| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Département de Neurobiologie des Rythmes (F.G.R., M.S., P.P., V.S.), Centre National de la Recherche Scientifique Unité Mixte de Recherche-7168/LC2, Institut des Neurosciences Cellulaires et Intégratives, Université Louis Pasteur–Institut Fédératif de Recherche des Neurosciences, 67084 Strasbourg, France; and Department of Translational Neurobiology (F.G.R., J.D.M.), NeuroSearch A/S, Pederstrupvej 93, 2750 Ballerup, Denmark
Address all correspondence and requests for reprints to: Jens D. Mikkelsen, Department of Translational Neurobiology, NeuroSearch A/S, Pederstrupvej 93, 2750 Ballerup, Denmark. E-mail: JDM{at}Neurosearch.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The Syrian hamster (Mesocricetus auratus) and the Siberian hamster (Phodopus sungorus) are well-known photoperiodic models (2, 4). In both species, sexual activity is promoted by long-day photoperiod (LD), whereas exposure to short-day photoperiod (SD) triggers arrest of the reproductive activity, as manifested by testicular regression and decrease in serum gonadotropin and testosterone levels (4, 5, 6). However, if SD exposure is prolonged over 20–30 wk, hamsters become refractory to inhibitory day lengths (SD-R) and a complete recrudescence of the reproductive system occurs spontaneously (4). Recently, we have found that in the arcuate nucleus of the Syrian hamster, expression of the Kiss1 gene is controlled by photoperiod, via melatonin (24, 25). This gene encodes kisspeptins (Kp), the endogenous ligand(s) of the G protein-coupled receptor 54 (GPR54), which is critical for the regulation of reproduction, both in adults and at puberty onset (26, 27, 28, 29). Via GPR54, Kp activate GnRH neurons, which induces GnRH release and downstream gonadotropin secretion (26, 27, 28, 29). We have found that in the Syrian hamster, Kiss1 expression is down-regulated in SD photoperiod, via a melatonin-dependent mechanism (24, 25). Importantly, we have shown that chronic intracerebroventricular administration of Kp to sexually quiescent SD hamsters is sufficient to restore testicular activity (24, 25), strongly suggesting that Kiss1 and GPR54 are critical for mediating the effect of photoperiod on the reproductive axis.
Kp is a member of the Arg-Phe-amide (RFamide) family of peptides (30, 31, 32, 33), which also includes the products of the RFamide-related peptide gene (RFRP). The mammalian RFRP gene is orthologous to the gene encoding the peptide gonadotropin-inhibiting hormone (GnIH) in birds, which possesses inhibitory effects on gonadotropin release from the anterior pituitary (34). The RFRP gene gives rise to the two biologically active peptides RFRP-1 and RFRP-3, and intracerebroventricular injections of RFRP-3 reduces plasma levels of LH (35, 36). In contrast to Kiss1, which is expressed in the arcuate nucleus (24, 27, 28, 37), RFRP is expressed in the area in between the VMH and DMH (33, 35, 36, 38, 39, 40, 41, 42, 43, 44, 45). Interestingly, neurons in the same area have earlier been described as Kp immunoreactive (45, 46, 47). However, the specificity of this labeling has been highly questioned (27, 28, 45, 46, 47, 48, 49, 50), because no Kiss1 mRNA could be found in this area (27, 28, 37), and antisera cross-reactivity is highly possible due to the close sequence similarity of the peptides.
