Recent Discoveries on the Control of GnRH Neurons in Nonhuman Primates

Abstract

Since Ernst Knobil proposed the concept of the GnRH pulse-generator in the monkey hypothalamus 3 decades ago, we have made significant progress in this research area with cellular and molecular approaches. First, an increase in pulsatile GnRH release triggers the onset of puberty. However, the question of what triggers the pubertal increase in GnRH is still unclear. GnRH neurons are already mature before puberty, but GnRH release is suppressed by a tonic GABA inhibition. Our recent work indicates that blocking endogenous GABA inhibition with the GABA_A receptor blocker, bicuculline, dramatically increases kisspeptin release, which plays an important role in the pubertal increase in GnRH release. Thus, an interplay between the GABA, kisspeptin, and GnRH neuronal systems appears to trigger puberty. Second, cultured GnRH neurons derived from the olfactory placode of monkey embryos exhibit synchronized intracellular calcium, [Ca²⁺]_i, oscillations and release GnRH in pulses at ~60 min intervals after 14 days in vitro (div). During the first 14 div GnRH neurons undergo maturational changes from no [Ca²⁺]_i oscillations and little GnRH release to the fully functional state. Recent work also shows GnRH mRNA expression increases during in vitro maturation. This mRNA increase coincides with significant demethylation of a CpG island in the GnRH 5′-promoter region. This suggests epigenetic differentiation occurs during GnRH neuronal maturation. Third, estradiol causes rapid, direct, excitatory action in GnRH neurons and this estradiol action appears to be mediated through a membrane receptor, such as GPR30.

Introduction

The study of reproductive neuroendocrinology was heralded by a series of experiments by Geoffrey Harris, John Everett, and Charles Sawyer in the late 1930’s to early 1940’s (see ¹). These pioneers proposed the concept that the hypothalamus controls reproductive function and a neurochemical signal from the hypothalamus is released into the anterior pituitary gland to stimulate gonadotropin secretion (2). After intensive studies over the decades by several investigators, Andrew Schally finally isolated the neurochemical substance from the pig hypothalamus and identified its molecular structure (3). The discovery of gonadotropin-releasing hormone (GnRH, aka LHRH) led to extensive research in this field. In the early 1970’s Ernst Knobil began investigating the hypothalamic control of gonadotropin secretion in nonhuman primates and reported that many reproductive neuroendocrine characteristics were not identical to those described in rodents (see ⁴). He also proposed the concept of the “hypothalamic GnRH pulse-generator” (4), as GnRH is released in a pulsatile manner in mammalian species. The study of the mechanism of GnRH pulse-generation became quite important not only for understanding control of reproductive function, but also for clinical usage of GnRH agonists and antagonists as a treatment tool in human patients. Nonetheless, three decades later, several unresolved questions remain. What triggers the onset of puberty? How do GnRH neurons attain the ability to release the decapeptide in a pulsatile manner? Does estradiol (E₂) directly modify the GnRH pulse-generating mechanism? In this review the authors discuss possible answers to these questions, focusing on our research conducted in nonhuman primates.

1. Possible mechanisms of puberty onset in primates

a). Pubertal changes in GnRH release

Pulsatile administration of GnRH induces puberty in sexually immature female rhesus monkeys (5) suggesting that the pubertal increase in the pulsatile release of GnRH triggers puberty. In fact, direct measurements of GnRH from the stalk- median eminence (S-ME) in female rhesus monkeys indicate that not only does the total output of GnRH increase, but also several parameters of GnRH release change at the time of puberty. For instance, pulse frequency, pulse amplitude, baseline levels, and mean GnRH levels start to increase shortly before the first signs of puberty, whereas pulse frequency stabilizes, pulse amplitude, baseline, and mean GnRH levels continue to increase through menarche until first ovulation (6).

