatic decrease in terms of LH responsiveness to central injection of kisspeptin, which resulted in acute responses that were less than 35% of those in control animals. Such a positive role of ER signaling in the modulation of LH responses to kisspeptin was fully confirmed by experiments in OVX rats, in which elimination of ovarian hormones resulted in a significant decrease (>2.5-fold) in net LH responses to kisspeptin, whereas selective replacement with ER agonist induced a state of increased responsiveness to kisspeptin that was maximal after cotreatment with P. Indeed, combined administration of ER agonist and P to OVX rats resulted in exaggerated LH responses to kisspeptin that were approximately 2.5-fold higher than those in cyclic D-1 controls. Altogether, the above findings are the first to demonstrate that estrogen signaling, via classical ER, plays a critical role in not only regulating KiSS-1 gene transcription at the AVPV but also increasing GnRH/LH responsiveness to kisspeptin. The relevance of the latter phenomenon for the generation of the preovulatory LH surge is illustrated by the fact that exogenous administration of an effective dose of kisspeptin-10, which should overcome the lack of endogenous activation of KiSS-1 gene expression at the AVPV due to ER blockade, failed to induce a genuine preovulatory surge of LH and did not rescue ovulation. The observation that, in contrast, administration of a bolus of GnRH to females rats pretreated with ER antagonist evoked fully maximal acute LH responses and restored ovulation demonstrates that the key permissive role of ER signaling on LH responsiveness to kisspeptin must take place at a step upstream the pituitary, likely at the hypothalamic level via modulation of GnRH release after kisspeptin stimulation. The latter is supported by previous observations that GnRH neurons do express GPR54, and are activated by kisspeptin in rodents (27). Nonetheless, the fact that female rats treated with ER antagonist did not display a normal endogenous preovulatory LH surge, even after GnRH injection, suggests that, in addition to its hypothalamic actions, ER signaling at the pituitary is also critically involved in the positive modulation of LH secretion during the preovulatory period, in keeping with previous references (28).
In contrast to the clear-cut positive role of ER, the function of ERβ signaling in the modulation of LH responses to kisspeptin appeared less discernible. Yet compelling evidence arising from our experiments suggests that ERβ might actually play a role opposite to that of ER because acute LH secretion after kisspeptin administration was unambiguously enhanced (rather than decreased) in female rats after short- and long-term blockade of ERβ. This observation unveils the putative inhibitory function of ERβ signaling in the control of GnRH/LH responsiveness to kisspeptin; a possibility that was further supported by the fact that treatment of OVX female rats with the ERβ agonist moderately, but significantly, decreased the magnitude of net LH responses to kisspeptin, even in the presence of ER agonist. Admittedly, the impact of such an inhibitory role of ERβ appears to be modest and could not be evidenced by our pharmacological antagonist approach in terms of basal LH secretion or the endogenous preovulatory surge of LH. Notwithstanding, comparative analysis of LH responses to kisspeptin and GnRH in control and ER-antagonist-treated groups reinforced the contention that ERβ might oppose to ER and contribute to partially restrain GnRH/LH secretion in response to kisspeptin in certain physiological states. Thus, GnRH administration evoked maximal acute LH secretory responses in all groups, regardless of ER signaling. However, whereas in control females LH responses to GnRH were significantly higher than those evoked by kisspeptin, blockade of ERβ resulted in similarly maximal LH secretory responses to GnRH and kisspeptin, the latter being similar to those elicited by GnRH in control animals. This observation suggests that responsiveness to kisspeptin is negatively modulated at the preovulatory period via activation of ERβ. Of note, we have not found evidence for such a suppressed responsiveness to kisspeptin via ERβ at other stages of the ovarian cycle, such as diestrus (our unpublished observations). Taken together with data of ER antagonism, it is tempting to hypothesize that the rise of estradiol at the preovulatory period induces, via ER, a state of high responsiveness to kisspeptin that is partially counteracted by the concomitant activation of ERβ. This mechanism might serve a role for the autolimitation of estrogen-induced LH responsiveness to kisspeptin at proestrus, a phenomenon that could contribute to optimize the secretory mass of LH peak preceding ovulation.
