Direct Action of Melatonin in Human Granulosa-Luteal Cells

The direct involvement of melatonin in modulation of ovarian steroidogenesis, the high levels of melatonin found in human follicular fluid, and the presence of melatonin binding sites in the ovary led us to hypothesize that melatonin acts as a modulator of ovarian function. In contrast to the hypothalamus and pituitary, the mechanism of melatonin action at the level of the ovary is still poorly understood. In the present study, we investigated the gene expression of the two different forms of melatonin receptors in human granulosa-luteal cells, using RT-PCR. PCR products corresponding to the expected sizes of the melatonin receptor subtypes, mt₁-R and MT₂-R, were obtained from granulosa-luteal cells, and the authenticity of the PCR products was confirmed by Southern blot hybridization with cDNA probes. Subsequent cloning and sequence analysis revealed that the ovarian mt₁-R and MT₂-R cDNAs are identical to their brain counterparts. Because gonadotropins and GnRH acting through specific receptors in the human ovary regulate cellular functions, we investigated the role of melatonin in the regulation of FSH receptor, LH receptor, GnRH, and GnRH receptor levels. Treatment with melatonin (10 pm–100 nm) significantly increased LH receptor mRNA levels without altering the expression of the FSH receptor gene. Both GnRH and GnRH receptor mRNA levels were significantly decreased, to 61% and 45% of control levels, respectively, after melatonin treatment. Melatonin treatment alone had no effect on basal progesterone production but enhanced the effects of human CG-stimulated progesterone production. Because MAPKs are activated in response to a diverse array of extracellular stimuli leading to the regulation of cell growth, division, and differentiation, and because melatonin has been shown to modulate cellular proliferation and differentiation, in this study, we demonstrated that melatonin activated MAPK in a dose- and time-dependent manner. In summary, our studies demonstrate, for the first time, that melatonin can regulate progesterone production, LH receptor, GnRH, and GnRH receptor gene expression through melatonin receptors in human granulosa-luteal cells, which may be mediated via the MAPK pathway and activation of Elk-1. Our results support the notion that melatonin plays a direct role in regulating ovarian function.

MELATONIN, A PINEAL hormone, regulates the dynamic physiological adaptations that occur in seasonally breeding mammals in response to changes in day length, but its role in reproduction in humans remains unclear (1, 2). It is generally believed that the reproductive actions of melatonin are mediated by way of regulating gonadotropin release after effects on hypothalamic monoamine and GnRH (1, 3, 4) and possibly on cAMP and Ca²⁺-dependent intracellular mechanisms in the hypophysis (5, 6). However, there is evidence to suggest that melatonin acts at the level of the ovary to modify ovarian function. High levels of melatonin, which undergo seasonal variation (7), are found in human preovulatory follicular fluid (8, 9). In the chicken, duck, and quail, autoradiographic studies show that 2[¹²⁵I]iodomelatonin binding is clearly concentrated around the follicles in which granulosa cells are known to proliferate (10). In the human, melatonin binding sites have been detected in granulosa-luteal cells (11), whereas melatonin can have a direct effect on ovarian steroidogenesis (12, 13).

Through a receptor-mediated process, pituitary gonadotropins, LH and FSH, play a central role in regulating ovarian functions. It is well established that melatonin can influence gonadotroph secretion of LH and FSH (14). These gonadotropins bind to receptors in the ovary, leading to effects on steroidogenesis and gametogenesis (15). The effects of gonadotropins on follicular growth, ovulation, and luteinization are associated with differences in FSH receptor (FSHR) and LHR concentrations (16, 17). Recently, GnRH and its receptor have been detected in the human ovary (18). Several reports suggest that it is an important paracrine/autocrine regulator in the ovary (16, 19, 20) and can modulate steroidogenesis (19, 21). Given that melatonin can functionally alter steroidogenesis, it is possible that melatonin may be interacting with gonadotropins and GnRH to modulate the amplitude of the transduction signal. The ability of these hormones to modulate ovarian function depends not only on the circulating levels of these hormones but also on the appropriate expression of their receptors. Previous work from our (22) and other (23) laboratories has shown that multiple forms of the melatonin receptor genes (24, 25), mt₁-R and MT₂-R, are expressed in human granulosa-luteal cells (hGLCs). These findings provide further evidence that melatonin may regulate ovarian function through binding to melatonin receptors, a G protein-coupled receptor. However, the role of melatonin in hGLCs has not been reported.