In primates, GnRH neurons appear active during the neonatal period and functionally mature well before the onset of puberty, but their activity is suppressed throughout the juvenile period (7–10). For instance, GnRH release, assessed by circulating LH levels in male rhesus monkeys, is pulsatile shortly after birth and elevated GnRH activity is observed throughout the first 12 weeks after birth (11). However, the GnRH neurosecretory system of juvenile primates subsequently enters a quiescent state (10), which is characterized by low release levels of GnRH (6). Similar developmental changes in LH have also been shown in patients with agonadal syndrome (13). Moreover, the low levels of GnRH release during the juvenile period are not attributable to the negative feedback inhibition from gonadal steroids, because removal of endogenous steroids does not eliminate the prepubertal hiatus in GnRH and LH secretion in gonadectomized rhesus monkeys (12,14). Therefore, the low level of GnRH release prior to puberty in primates is neither due to an inability of GnRH neurons to secrete the GnRH peptide nor suppression by gonadal steroids. Rather, it appears that the prepubertal GnRH neurosecretory system is under the control of a central inhibition in the brain (9,10).

b). Reduction in tonic inhibition: Role of γ-amino butyric acid (GABA)

Research in this lab has focused on elucidating the central mechanism(s) responsible for restraining GnRH release prior to puberty and has revealed that GABA, a major inhibitory neurotransmitter in the hypothalamus, appears to be largely responsible. Although there are no reports showing that mutation of GAD67 or the GABA_A receptor subunits in mice cause an abnormal timing of puberty (some of these mutants do not survive until the pubertal age, 9,15), and although there are conflicting reports whether GABA is stimulatory or inhibitory in control of GnRH release (see ^9,10), several studies in our laboratory have revealed compelling evidence that tonic GABA input is a crucial component of the central inhibition that restrains GnRH release in female monkeys prior to puberty. First, direct measurement of GABA and GnRH in the S-ME of the hypothalamus, where GnRH neuroterminals are located, indicates that GnRH and GABA levels are generally inversely related during the pubertal transition: in prepubertal monkeys, when GnRH release is low, GABA levels are greater than in pubertal monkeys, and GABA levels are lower when GnRH release increases after the onset of puberty (16). Second, whereas infusion of the GABA_A receptor antagonist, bicuculline, into the S–ME stimulates GnRH release to a much greater extent in prepubertal than in pubertal monkeys, infusion of GABA is effective in suppressing GnRH release in only pubertal monkeys, presumably because of the reduction in tonic GABA inhibition at the onset of puberty (16). The most compelling evidence that GABA is responsible for the central inhibition of GnRH release prior to puberty is that reducing GABAergic neurotransmission leads to precocious puberty. In fact, blocking GABA_A receptor activation by chronic infusion of bicuculline substantially advances the timing of menarche and first ovulation in female monkeys: the ages of menarche and first ovulation in control females were approximately 28 and 45 months, respectively, whereas the ages of menarche and first ovulation in bicuculline-treated animals were 18 and 31 months, respectively (17).

The role of glutamatergic neurons through NMDA receptors in puberty has been extensively reported (see ^7,8,18). NMDA stimulates GnRH release and pulsatile infusion of NMDA results in precocious puberty in rodents and primates (19,20). However, it is important to note that reduction in GABAergic inhibition of GnRH release, through the GABA_A receptor mediated mechanism, appears to be a prerequisite for a subsequent role of excitatory neurotransmitters, such as glutamate, in further stimulating GnRH release at the onset of puberty (21). The role of other excitatory and inhibitory neurotransmitters/neuromodulators, such as catecholamines, neuropeptide Y and opioids in puberty will not be discussed in this review, as it has been reviewed elsewhere (9,10,22). However, presently, it is unknown whether pruning of GABAergic inhibitory synapses on primate GnRH neurons occurs before puberty, or GABA_A subunit composition undergoes developmental changes.

c). Possible role of kisspeptin in puberty

Kisspeptin-54 and its receptor, GPR54, have been implicated as key components of the mechanism underlying the onset of puberty (see ^23–28). In humans, a single amino acid mutation of GPR54 results in the absence of puberty, delayed puberty or precocious puberty (29–31). Mice lacking the GPR54 gene fail to undergo puberty and have immature gonads (32). Similarly, mice lacking the KiSS-1 gene exhibit an absence of reproductive development (33,34). However, the results from GPR54 null mice and KiSS-1 null mice (35), indicate that there is a kisspeptin-GPR54 independent GnRH regulatory mechanism. This notion is also supported by reports that low-amplitude LH pulses are observed in human patients with severe forms of GPR54 mutations (29,36).

We investigated a potential role of kisspeptin-54 and GPR54 in the mechanism of puberty in female monkeys. The results from “archived” push-pull perfusate samples suggest that kisspeptin-54 release in the S-ME increases in association with the pubertal increase in GnRH release and that a nocturnal increase in kisspeptin-54 release is already observed in prepubertal monkeys and continues throughout the pubertal period (37). Our recent work with a microdialysis method further indicates that the pulse frequency, but not pulse amplitude, of kisspeptin-54 release increases at the time of puberty (38). Therefore, developmental changes in pulsatile kisspeptin-54 release appear to be parallel to those previously observed with the pubertal changes in pulsatile GnRH release. These observations are consistent with a pubertal increase in Kiss-1 mRNA in the MBH in both male and female monkeys (39).