Our studies are also the first to demonstrate a divergent role of PR signaling in the control of LH responses to kisspeptin and the endogenous LH surge at the preovulatory period. In contrast to ER, blockade of PR did not significantly modify the magnitude of acute LH responses to kisspeptin or GnRH, which were grossly similar to those of control animals (see Fig. 2). Conversely, as was the case for females pretreated with ER antagonist, the secretory mass of the endogenous LH surge was significantly attenuated after antagonism of PR and could not be normalized even after kisspeptin administration (see Fig. 3). The latter might reflect the impact of blockade of PR at the pituitary because functional PRs at the gonadotrophs are mandatory for GnRH self-priming and the preovulatory LH surge (23). In contrast, the lack of impact of PR antagonist on acute LH responses might indicate that PR is not significantly involved in the modulation of GnRH secretion in response to kisspeptin, a possibility that is yet to be confirmed.
The question arising from our studies is at what hypothalamic site(s) ER and ERβ might conduct their opposite modulatory effects on kisspeptin-induced LH secretion. This is particularly relevant, given the ongoing debate on the mode of action (direct vs. indirect) of estrogen on GnRH neurons. Thus, whereas it is generally accepted that GnRH neurons do not physiologically express ER, it is also assumed that they possess ERβ (21). Yet ER is the dominant receptor for mediating the positive feedback effects of estrogen on LH release, whereas the role, if any, of ERβ in this particular aspect of gonadotropin secretion remains obscure (22). Accordingly, it has been proposed a predominant indirect mode of action of estrogen on GnRH neurons, via primary effects on glial and/or neuronal afferents to GnRH cells (21). Considering its stimulatory role on GnRH/LH responsiveness to kisspeptin reported herein, it is tempting to hypothesize that blockade of ER signaling in some of those afferent pathways may render GnRH neurons less responsive to an exogenous (or endogenous) bolus of kisspeptin. The nature of such afferents remains to be elucidated, yet it is possible that kisspeptin input itself, before the rise of KiSS-1 expression at late proestrus, might contribute to the setting of GnRH responsiveness to the peak of kisspeptin that apparently takes place as surge generator. Anyhow, it is remarkable that, even by the short-term blockade of ER (<12 h before estrogen surge), LH responses to kisspeptin were severely blunted and the endogenous preovulatory surge and ovulation abrogated. In clear contrast, antagonism of ERβ (at GnRH neurons and/or their different afferents) did not compromise the generation of the preovulatory LH surge but rather enhanced acute responses to kisspeptin. The site for the latter effect remains unexplored, but given the expression and functionality of ERβ in GnRH neurons (29, 30), the possibility that this receptor subtype, directly in GnRH cells, might contribute to attenuate their response to kisspeptins warrants further investigation.
Besides their potential physiological relevance, our present findings may pose interesting pharmacological implications in terms of manipulation of female gonadotropic axis and ovulation. Among those, it is striking that blockade of ER shortly before ovulation (30-h) was sufficient to fully prevent the preovulatory surge of LH and oocyte release. Moreover, this manipulation was able to significantly reduce acute LH responses to exogenous kisspeptin, despite the fact that KiSS-1 peptides have been proven as (one of) the most potent elicitors of GnRH/gonadotropin secretion in a wide diversity of species and functional states of the reproductive axis (3, 4). On the other hand, the observation that antagonism of ERβ, even on a short-term basis, was able to enhance LH responses to kisspeptin at the preovulatory period is also of interest and might be of help to refine protocols of hyperstimulation of gonadotropin secretion. Finally, it is also remarkable that, despite the use of maximally effective doses, LH responses to kisspeptin in cyclic females were consistently lower than those elicited by GnRH, a finding that we previously reported also in the male rat (26). Overall, it is tempting to suggest that kisspeptin, by inducing the secretion of the endogenous releasable pool of GnRH (rather than the pharmacological activation of GnRH receptors at the pituitary), might constitute an optimal physiological procedure for stimulation of gonadotropin secretion.