We found it interesting to examine further this population of neurons in photoperiodic species, because their location matches the site identified in the Syrian hamster for the action of melatonin on reproduction (16, 17, 18, 19, 20). We speculated that the gene product(s) recognized in this region by the Kp antisera may be that of another member of the family (39, 51, 52), presumably RFRP gene products. We first examined RFRP mRNA expression in the brain of two photoperiodic rodents, the Syrian hamster and the Siberian hamster, and we found in accordance with earlier studies in the rat that RFRP mRNA is also located in the medial hypothalamus. In these seasonal species, gonadal activity is modulated by day length, and we hypothesized that RFRP expression may also change in SD vs. LD hamsters. Given the biological roles suggested for RFRP and its avian ortholog GnIH (i.e. inhibitory action of the reproductive axis) (34, 35, 36, 53, 54, 55), we speculated that RFRP expression would increase in the medial hypothalamus of sexually inactive SD hamsters. We observed that RFRP displayed a photoperiodic-dependent expression in both Syrian and Siberian hamsters but, surprisingly, exhibited a reduced expression under short photoperiod. Thus, we performed a series of experiments in the Syrian hamster to better characterize this photoperiodic regulation. We examined the influence of gonadal hormones and melatonin and determined that melatonin is responsible for the photoperiodic regulation of RFRP expression. We discuss these results in the light of the presumed biological functions of RFRP peptides, and the implications for understanding the mechanisms of seasonal breeding are also exposed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The adult male Syrian (M. auratus) and Siberian (P. sungorus) hamsters used in these experiments were bred in-house. From birth they were maintained in LD photoperiod consisting of 14 h light (200 lux) and 10 h dark (2 lux dim red light) for the Syrian hamsters, or 16 h light and 8 h dark for the Siberian hamsters, with lights off at 1900 h, at 22 ± 2 C with ad libitum access to water and food. Adult male Han Wistar rats (250–350 g) were bred and obtained from the Faculté de Médecine (Strasbourg, France). The animals were allowed to adapt to our animal facility for 2 wk under standard laboratory conditions (12 h light, 12 h dark, lights off at 1900 h) before the lighting protocol was changed.
Tissue collection consistently occurred 3–4 h after lights on, except where indicated. Animals were anesthetized with isoflurane vapors and killed by decapitation. Brains were rapidly removed from the skull, snap-frozen at –30 C, and stored at –80 C until in situ hybridization. Testes were dissected out, measured, and weighed. When required, trunk blood was collected and centrifuged at 1500 x g for 15 min, and plasma was stored at –20 C until assayed.
Experimental design
Photoperiodic modulation of RFRP gene expression.
This experiment aimed to examine the photoperiodic expression of RFRP in the hypothalamus of two photoperiodic and one non-photoperiodic rodent model. Groups (n = 6) of Syrian and Siberian hamsters were placed in LD (Syrian hamsters, 14 h light, 10 h dark; Siberian hamsters, 16 h light, 8 h dark; for both, lights off at 1900 h) and SD (10 h light, 14 h dark; lights off at 1900 h) photoperiods for 10 wk. Wistar rats were divided into two groups (n = 4 per group) that were transferred for 10 wk into a LD photoperiod consisting of 16 h light, 8 h dark, or into a SD photoperiod consisting of 8 h light, 16 h dark.
The 24-h time course for RFRP gene expression.
This experiment aimed to examine whether RFRP displays a daily rhythm of expression, under both LD and SD photoperiods. Syrian hamsters were placed in LD and SD for 10 wk, after which they were killed every 4 h over 24 h, starting at 1800 h (n = 5 per group).
Photoperiodic change of RFRP immunoreactivity.
This experiment was to investigate whether the photoperiodic change of RFRP mRNA in hamsters is translated at the peptide level. Syrian and Siberian hamsters were raised for 10 wk under LD and SD (n = 5 per group) before they were processed for immunocytochemistry, as described below.
Modulation of RFRP expression by sex steroids.
This experiment was to determine whether the photoperiodic modulation of the RFRP expression in the Syrian hamster results from the secondary changes in circulating levels of gonadal steroids. Two groups (n = 5 per group) of LD Syrian hamsters were either castrated (LD-cast) or sham operated (LD-sham), and two groups (n = 5 per group) of SD Syrian hamsters were implanted with silastic capsules either filled with testosterone (SD-testo) or left empty (SD-empty). All animals were killed after 4 wk of treatment, after which RFRP expression was examined.
Modulation of RFRP expression by melatonin.