Because primate GnRH neurons express GPR54 (39), it is possible that GnRH release is a consequence of the pubertal increase in kisspeptin-54 release. Similarly, it is possible that responsiveness of GPR54 to kisspeptin undergoes developmental changes. To test these possibilities we examined the effect of the kisspeptin agonist, human kisspeptin-10, in prepubertal and pubertal monkeys. Kisspeptin-10, infused through the microdialysis probe while dialysates are collected, stimulates GnRH release in both prepubertal and pubertal monkeys in a dose dependent manner. However, while GnRH peak levels in response to kisspeptin-10 are higher in pubertal animals, the percent stimulation of GnRH from the baseline level is not different between the age groups, because baseline GnRH levels in prepubertal monkeys are significantly lower than that in pubertal monkeys (40). Moreover, recent observations demonstrate that while the GnRH response to kisspeptin does not change in ovariectomized prepubertal monkeys, ovariectomy in midpubertal monkeys eliminates the GnRH response to kisspeptin, indicating that GPR54 becomes dependent on circulating estradiol after the initiation of puberty (41). The kisspeptin antagonist, peptide 234, was similarly examined in both age groups. Results show that peptide 234 significantly suppresses GnRH release in both prepubertal and pubertal monkeys with a larger suppression in pubertal monkeys compared to prepubertal monkeys (40,42). However, again, when considering the difference in baseline GnRH levels between the two developmental stages, there is no difference in the degree of peptide 234 suppression of GnRH release. Collectively, these observations suggest that the GnRH response to kisspeptins does not appear to undergo a developmental change at the onset of puberty, and that the pubertal increase in kisspeptin-54 release appears to be the factor facilitating the pubertal increase in GnRH release. Interestingly, after the onset of puberty GPR54 activity appears to become dependent on the presence of circulating estradiol. Our findings in monkeys at the time of puberty are quite different from those reported by Clarkson et al. (43), as mouse GnRH neurons do not respond to kisspeptin-10 until after the onset of puberty. Gonadal steroid independent GPR54 activation in primate GnRH neurons is consistent with a report by another group in male rhesus monkeys (39). However, the property of GPR54 activation in primate GnRH neurons becomes dependent on estrogen signaling after the onset of puberty (41) is similar to that described in mouse GnRH neurons, in which pubertal estradiol increases GPR54 activation (43).

d). Interaction between the kisspeptin neuronal system and GABA neuronal system

If the pubertal increase in kisspeptin-54 plays a role in puberty, an important question arises as to whether the pubertal reduction in GABA in the S-ME stimulates kisspeptin-54 or the pubertal increase in kisspeptin-54 reduces GABA inhibition. Recently, we have begun to investigate the role of GABA input in the pubertal increase in kisspeptin-54 release. Preliminary results indicate that the GABA_A receptor antagonist bicuculline dramatically stimulates kisspeptin-54 release in the S-ME of prepubertal monkeys (Kurian and Terasawa, unpublished observation). In fact, 0.1 μM bicuculline infusion through a microdialysis probe into the S-ME stimulates a 1,000-fold increase in kisspeptin-54 release over baseline levels. Notably, this fold increase of kisspeptin-54 release by bicuculline infusion in prepubertal monkeys is significantly greater than the previously observed bicuculline effects on GnRH release, which is ~10 fold over baseline levels (16). These preliminary findings are quite exciting, as it appears that the reduction in tonic inhibition by GABA leads to increased release of kisspeptin-54, which may, in turn, stimulate the pubertal increase in GnRH release (Figure 1A). Nonetheless, it is possible that an initial small increase in kisspeptin-54 triggers the reduction in inhibitory GABA tone. If this is the case, kisspeptin-54 may mediate a metabolic signal to GnRH neurons directly or indirectly. The importance of metabolic signals on puberty has been well documented (see ⁴⁴) and the role of kisspeptin neurons as metabolic signal mediators in rats has also been documented (23). We will examine this possibility in future experiments.

Figure 1.