Recent interest concerning the role of melatonin in regulating cellular proliferation and differentiation in a number of different tissues raises the question of how does melatonin act to regulate these cellular processes (26, 27). Little is known about the molecular events that mediate the actions of melatonin in these tissues. Because a number of effectors are regulated by the heterotrimeric G protein subunits, including serine/threonine kinases[ a family of kinases (MAPKs) involved in the transduction of externally derived signals to regulate cell growth, division, and differentiation] (28), it is possible that the effects of melatonin on cellular proliferation and differentiation may be mediated by the MAPK signaling cascade.

In the present study, to examine the potential direct regulatory role of melatonin in the ovary, we first confirmed the expression of melatonin receptors in hGLCs. To further understand the physiological significance of melatonin in the human ovary, we investigated the effects of melatonin on the regulation of gonadotropin receptors, GnRH and its receptor gene regulation, and the possible involvement of MAPK in the signal transduction mechanism of melatonin.

Methods and Materials

hGLC culture and treatment

The use of hGLCs was approved by the Clinical Screening Committee for Research and Other Studies Involving Human Subjects of the University of British Columbia. Follicular aspirates were collected during oocyte retrieval from women undergoing in vitro fertilization and ovarian stimulation, as described previously (29). Briefly, granulosa cells were separated from red blood cells in follicular aspirates by centrifugation through Ficoll Paque (Amersham Pharmacia Biotech, Bale D’Urfé, Québec, Canada) (30). Cells in the interface were collected, washed twice, and resuspended in DMEM (Life Technologies, Inc., Burlington, Ontario, Canada) supplemented with 10% FBS (Life Technologies, Inc.), 100 U/ml penicillin G, and 100 μg/ml streptomycin (Life Technologies, Inc.) and seeded at a density of approximately 1 × 10⁵ cells/ml in 35-mm culture dishes. The cells were allowed to adhere for 48 h at 37 C in a humidified atmosphere of 5% CO₂-95% air. On d 4 in culture, hGLCs were incubated in serum-free medium for 24 h, as performed previously (30). For the regulation of gene expression studies, cells were treated with various concentrations of melatonin (10 pm–100 nm) for 24 h on d 5 in culture. To examine the effects of melatonin on progesterone production in our culture system, hGLCs were treated with increasing concentrations of melatonin in the presence or absence of human CG (hCG) (1 IU). To study the regulation of MAPK activation by melatonin, cells were treated with melatonin (10 pm–10 μm) for varying intervals (1–20 min). For the in vitro MAPK assays, cells were serum-starved for 4 h and treated with 1 nm melatonin for 5 min. Control cultures were treated with vehicle (0.01% wt/vol ethanol).

Total RNA isolation and first-strand cDNA synthesis

Total RNA was prepared from the cultured hGLCs following the acid phenol-chloroform protocol of Chomczynski and Sacchi (31). In brief, hGLCs were disrupted in lysis buffer [4 m guanidine thiocyanate, 25 mm sodium citrate (pH 7.0), 0.5% N-lauroyl sarcosine, and 0.1 mβ -mercaptoethanol] followed by acid-phenol extraction. The RNA concentration was determined based on absorbance at 260 nm, and the mRNA integrity was confirmed by agarose-formaldehyde gel electrophoresis. One microgram of total RNA obtained from hGLCs was reverse-transcribed into cDNA using the First Strand cDNA Synthesis Kit (Pharmacia Biotech, Morgan, Canada).