In summary, kisspeptins and their receptor GPR54 have recently emerged as major gatekeepers in the central control of the reproductive axis, thus forcing us to identify the role and physiological relevance of this new signaling system within the classical neuroendocrine networks controlling key aspects of reproduction, such as puberty and ovulation (3). On the latter, compelling evidence in rodents indicates that a population of KiSS-1 neurons at the AVPV, which respond to estrogen with an increase in KiSS-1 gene transcription, is crucially involved in the generation of the preovulatory LH surge. Our present findings enlarge those original observations and provide novel functional evidence for the indispensable role of central ER signaling in mediating the positive roles of estrogen on GnRH/LH responsiveness to kisspeptin, the preovulatory surge, and ovulation. In addition, our data are the first to disclose a discernible role of ERβ in modulating a putative component of feedback actions of estrogen, namely GnRH/LH responses to kisspeptin. Altogether, our current data further document the multifaceted nature and physiological relevance of estrogen/kisspeptin interactions for the generation of the preovulatory surge of gonadotropins, i.e. the hormonal trigger for ovulation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: F.A.D., A.G.H.E., and P.I.v.N. are employed by Organon NV; M.T.-S. is a recipient of grant support from Organon NV; other authors have nothing to declare.
First Published Online January 3, 2008
Abbreviations: AUC, Area under the curve; AVPV, anteroventral periventricular; D-1, diestrus 1; ER, estrogen receptor, GPR, G protein-coupled receptor; i.c.v., intracerebroventricular; OVX, ovariectomy; P, progesterone; PR, P receptor.
Received November 8, 2007.
Accepted for publication December 27, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERβ on progesterone receptor expression and action and on gonadotropin and prolactin secretion in the rat. Neuroendocrinology 79:247–258[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  G. Neal-Perry, D. Lebesgue, M. Lederman, J. Shu, G. D. Zeevalk, and A. M. Etgen The Excitatory Peptide Kisspeptin Restores the Luteinizing Hormone Surge and Modulates Amino Acid Neurotransmission in the Medial Preoptic Area of Middle-Aged Rats Endocrinology, August 1, 2009; 150(8): 3699 - 3708. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. M. Navarro, M. A. Sanchez-Garrido, J. M. Castellano, J. Roa, D. Garcia-Galiano, R. Pineda, E. Aguilar, L. Pinilla, and M. Tena-Sempere Persistent Impairment of Hypothalamic KiSS-1 System after Exposures to Estrogenic Compounds at Critical Periods of Brain Sex Differentiation Endocrinology, May 1, 2009; 150(5): 2359 - 2367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Veiga-Lopez, O. I. Astapova, E. F. Aizenberg, J. S. Lee, and V. Padmanabhan Developmental Programming: Contribution of Prenatal Androgen and Estrogen to Estradiol Feedback Systems and Periovulatory Hormonal Dynamics in Sheep Biol Reprod, April 1, 2009; 80(4): 718 - 725. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.K. Roseweir and R.P. Millar The role of kisspeptin in the control of gonadotrophin secretion Hum. Reprod. Update, March 1, 2009; 15(2): 203 - 212. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Roa, E. Vigo, J. M. Castellano, F. Gaytan, D. Garcia-Galiano, V. M. Navarro, E. Aguilar, F. A. Dijcks, A. G. H. Ederveen, L. Pinilla, et al. Follicle-Stimulating Hormone Responses to Kisspeptin in the Female Rat at the Preovulatory Period: Modulation by Estrogen and Progesterone Receptors Endocrinology, November 1, 2008; 149(11): 5783 - 5790. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