Figure 1A: A possible neurocircuit controlling GnRH release at the onset of puberty in nonhuman primates. This schematic illustration was created primarily based on the findings from the authors’ laboratory. Inhibitory and excitatory signals to GnRH neurons (yellow) are shown by blue cell bodies with broken lines and red cell bodies with solid lines, respectively. Predominant tonic inhibition of GABA to GnRH neurons and/or kisspeptin neurons keeps GnRH release very low during the juvenile period in primates (see ⁷, ⁹, ¹⁷). At the time of puberty, a reduction in tonic GABA inhibition (mechanism yet to be clarified) allows an increase in glutamatergic signal to kisspeptin neurons (see ²¹), resulting in the pubertal increase in GnRH release (see text). The role of glutamatergic input through NMDA receptors on puberty (see ⁷) and stimulation of GnRH release by kisspeptin (see ²³–²⁸) have been extensively demonstrated. The reduction in tonic GABA inhibition and increase in glutamatergic signal may, in part, occur directly at GnRH neurons. Also, inhibitory and excitatory input to GnRH neurons could occur at the soma/dendrite as well as the neuroterminals. In the schema the hypothetical role of kisspeptin neurons mediating metabolic signals to GABA, glutamate, and GnRH neurons in primates is depicted by arrows. Other neurotransmitters and neuromodulators, such as catecholamines, neuropeptide Y, and opioids, are likely involved in the mechanism of puberty, but we do not discuss these in the article.

Figure 1B: Schematic illustration of possible mechanisms for rapid estrogen action in GnRH neurons. The role of GPR30 (green) and membrane receptors (mER, yet to be identified) is shown. Estradiol (E₂) induces an increase in [Ca²⁺]_i oscillations and synchronization of [Ca²⁺]_i oscillations and stimulates GnRH release (see ⁷⁵). We have also reported the involvement of GPR30 in rapid E₂ action. Bovine serum albumin (BSA) conjugated E₂ (E₂-BSA), the nuclear membrane impermeable estrogen dendrimer conjugate (EDC), and the GPR30 agonist G1, all induce effects similar to E₂. There is controversy over the localization of GPR30 (see ⁷⁵). E₂ action through GPR30 in the plasma membrane and possible interaction with epidermal growth factor receptors (EGFR) and adenylyl cyclase (AC) are adapted from studies in cancer cells (89). The identity of mERs is still unclear. Classical estrogen receptors, such as ERβ, have been reported in mouse GnRH neurons as well as in GT1 cells, while ERX, and STX sensitive receptors (see ⁶⁹), and ERα splice variant ERα–36 in GnRH neurons have not been shown (75). Some of the mERs may interact with intracellular GPR30 in the Golgi apparatus and/or endoplasmic reticulum (see text).

2. Development of primate GnRH neurons coincides with epigenetic changes

a). GnRH neurons require a maturational period prior to becoming functional

Pulsatile peptide release appears to be an intrinsic property of GnRH neurons. Mouse GnRH neurons exhibit periodic burst firing activity (45), and they have dendro-dendritic bundlings and shared synapses (46). These morphological features are necessary for intercellular communication among GnRH neurons, as have been shown in oxytocin neurons (47). GnRH neuronal cultures derived from the monkey nasal placode, which rarely contain nonGnRH neurons (48), and GT1 GnRH neurosecreting cells (49–52) release the decapeptide in a pulsatile manner. Moreover, cultured GnRH neurons also exhibit spontaneous intracellular calcium ([Ca²⁺]_i) oscillations, which periodically synchronize at a frequency similar to pulsatile GnRH release in vivo (53,54).

GnRH neurons are not functionally mature when they originate from the nasal placode i.e., embryonic day (E) 35–36, the time when we start cultures. We have found that there are no spontaneous [Ca²⁺]_i oscillations in primate GnRH neurons during the first 8–9 days in vitro (div), infrequent [Ca²⁺]_i oscillations appear in a small percent of GnRH neurons at 12–13 div, and subsequently, at 14 div [Ca²⁺]_i oscillations appear in most GnRH neurons (53; unpublished observation). Likewise, perifusion experiments with placode cultures indicate that GnRH is rarely detectable during the first 14–17 div (48). Recently, Susan Wray and her colleagues report similar observations in mouse GnRH neurons derived from the nasal placode: GnRH pulses are detectable at 3 div (cultures started at E11.5), and the amplitude of GnRH pulses and the secretory rate gradually increase with culture days, reaching the maximum at 14 div (55). This group also has found that synchronization of [Ca²⁺]_i oscillations among GnRH neurons are sparse when cells are young in culture, but becomes more frequent when cells mature by 14 div. This parallels an increase in tissue contents of GnRH between 14 and 21 div (56). Although there are subtle differences in monkey and mouse studies, the results are consistent with a concept that GnRH neurons undergo functional maturation during the first 2 weeks after differentiation from their progenitor cells.