PCR, cloning, and sequencing of the mt₁-R and MT₂-R cDNA

Based on the published brain and retinal sequence for the human melatonin receptors, mt₁-R and MT₂-R (24, 25), two pairs of primers were designed to amplify the mt₁-R and MT₂-R from hGLCs (Fig. 1, A and B). PCR reactions were performed in the presence of 2.5 U Taq polymerase (Life Technologies, Inc.) and its buffer, 1.5 mm MgCl₂, 10 mm deoxynucleotide triphosphate (dNTP), and 50 pmol of each primer. PCR amplification was carried out for 33 cycles with denaturing at 94 C for 60 sec, annealing at 58 C for 35 sec, and extension at 72 C for 90 sec, followed by a final extension at 72 C for 15 min. Amplified PCR products were subjected to Southern blot analysis (Fig. 1, C and D). The PCR products were fractionated by agarose gel electrophoresis and visualized with ethidium bromide staining and UV light. The PCR products were transferred to a nylon membrane and hybridized with digoxigenin-labeled cDNA probe (Roche Molecular Biochemicals, Laval, Canada) for human mt₁-R and MT₂-R. The cDNAs for human mt₁-R and MT₂-R were kindly provided by Dr. S. M. Reppert (Harvard Medical School, Boston, MA). After high-stringency washes, detection was carried out following the manufacturer’s recommended procedures (Roche Molecular Biochemicals), and the membrane was exposed to Omat x-ray film (Eastman Kodak Co., Rochester, NY). The PCR products of the mt₁-R and MT₂-R cDNAs were cloned into PCRII Vector (Invitrogen, San Diego, CA) and sequenced by the dideoxynucleotide chain termination method using the T7 DNA Polymerase Sequencing Kit (Pharmacia Biotech).

PCR and quantification of FSHR, LHR, and GnRH mRNA from hGLCs

Using a semiquantitative PCR system, the expression and regulation of FSHR, LHR, and GnRH mRNA levels were compared. PCR amplifications were carried out under increasing cycle numbers in 50-μl reactions containing 2.5 U Taq polymerase and its buffer, 1.5 mm MgCl₂, 2 mm dNTP, and 50 pmol of sense and antisense primer. The number of PCR cycles used for subsequent experiments fell within the linear range (Fig. 2). PCRs for FSHR and LHR were carried out for 26 cycles with denaturing for 1 min at 94 C, annealing for 35 sec at 55 C, extension for 90 sec at 72 C, and a final extension for 10 min at 72 C. The primers for LHR were: sense, 5′-GCCCACCTTGGACCCTCAGAG-3′; and antisense, 5′-CCAATAAAGTGTGAGGTTCTCCG-3′. Primers for FSHR were described previously (31A ). The cDNAs for human FSHR and LHR were kindly provided by Dr. T. Minegishi (Gunma University, Maebashi, Japan) and used as a template for DIG-labeled probes. PCR amplification for GnRH was carried out for 25 cycles with denaturing for 1 min at 94 C, annealing for 35 sec at 53 C, extension for 90 sec at 72 C, followed by a final extension for 10 min at 72 C. Primers for GnRH were designed based on the published sequence of human hypothalamic GnRH (32), as described previously (33). To standardize for first-strand cDNA synthesis efficiency, PCR forβ -actin was performed. Primers for β-actin were designed from the human published sequence (34), as described previously (35). Amplified PCR products were separated by agarose gel electrophoresis and subjected to Southern blot analysis. Quantitation was performed using Scion Image-Released β 3b (Scion Corp., Bethesda, MD).

RT-PCR and quantification of GnRH receptor (GnRHR) mRNA from hGLCs

For competitive PCR, 4 μl first-strand cDNA from 1 μg total RNA were coamplified with 0.08 pg mutant GnRHR cDNA, constructed in our laboratory and previously described (20). The PCR for GnRHR was carried out in a 50-μl PCR reaction containing 2.5 U Taq polymerase, its buffer, 1.5 mm MgCl₂, 2 mm dNTP, and 50 pmol specific primers (P44F and P45R), with denaturing for 1 min at 94 C, annealing for 35 sec at 60 C, extension for 90 sec at 72 C, and a final extension for 10 min at 72 C for 33 cycles. Quantitation was performed using Scion Image-Releasedβ 3b.