This experiment was to examine the role of melatonin in the photoperiodic regulation of RFRP expression in the Syrian hamster. Two groups of Syrian hamsters (n = 6 per group) were either pinealectomized (Pin-X) to remove the primary source of melatonin or underwent sham surgery (Pin-sham) before being transferred to SD for 10 wk. In addition, six groups of hamsters (n = 6 per group) were left in LD and given daily sc injections for 7, 21, or 60 d of either saline solution (Ringer-5% ethanol) or melatonin (Sigma-Aldrich, Lyon, France) (50 µg in Ringer-5% ethanol) to mimic the effects of SD. Injections occurred 1.5 h before lights-off, as previously described.
RFRP gene expression in photorefractory Syrian hamsters.
This experiment aimed to investigate RFRP expression in SD-refractory (SD-R) Syrian hamsters, which undergo spontaneous gonadal reactivation after prolonged exposure to SD. Two groups of SD-R hamsters were prepared by leaving the animals in SD for 28 wk, whereas age-matched control hamsters were left in LD (n = 6 per group). The animals were then processed for in situ hybridization and immunocytochemistry, as described below.
Surgical procedures
Castration, steroid replacement and Pin-X were performed under anesthesia as previously described (11, 24). For anesthesia, a mixture of Zoletil 20 (Virbac, Carros, France) and Rompun (Bayer Pharma, Puteaux, France) was used. For castration, the gonads of LD animals were removed, whereas control animals left intact underwent sham surgery. For steroid replacement, testosterone (4-androsten-17β-ol-3one; Sigma)-filled silastic capsules (inner diameter, 1.47 mm; outer diameter, 1.95 mm; length, 13 mm) were implanted sc to SD adapted hamsters (testicular regression verified by scrotal palpation), whereas control untreated animals received empty capsules. For Pin-X, a circular hole was drilled in the skull and the pineal gland was extracted or left in place (Pin-sham).
In situ hybridization histochemistry
Coronal brain sections (20 µm) were cut through the mediobasal and posterior hypothalamic regions in seven similar series of sections. The sections were thaw-mounted and stored at –80 C until required. Radioactive in situ hybridization was carried out as previously described (11, 56). RFRP mRNA was probed with an antisense oligonucleotide labeled with [35S]deoxy-ATP (1250 Ci/mmol; PerkinElmer, Zaventrum, Belgium) using terminal transferase (Roche, Meylan, France). The sequence of the oligonucleotide used for quantitative analysis was 5'-GAC TCA TCT TAA TAT CTT TCT TTG CCC CCC AAT CTT TGA GTT CTT GAA A-3', based on the rat RFRP sequence (bases 181–229 from GenBank accession no. NM_023952). Another nucleotide sequence complementary to the bases 431–480 used by Pertovaara et al. (39) produced identical results. The signal obtained with both probes matched that previously described in the literature and was clearly absent from the arcuate nucleus where Kiss1 is expressed. The sections were rapidly thawed, fixed in 4% paraformaldehyde, acetylated in triethanolamine buffer, and dehydrated in graded ethanols. The radiolabeled oligoprobe was added at a specific activity of 107 cpm/ml to the hybridization medium containing 50% formamide (vol/vol), 4x saline sodium citrate (SSC) (1x SSC is 0.15 M NaCl, 0.015 M sodium citrate·2H2O, pH 7.2), 1x Denhardts solution (0.02% Ficoll, polyvinylpyrrolidone, and BSA), salmon sperm ssDNA (0.5 mg/ml), 0.25 mg/ml yeast tRNA, 10% (wt/vol) dextran sulfate, and 10 mM dithiothreitol. The hybridization mixture was applied to slides, which were coverslipped and incubated overnight at 37 C in a humid chamber. The sections were washed four times for 15 min in 1x SSC at 55 C, followed by two times for 30 min in 1x SSC at room temperature. Finally, the sections were dehydrated, and all were exposed against BioMax MR Film for 7 d together with 14C radioactive standards to allow standardization of densitometric measurements. Selected sections were dipped in EM-1 nuclear emulsion (Amersham Bioscience, Arlington, IL) for 2 wk before being developed for microscopic localization of RFRP mRNA transcripts. These sections were counterstained in thionin to determine anatomical structures in details. For semiquantitative analysis, x-ray films were scanned on an Epson 4990 transmittance scanner, and background-subtracted calibrated OD measurements of gene expression were performed using ImageJ (NIH). Integrated density was measured on three sections from the same slide [region situated 2–2.6 mm posterior to the bregma for Syrian hamsters, according to the stereotaxic atlas of the golden hamster brain, Morin and Wood (64)] to calculate the average integrated density per animal. We did not detect any subregional differences for the photoperiodic regulation of RFRP expression, and the analysis was performed by measuring the OD of the entire signal for RFRP. An observer blind to treatment identity performed all analyses.