In vivo maturational events corroborate in vitro data. GnRH neurons migrate into the forebrain within a few days after their birth in the nasal placode, and most, but not all, reach the preoptic area and medial basal hypothalamus within 2 weeks (57,58). GnRH neurons are functional at E50-E70, as active gonadotropes are found in the pituitary at E50 (57), and sex-specific gonadal steroids are detectable in the umbilical cord at E70 (59).

It has been consistently shown that in vitro maturation recapitulates in vivo maturation in neurons, suggesting a 2-week period for functional maturation is not an artifact. For example, protein expression of metabolic glutamate receptors in cultured rat hypothalamic neurons started at birth, increase with culture days in parallel to the postnatal ages (60) and GABA neurotransmission is reversed from excitatory to inhibitory between 4 and 18 div in cultured cells, similar to those observed in slice preparations obtained from the hypothalamus of P4 or P18 rats (61). Likewise, GnRH peptide contents per cell gradually increased with a similar developmental time course both in vivo and in vitro, i.e., GnRH contents in mouse nasal placode cultures (started at E11.5) significantly increased between 7 and 10 div, which parallels the in vivo increase between E14.5 and P1. Moreover, GnRH peptide contents at 10 div are not distinguishable from peptide contents at P1, a comparable time point in vivo (56).

A question arises as to what influences GnRH gene expression and neurosecretory capacity of GnRH neurons during embryonic development. Gonadal steroids profoundly alter mRNA levels and GnRH neurosecretion (62,63). However, gonadal steroids are not likely involved in the process of neuronal maturation at this embryonic stage. At the time GnRH neurons are generated at ~E34, there are no gonadal steroids of fetal origin, as gonads are not differentiated yet. Although gonadal steroids of maternal origin cannot be excluded in vivo, in vitro studies by Wray’s group use serum free media and in our studies steroid levels in media are negligible. The position of GnRH neurons in the nasal area or in the brain is also not likely involved in this process, as GnRH neurons are already in the brain at E38 in rhesus monkeys and E14.5 in mice, yet an increase in GnRH mRNA requires more than 14 days in vitro in both species. Collectively, temporally-regulated, rather than spatially regulated, developmental factors appear to be critical for maturation of GnRH neurons, as suggested by Wray and colleagues (55).

b). Epigenetic changes during GnRH neuronal cell maturation

The epigenetic regulation of cellular phenotypes is organized by several modifications to chromatin, which include DNA (CpG) methylation and several covalent modifications to the histone core. The accumulation of these modifications alters rates of gene expression by organizing transcriptionally permissive or repressed chromatin states. DNA methylation itself is thought to alter gene transcription rates by two general mechanisms. First, CpG methylation can interfere with transcription factor assembly and alter the rate of transcription. Second, methylated CpGs may be recognized by methyl-CpG-binding proteins, which recruit co-repressor complexes that modify the histone core of chromatin. Therefore, investigation of DNA methylation status is important for initial evaluations of potential epigenetic gene regulation, as CpG methylation is 1) a direct mediator of altered gene expression and 2) related to the presence of major histone modifications that impact gene expression.

We recently began investigating the CpG methylation status of a distal 5′ CpG rich region of the rhesus monkey GnRH gene. We focused on the 5′ portion of the gene as previous investigations had already identified this region as pertinent to GnRH neuronal migration into the brain (64), gene expression (65), and tissue specific expression (66). Kepa et al. (65) also identified sequence similarities between rodent and human genes in the distal 5′ region. Remarkably, during the development of embryonic nasal placode derived GnRH neurons, we found a tight temporal correlation between decreasing CpG methylation status and increasing GnRH mRNA levels (67). Thus, it appears that higher levels of GnRH mRNA expression are associated with lower levels of CpG methylation in this 5′ region of the rhesus monkey gene. In fact, this relationship is consistent across several comparisons. Specifically, we have found that GnRH mRNA levels are 1) higher in the adult female than adult male in the MBH (68), 2) higher in the young adult male than the prepubertal male MBH (67), and 3) as stated above, increase with culture age over the first three weeks in GnRH neurons derived from nasal placode (67). In each of these cases, higher mRNA levels are associated with a lower CpG methylation status. These findings are the first to suggest that CpG methylation in this region is associated with GnRH mRNA expression in the primate hypothalamus. Further, they suggest that CpG methylation status of the primate GnRH gene changes during embryonic development as well as at the time of puberty.