RIA for progesterone

The progesterone concentration in culture media was measured by an established RIA (36). Antiprogesterone antibody was kindly provided by Dr. D. T. Armstrong (University of Western Ontario, London, Canada). Briefly, samples were incubated with antibody and tracer, with a final concentration of 7000 cpm/ml[ 1,2,7,6,17-³H]progesterone (Amersham Pharmacia Biotech). After incubation for 16–24 h, a charcoal/dextran solution was added to remove unbound progesterone and tracer. Scintillation cocktail (Amersham Pharmacia Biotech) was added to each sample, and the vials were counted in a β-counter (LKB Wallac, Inc., Turku, Finland). The cells were harvested and lysed for quantification of protein amount using the protein assay kit from Bio-Rad Laboratories, Inc. (Richmond, CA). Samples were assayed in triplicates, and progesterone concentrations were normalized against protein content.

Western immunoblot analysis

Phospho-specific MAPK antibody (New England Biolabs, Inc., Beverly, MA), which detects p44 and p42 MAPK only when phosphorylated at Thr²⁰² and Tyr²⁰⁴, was used to measure MAPK activity. The hGLCs were washed with ice-cold PBS and lysed with 100 μL RIPA [150 mm NaCl, 50 mm Tris-HCl (pH 7.5), 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 1.0 mm PMSF, 10μ g/ml leupeptin, and 100 μg/ml aprotinin]. The cellular extracts were placed on ice for 15 min and centrifuged to remove cellular debris, and protein content of the supernatants was determined using the Bio-Rad Laboratories, Inc. protein assay kit. Thirty micrograms of total protein were subjected to 10% SDS-PAGE under reducing conditions as previously described (37) and then electrophoretically transferred from the gels onto nitrocellulose membranes (Amersham Pharmacia Biotech) according to Towbin et al. (38). The membranes were probed with a mouse monoclonal antibody directed against the phosphorylated forms of ERK1 and ERK2 (p42^MAPK and p44^MAPK) at 4 C for 16 h. After washing, the signals were detected with horseradish peroxidase-conjugated goat-antimouse secondary antibody and visualized using the ECL chemiluminescent system (Amersham Pharmacia Biotech), followed by autoradiography. Intensities of the signals were quantified with Scion Image-Released β 3b.

In vitro MAPK assay

After treatment, the cells were washed in ice-cold PBS and lysed in RIPA buffer as described above. Cellular protein (200 μg) was immunoprecipitated with 15 μl resuspended immobilized phospho-p44/42 MAPK (Thr²⁰²/Tyr²⁰⁴) monoclonal antibody for 16 h at 4 C with gentle rocking. Active MAPK (phosphorylated p42 MAPK; 20 ng) was also immunoprecipitated as a positive control. The tubes were centrifuged, and the pellets were washed once with 500 μl of 1× Lysis Buffer and twice with 500 μl of 1× Kinase Buffer [25 mm Tris (pH 7.5), 5 mm β-glycerolphosphate, 2 mm dithiothreitol, 0.1 mm Na₃V0₄, 10 mm MgCl₂]. In vitro MAPK assays were performed using the Elk-1 fusion protein as a substrate for the MAPKs, according to the manufacturer’s suggested procedure (New England Biolabs, Inc.). The pellet was resuspended in 50 μL of 1× Kinase Buffer supplemented with 200μ m ATP and 2 μg Elk-1 fusion protein and incubated for 30 min at 30 C. The reaction was terminated by adding 25μ l of 3× SDS loading buffer and boiled for 5 min. Forty microliters of reaction mixture were run on 10% SDS-PAGE gels and electrotransferred to a nitrocellulose membrane, as described earlier. The membrane was immunoblotted using a rabbit polyclonal antibody specific to the phosphorylated Elk-1 (Ser³⁸³). After washing, the signals were detected with horseradish peroxidase-conjugated secondary antibody, and visualized using the ECL chemiluminescent system, followed by autoradiography.