Immunohistochemistry
Animals were deeply anesthetized with isoflurane, and transcardially perfused with physiological saline (with heparin) and then with 4% paraformaldehyde in 0.1 M phosphate buffer. The brains were removed from the skull, postfixed overnight, stored in PBS, and cryoprotected in 30% sucrose/PBS for 48 h before sectioning in series of five (Syrian hamsters) and four (Siberian hamsters) at a freezing microtome (40 µm). Immunostaining was performed on free-floating sections as described in a previous report (57), using the avidin-biotin method and diaminobenzidine as chromogen. To avoid animal-animal and section-section variability, one series of sections from all animals were incubated in the same dilutions of the same antisera as indicated below. Briefly before incubating in the primary antiserum, the sections were incubated three times for 5 min in PBS, for 10 min in 1% H2O2-PBS, and for 20 min in 5% swine serum mixed in PBS with 1% Triton X-100.
The sections were incubated overnight at 4 C in the primary antiserum. Two commercial polyclonal antisera raised in rabbits were used for this study. One antiserum was raised against YNWNSFGLRF-amide or human Kp10 (hKp10) (catalog item H-048-56, lot 00064; Phoenix Pharmaceuticals Inc., Belmont, CA), and the other was raised against FMRF-amide (catalog item RA20002, lot 400206; Neuromics, Edina, NM). The antiserum from Phoenix was diluted 1:1500, and the antiserum from Neuromics was diluted 1:5000. Both antisera produced similar strong staining of neuronal cell bodies in the same area of the hypothalamus, and this staining could be completely abolished by preincubation (10 µM) with RFRP-1, RFRP-3, hKp10, and FMRF-amide. After careful washing in PBS with 0.3% Triton X-100 for three times 10 min, the sections were processed by incubating in a biotinylated donkey antirabbit (dilution 1:1000; Jackson Laboratories, Bar Harbor, ME) and subsequently incubated in an avidin-biotinylated horseradish peroxidase complex (Vector Elite Kit; Vector Labs, Burlingame, CA; diluted 1:125), both for 60 min. Finally, the sections were reacted for peroxidase activity by incubation for 20 min with a solution of 0.125% diaminobenzidine, rinsed twice for 10 min in distilled water, and mounted on gelatinized glass slides. The number of RFRP-immunoreactive cell bodies was counted through the entire length of the mediobasal hypothalamus by an observer not aware of the nature of the animals. The number of cell bodies was counted in a bright-field microscope (Zeiss Imager Z1) at x20 magnification in one series of sections (every fifth section for Syrians and every fourth for Siberians). The number of positive cells counted from each animal was multiplied by the number of series to give a rough estimate of the total number of cells per brain.
Testosterone assay
Plasma concentrations of testosterone were determined by RIA as described (58). Briefly, 3 ml diethyl ether was added to 50 µl of the blood sample, which was then vortexed for 1 min and centrifuged for 5 min (4 C, 2000 rpm). The ether phase was decanted and evaporated, and the dried extract was redissolved in 300 µl phosphate buffer before tritiated testosterone (1000 cpm) (Amersham Pharmacia Biotech Europe, Orsay, France) was added. The testosterone concentration was determined by using duplicate aliquots of the extracts, which were redissolved in 0.01 M PBS (pH 7.4) containing 0.1% BSA and incubated overnight at 4 C with approximately 6000 cpm of [3H]testosterone and a specific antibody. The specific testosterone antibody was kindly provided by Dr. G. Picaper (Médecine Nucléaire, La Source, France). Bound and free fractions were separated by absorption with dextran-coated charcoal and centrifuged. Aliquots of the bound fractions were counted with a Packard 1600 liquid scintillator counter. The sensitivity of the assay was 50 pg/ml, and the intra- and interassay coefficients of variation were 8 and 9.5%, respectively.