3. Rapid action of estradiol on GnRH neurons

Classically, 17β-estradiol (E₂) modifies neuronal function through the actions of a soluble estrogen receptor (ER). E₂ binds either ERα or ERβ, which promotes receptor dimerization, nuclear translocation, co-factor recruitment, and genomic activity. This occurs within a time scale of hours to days. Findings in rodent neurons, however, suggest that E₂ causes rapid actions, occurring within a time scale of seconds to minutes (69). Currently, the mechanisms of the E₂ action on GnRH neurons remain unclear.

Effects of E₂ on GnRH neurons, such as negative and positive feedback, are generally thought to occur through interneurons that appose GnRH neurons (70,71). However, direct actions of E₂ on mouse GnRH neurons through ERβ have also been reported (72–74). We also found that E₂ causes a direct stimulatory action on primate GnRH neurons. Because we recently wrote a review article on this particular topic (75), our review will be brief in this article.

a). Estradiol rapidly and directly stimulates GnRH neurons

As previously described (53,76), our olfactory placode cultures derived from monkey fetuses between E35–37 contain GnRH neurons, GnRH progenitor cells, epithelial cells, fibroblasts, and other unidentified non-neuronal cells, but essentially no non-GnRH neurons. Nonetheless, we always confirm that recorded cells were GnRH neurons after each experiment. Using placode cultures, we examined the effect of E₂ on GnRH neurons with patch clamp recording and calcium imaging methods. We found E₂ rapidly stimulates GnRH neural firing activity (77), the frequency and synchronizations of [Ca²⁺]_i oscillations (78), and GnRH release (79). Remarkably, the neural firing and [Ca²⁺]_i oscillations occurred within one minute (77,78) and GnRH release was stimulated in less than 10 minutes (79). These time scales clearly suggested non-classical/non-genomic effects of E₂ on GnRH neurons.

The excitatory effect of E₂ on primate GnRH neurons appears to occur at the membrane level and does not require entry of E₂ into cells, i.e., an E₂-BSA conjugate stimulates neural firing activity (77) and a nuclear membrane impermeable estrogen dendrimer conjugate, described by Harrington et al. (80), results in a robust increase in the frequencies of [Ca²⁺]_i oscillations and synchronizations (78). Moreover, the stimulatory effect of E₂ on [Ca²⁺]_i oscillations is not blocked by TTX, suggesting that E₂ causes a direct action on GnRH neurons, rather than through synaptic input from other neurons.

b). Role of G-Protein Coupled Receptor 30 (GPR30) in mediating estradiol action in GnRH neurons

Recent studies by several research groups suggest that E₂ causes rapid actions in neurons through ERα or ERβ (72,73,81–83) or a novel membrane receptor (84–86).

In order to elucidate the type of receptors mediating E₂ action in primate GnRH neurons we first examined whether the estrogen receptor antagonist, ICI 182,780, blocks the E₂-induced increases in [Ca²⁺]_i oscillations, and GnRH release. The E₂-induced increases in [Ca²⁺]_i oscillations and GnRH release are not blocked by ICI 182,780, suggesting that the classical estrogen receptors, ERα and ERβ, are not likely involved in the rapid action of E₂ in primate GnRH neurons. To further investigate the mechanism of E₂ action in primate GnRH neurons, we examined whether E₂-induced [Ca²⁺]_i oscillations and GnRH release can be seen in GnRH neurons treated with pertussis toxin (PTX). PTX is a broad spectral G-protein coupled receptor (GPCR) blocker. Results suggest that E₂ application to cells treated with PTX fails to induce both [Ca²⁺]_i oscillations and GnRH release (79), indicating that E₂ action in primate GnRH neurons is mediated by a GPCR.

Next we examined the role of GPR30 in E₂ actions on GnRH neurons, as it has been suggested that E₂ action can occur through GPR30 in cancer cells (87). We found GPR30 mRNA is expressed in the olfactory placode, from which GnRH neurons originate. The deduced amino acid sequence of GPR30 through sequencing this mRNA in the monkey olfactory placode is identical to the GPR30 encoded by the human genome (GenBank accession numbers AF027956). GPR30 protein is also found in a subset of GnRH neurons in the adult monkey hypothalamus (79).