Statistical analysis

Relative LHR, FSHR, and GnRH mRNA levels were expressed as the ratio of LHR, FSHR, and GnRH to β-actin. The amount of GnRHR transcript was calculated from the ratio of the target to competitive cDNA. Expression levels of LHR, FSHR, GnRH, and GnRHR mRNA are expressed as the percent change from the control value. Data are shown as the means of three individual experiments from three different patients with duplicate samples and are presented as the mean ± se. Progesterone levels are expressed as the mean ± se of three individual experiments from three different patients with triplicate samples. MAPK levels are expressed as fold changes compared with basal levels and represented as the mean ± se of three individual experiments from three different patients. Elk-1 activity is expressed as fold change compared with control and represented as the mean ± se of three individual experiments from three different patients. Statistical analysis was performed by one-way ANOVA followed by Fisher’s PLSD test (Statview for Windows 4.53, Abacus Concepts, Inc., Berkeley, CA). A P-value less than 0.05 was considered statistically significant.

Results

Expression of mt₁-R and MT₂-R mRNA in hGLCs

To investigate the expression of the multiple forms of the melatonin receptor mRNA in hGLCs, two pairs of primers derived from human brain and retinal mt₁-R and MT₂-R cDNAs were designed. The positions and sequences of primers are shown in Fig. 1, A and B. Using primers P54F and P55R, PCR amplification yielded a 326-bp DNA fragment from the hGLCs and the positive control cell line, human embryonic kidney (HEK) cell line 293 (Fig. 1C). When the PCR products were transferred to a nylon membrane and hybridized with a specific probe for mt₁-R cDNA under high-stringency conditions, specific hybridization signals were obtained, thereby confirming the identity of the PCR product (Fig. 1C, lower panel). Using primers P64F and P65R, the predicted PCR products (321-bp DNA fragment) were obtained in hGLCs and the positive control cell line, HEK-293, and were validated as MT₂-R by hybridization with a specific probe for MT₂-R cDNA (Fig. 1D, lower panel). Because both sets of primers were located in different exons, the amplified products were not attributable to genomic DNA contamination but to specific amplification of mRNA. The possibility of cross-contamination was also ruled out, because no PCR products were detected in the negative control, without template, by ethidium bromide staining and Southern blot analysis for both the mt₁-R and MT₂-R (Fig. 1, C and D). Sequence analysis revealed that the mt₁-R and MT₂-R from hGLCs have sequences identical to those found in the brain and retina, respectively (data not shown).

Validation of PCR for FSHR, LHR, GnRH, and GnRHR transcript

Using semiquantitative RT-PCR with primers specific for the human FSHR, LHR, and GnRH cDNA, the levels of FSHR, LHR, and GnRH mRNA were examined in hGLCs. To determine where PCR amplification for FSHR, LHR, GnRH, and β-actin mRNA were, in the logarithmic phase, 1 μg total RNA was reverse transcribed and was amplified under different cycle numbers (Fig. 2). A linear relationship between PCR products and amplification cycles was observed for FSHR (Fig. 2A), LHR (Fig. 2B), GnRH (Fig. 2C), and β-actin (Fig. 2D). Consequently, 26 cycles for FSHR and LHR, 25 cycles for GnRH, and 18 cycles for β-actin were employed for quantification. GnRHR mRNA levels were analyzed using competitive RT-PCR, as validated previously (20).

Effect of melatonin on FSHR, LHR, GnRH, and GnRHR expression in hGLCs

Treatment with varying doses of melatonin (10 pm–100 nm) did not induce any significant response in FSHR mRNA levels (Fig. 3A). In contrast, an up-regulation of LHR mRNA levels was observed in cells treated with melatonin at all concentrations used. As shown in Fig. 3B, doses as low as 10 pm increased LHR mRNA levels by 29% (P < 0.05) relative to control. A maximum increase of 80% (P < 0.001) in LHR mRNA was observed with 1 nm melatonin treatment (Fig. 3B). At a higher melatonin treatment (100 nm), there was no further effect on LHR mRNA levels, compared with a 1-nm dose (P > 0.05). Treatment of hGLCs with various concentrations of melatonin induced a significant decrease in GnRH mRNA levels. As depicted in Fig. 3C, a low dose of melatonin (10 pm) was able to inhibit GnRH levels (12%; P < 0.05). A maximum effect was observed at a melatonin treatment of 1 nm, significantly decreasing GnRH, to 61% (P < 0.0001) of control. Further increase in melatonin concentration had no further effect on GnRH mRNA levels. Similarly, down-regulation of GnRH-R was observed in all melatonin concentrations tested (Fig. 3D). A 24% (P < 0.05) decrease and 34% (P < 0.001) decrease in GnRH-R mRNA levels were observed with a dose of 10 pm and 1 nm melatonin. With an even-higher dosage of melatonin (100 nm), there was further down-regulation of GnRH-R levels, to 45% of control levels (P < 0.001 vs. control, P < 0.05 vs. 1 nm).

Effect of melatonin on basal and hCG-induced progesterone production

hGLCs were treated with melatonin, alone or in combination with hCG. Melatonin had no significant effects on basal progesterone secretion from hGLCs. As expected, hCG significantly increased progesterone accumulation (Fig. 4). Human GLCs treated with 1 nm melatonin plus 1 IU hCG enhanced hCG-stimulated progesterone production (P < 0.05). However, increasing the amount of melatonin had little further effect.

Effect of melatonin on MAPK activation and Elk-1 phosphorylation

Melatonin stimulates MAPK activation in a dose-dependent manner in hGLCs treated with melatonin for 5 min. MAPK activity is stimulated at concentrations of 10 pm–10 μm (data not shown for 10 μm). Maximal activity (5.8-fold over basal levels) was observed at a concentration of 1 nm (Fig. 5). To assess the duration of MAPK activation, hGLCs were treated with melatonin (1 nm) in a time-dependent manner. The results showed that melatonin stimulated a rapid MAPK activation, observed within 1 min, with maximum stimulation (6.1-fold over basal levels) within 5 min (Fig. 6). MAPK levels fell by one-third within 10 min (4.1-fold over basal levels); and within 20 min, the MAPK activity returned to basal levels (P > 0.05).

The Ets family transcription factor, Elk-1 is a physiological substrate for ERK1/2 MAPK. To investigate whether melatonin phosphorylates Elk-1, hGLCs were treated with 1 nm melatonin for 5 min. As shown in Fig. 7, melatonin (1 nm) stimulated a significant increase in Elk-1 phosphorylation. As expected, active p42 MAPK (ERK-2) substantially stimulated Elk-1 phosphosphorylation.

Discussion

It is generally thought that melatonin exerts its reproductive actions at the level of the brain and pituitary (3–6). However, the presence of high levels of melatonin in follicular fluid (8) and its binding sites in the granulosa-luteal cells (10, 11) suggest that melatonin exerts its actions via a receptor-mediated event at the level of the ovary, to modulate steroidogenesis (12, 13) and possibly luteolysis. Although the emphasis on the role of melatonin in the hypothalamus and pituitary has overshadowed the events within the ovary itself, we now confirm our preliminary data (22) and the report by Niles et al. (23) that melatonin mt₁ and MT₂ receptors are expressed in hGLCs. The present study demonstrates, for the first time, that melatonin regulates LHR, GnRH, and GnRHR and that its actions may be mediated via the MAPK pathway.