If GPR30 mediates rapid E₂ actions in primate GnRH neurons, knockdown of GPR30 by transfection with small interference (si)RNA specific to human GPR30 would eliminate the E₂-induced increase in [Ca²⁺]_i oscillations. The results indicate that transfection of GnRH neurons with siRNA for GPR30 completely abrogates the E₂-induced changes in [Ca²⁺]_i oscillations, whereas transfection with control siRNA does not. Similarly, the EDC-induced increase in [Ca²⁺]_i oscillations also does not occur in GnRH neurons transfected with GPR30 siRNA. Subsequently, we examined whether G1, a GPR30 agonist (88), can induce an effect similar to E₂. Results suggest that G1 stimulates the frequency of [Ca²⁺]_i oscillations, similar to those observed with E₂, although it does not cause an increase in synchronization of [Ca²⁺]_i oscillations (79). Preliminary data further indicate that G1 also stimulates GnRH release (Shiel and Terasawa, unpublished observation). Therefore, GPR30, at least in part, mediates the rapid action of E₂ in primate GnRH neurons (Figure 1B).

Subcellular localization of GPR30 is still controversial. Filardo and collaborators (89) have shown the presence of GPR30 in the plasma membrane, whereas Prossnitz and collaborators have shown the presence of GPR30 in the endoplasmic reticulum, Golgi apparatus, and nuclear membrane (90). More recently Filardo and colleagues (91) have suggested that GPR30 is located at the cell surface, but upon E₂ stimulation GPR30 is partly sequestered to clathrin-coated vesicles, which makes it difficult to demonstrate GPR30 presence in the plasma membrane. Internalization of seven-transmembrane GPCRs commonly occurs as a consequence of receptor endocytosis following ligand binding (92). While subcellular localization of GPR30 remains unsettled, another hypothesis regarding GPR30 function in cancer cells was recently proposed (93). This group finds GPR30 induces expression of a 36-kDa variant of ERα (ERα-66), ERα-36, which associates with the plasma membrane (94). Consequently, ERα-36 and GPR30 together mediate rapid non-genomic effects of E₂ in breast cancer cells (93). The generalizability of this pathway needs to be tested. Nonetheless, if it is proven that GPR30 interacts with an E₂ binding partner such as ERα-36, it is possible that the subcellular localization of GPR30 may not be of significant importance to E₂ actions through GPR30. At this time, neither the subcellular localization of GPR30 nor presence of ER α-36 has been determined in primate GnRH neurons.

Concluding Remarks

In this article we have summarized recent progress in understanding the physiology of GnRH neurons in nonhuman primates. During the juvenile period in primates, the activity of GnRH neurons is suppressed by a central inhibition until the time of puberty. A series of our studies suggests that a reduction in tonic GABAergic inhibition over GnRH release through a GABA_A receptor-mediated mechanism appears to be crucial for subsequent stimulation of kisspeptin-54 release, which likely results in the pubertal increase in GnRH release. Presently, it is unknown whether GABAergic inhibitory synaptic pruning occurs prior to puberty or GABA_A subunit compositions on GnRH neurons undergo developmental changes. Developmental modification of GABA_A subunit composition in kisspeptin and GnRH neurons is an important issue, as GABA_A receptors are modulated by various allosteric agonists, such as the neurosteroid allopregnanolone. Nonetheless, it appears that intricate interactions between kisspeptin, GABA, and glutamate neurons play a key role in the pubertal GnRH increase in nonhuman primates (Figure 1A). The question of how metabolic signals impinge on these neuronal systems and what role they play in the mechanism of puberty remains to be investigated.

Using our cell culture system we have found changes in GnRH mRNA expression and maturation of GnRH neurons (i.e., ability to release the decapeptide in a pulsatile manner) are associated with epigenetic changes. We also found E₂ causes rapid direct actions on GnRH neurons through the membrane receptor, GPR30 (Figure 1B). However, these findings do not answer all necessary questions. A full understanding of the mechanism of GnRH pulse-generation and estrogen action in primates is far from clear. Nonetheless, significant progress is being made toward a full understanding of the mechanisms controlling the nonhuman primate reproductive neuroendocrine system.