The radioreceptor assays for peripheral sites of melatonin action have reported high-affinity binding sites for 2-[¹²⁵I]iodomelatonin. These binding sites are linked to a G protein, and the specificity of iodomelatonin binding is characteristic of a typical melatonin receptor, as found in the brain and retina, suggesting that the molecular structure of the brain melatonin receptors may be identical to that of peripheral melatonin receptors (10, 11, 39–41). Sequence analysis of the fragments revealed that the ovarian mt₁ and MT₂ receptors were identical to the published brain and retina cDNA sequences, respectively. Similar results were also obtained when the melatonin receptors in the epididymis, uterus, and choriocarcinoma cells were compared with their brain counterparts (24–26, 42, 43). These findings are in line with the binding studies, which demonstrate significant pharmacological and biochemical similarities among the brain and peripheral melatonin receptors. The expression of melatonin receptors in hGLCs supports the hypothesis that melatonin may play a role in the direct regulation of ovarian function. However, the question remains as to the differential roles of the melatonin receptor subtypes, because mt₁ (but not MT₂) receptors were consistently detected in freshly isolated hGLCs (23). In our study, cultured hGLCs expressed both receptor subtypes, suggesting differential regulation of the receptors in response to microenvironmental factors. Possibly, high melatonin levels normally found in follicular fluid, but not in cell culture media, may be down-regulating the expression of the MT₂ receptors in hGLCs. Further studies are required to understand the regulation of the mt₁ and MT₂ receptors in hGLCs by hormones, melatonin, and other factors.

The ovary, during the fertile period, undergoes cyclic changes in morphology and function. The monthly profile results from cyclic changes in pituitary gonadotropins, LH and FSH, and also on the appropriate expression of their receptors (44, 45). It is known that FSH stimulates both estrogen and progestin production in cultured granulosa cells (46, 47) and induces the formation of LH/hCG receptor in granulosa cells in vivo and in vitro (44, 48). Given that melatonin can functionally alter steroidogenesis, as observed in this study and shown by others, to regulate ovarian function, it is possible that melatonin may be interacting with FSH at multiple cellular site(s) to modulate the amplitude of the transduction signal. Indeed, a melatonin-induced increase or decrease in granulosa cell FSHR may account, in part, for the additive or antagonistic interaction between these two hormones, respectively. However, our current findings suggest that melatonin is without significant effect on the regulation of granulosa cell FSHR gene expression.

The new formation of LH receptors (LHRs), induced by FSH, is an obligatory step in the differentiation and maturation of the granulosa cell and is essential for the initiation of luteinization (16, 45, 48). We have demonstrated that melatonin treatment in our culture system is capable of up-regulating LHR mRNA. Previous observations show that stimulatory effects of FSH on granulosa LH binding, induced by different hormones, are closely related to the regulation of LHR mRNA levels (49). Thus, our results indicate that physiological doses of melatonin may help maintain LHRs, which mediate the luteotropic action of LH. Because a significant reduction in available LHRs is associated with impaired ovarian function (50, 51), functionally, the dynamic changes in LHR during luteinization may suggest a possible regulatory role for melatonin during the luteal phase of the menstrual cycle. Because the protein kinase A pathway mediates the actions of LH and FSH on LHR number, it may be possible that melatonin affects gonadotropin-induced cAMP production in these cells and thus also alter progesterone production. As observed in this and other studies, melatonin stimulates hCG-induced progesterone production, possibly via the up-regulation of receptors for LH. The mechanism of melatonin action in altering LHR mRNA levels is unknown. Nonetheless, we have shown that melatonin can up-regulate LHR mRNA; and therefore, it is possible that melatonin may be involved in maintaining a critical level of LHR expression for ovarian function.

Another potential regulatory factor of ovarian steroidogenesis and ovarian follicle development is the peptide, GnRH (18, 20). GnRH has a direct inhibitory effect on hGLCs (18, 52). Results from the present study indicate that levels of GnRH and its receptor mRNA significantly decrease with melatonin treatment in vitro. Because the role of melatonin in reproduction has previously been focused at the hypothalamo-pituitary level, no information is available on the role of melatonin in regulating GnRH and its receptor in the human ovary. Reports have implicated GnRH as a luteolytic factor that can directly increase the incidence of apoptosis in granulosa cells (20, 53). At the level of the ovary, LH/hCG, an important rescue factor for the maintenance of the corpus luteum during the early stages of pregnancy, down-regulates GnRH-R, suggesting a possible involvement of the maintenance of the corpus luteum during pregnancy. In addition, GnRH and its receptor mRNA increase during spontaneous luteinization in vitro, suggesting that GnRH, in combination with other factors, may play an important role in corpus luteum regression (20). With this in mind, melatonin-induced down-regulation of GnRH and its receptor in hGLCs may play an important role in interfering with the demise of the corpus luteum during the mid- to late luteal phase. However, further studies are required to substantiate this hypothesis.

The mechanism of action of melatonin in the human ovary is not known. Melatonin in other tissues is known to be capable of modulating a variety of second messengers. We report, for the first time, that melatonin can induce a biphasic effect on MAPK activation. It has been shown that follicular fluid levels of melatonin follow a seasonal rhythm. High levels of melatonin during the winter months are associated with anovulation and high progesterone levels. In this study, MAPK activity was the highest after treatment with 1 nm melatonin. Progesterone secretion was also observed at this concentration of melatonin treatment. However, at higher concentrations of melatonin, MAPK activity decreased. This may be attributable to receptor regulation by melatonin itself. It is hypothesized that cellular responses to MAPK may be influenced by the duration of its activation, because sustained activation of MAPK is associated with cell differentiation, whereas transient activation of MAPK leads to cell proliferation (54, 55). Melatonin has been found capable of regulating cellular proliferation in epididymal cells and choriocarcinoma cells (26, 42) as well as regulating differentiation of human neuroblastoma cells (56). It is not known whether melatonin has any effect on proliferation in hGLCs, but activation of the MAPK by melatonin in hGLCs may be associated with differentiated cellular functions such as steroidogenesis. The functional significance of MAPK activation by melatonin in hGLCs awaits further investigation.

The ability of melatonin to activate a downstream effector of the MAPK pathway was investigated. Studies have shown that ternary complex factor proteins, such as Elk-1 and SAP-1 (57, 58), are phosphorylated by MAPK. Activated ternary complex factor proteins regulate the expression of coregulated genes, such as c-fos, through their actions with serum response element. This study demonstrated, for the first time, that treatment with melatonin resulted in a substantial activation of Elk-1 fusion protein in vitro. Further studies on melatonin activation of the MAPK cascade will help clarify the intracellular signaling pathway leading to cellular functions, such as modulation of steroidogenesis and gene transcription, in response to melatonin in hGLCs.

In summary, we have isolated the ovarian mt₁-R and MT₂-R transcripts that are equivalent to the mt₁-R and MT₂-R cDNA derived from the brain. In addition, we observed no effect of melatonin on FSHR expression except an up-regulation of LHR expression (which may mediate the stimulatory effects of melatonin on hCG-induced progesterone production) and down-regulation of GnRH and its receptor mRNA. Our findings demonstrate, for the first time, that melatonin activates the MAPK cascade. Taken together, our results strongly suggest that the action of melatonin is a receptor-mediated event that modulates cellular functions, possibly via a MAPK signal transduction mechanism.

Acknowledgements

We thank Dr. Margo Fluker and the Genesis Fertility Center (Vancouver, Canada) for the provision of hGLCs.

This study was supported by the Canadian Institutes of Health Research (to P.C.K.L.) and the Neuroendocrinology Fund (to S.F.P.). M.M.M.W. is a recipient of the Croucher Foundation scholarship. S.K.K. and P.S.N. are studentship recipients of the British Columbia Research Institute for Children’s and Women’s Health. P.C.K.L. is a Career Investigator of the British Columbia Research Institute of Children’s and Women’s Health.

Abbreviations:

 
• dNTP,

  Deoxynucleotide triphosphate;

 
• FSHR,

  FSH receptor;

 
• GnRHR,

  GnRH receptor;

 
• hCG,

  human CG;

 
• HEK,

  human embryonic kidney;

 
• hGLCs,

  human granulosa-luteal cells;

 
• LHR,

  LH receptor.