Statistical analyses
Results are shown as mean ± SEM. Data were analyzed by t test, Mann-Whitney rank sum test, or two-way ANOVA, followed by Tukey’s analysis, as appropriate. Statistical significance was set at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A–D). A group of RFRP mRNA-containing neurons was also observed medial and dorsal to the DMH, inside the boundaries of the DMH as well as in the periventricular area at the same rostrocaudal level of the hypothalamus (Fig. 1, A–D). RFRP mRNA expression in all groups of neurons was clearly diminished in Syrian hamsters maintained in SD for 10 wk (Fig. 1E). SD animals underwent testicular regression [Fig. 1F; paired testicular weights (PTW): 0.30 ± 0.05 g vs. 4.00 ± 0.19 g in LD hamsters] and displayed RFRP mRNA level significantly lower than in LD animals (Fig. 1F; P < 0.001). The RFRP mRNA level in SD was about 28% of that in LD.
|
). By two-way ANOVA, there was an overall effect of photoperiod on the mRNA level (F = 72.793; P < 0.001), but no effect of time (F = 0.355; P = 0.876) and no interaction between photoperiod and time (F = 0.287; P = 0.918). At all time points, multiple pair-wise comparisons showed that LD and SD values were significantly different (P < 0.01).
Similar to Syrian hamsters, RFRP expression was detected in the same part of the hypothalamus in Siberian hamsters kept under LD, and the mRNA level was highly depressed in SD animals (Fig. 2, A and C; P < 0.001). In SD-adapted Siberian hamsters, RFRP expression was about 18% of that in LD animals (Fig. 2C). The mean PTW in these animals was 690 ± 50 mg in LD and 40 ± 10 mg in SD.
|
), with no significant variations of RFRP mRNA level (Fig. 2B; P = 0.428). In the rat, the positive neurons were confined more exclusively to the zone in between the VMH and the DMH (Fig. 2B). In the posterior hypothalamus at the level of the very caudal part of the DMH, a group of RFRP mRNA-expressing neurons was present above the third ventricle (not shown).
Photoperiodic change of RFRP immunoreactivity
The male Syrian and Siberian hamsters used for immunohistochemical analysis were maintained in SD for 10 wk, and it was observed that they all underwent testicular regression. For the Syrian hamsters, the PTW was 3.26 ± 0.46 g in LD and 0.3 ± 0.05 g in SD. For the Siberian hamsters, the PTW was 688.33 ± 47.84 mg in LD and 30.93 ± 1.93 mg in SD.
In LD Syrian hamsters, a strong RFRP-immunoreactive staining was observed in the neuronal cell bodies in the same area as observed to contain RFRP mRNA (Fig. 3, A and B). This distribution matches that previously described for RFRP immunoreactivity in rodents (35, 36, 41, 42, 43, 44), including the Syrian hamster (36). Furthermore, despite being raised against two different RFamide peptides, the two antisera produced virtually the same staining. Preincubation of both antisera with RFRP-1, RFRP-3, mouse Kp10 (mKp10), and human Kp10 (hKp-10) totally abolished the staining, and application of the same peptides on dot-blots revealed that both antisera bind to all three peptides, although with apparent lower affinity for mKp10 than RFRP-1 (not shown). The quantification shown here was carried out in sections from animals incubated with the antiserum from Phoenix Pharmaceuticals.
|
). Quantification of the number of cells in LD and SD animals revealed that significantly fewer immunoreactive cells were present in the hypothalamus of SD compared with LD animals (Fig. 3G). In Syrian hamsters, the number of immunoreactive cell bodies was 50% less in SD compared with LD condition (Fig. 3G; LD, 346.0 ± 29.5 cells per brain; SD, 189.0 ± 12.8 cells per brain; P = 0.001).
Similarly, the number of cells and the intensity of positive staining in hypothalamic neurons of Siberian hamsters kept under LD (Fig. 3, D and E) was much higher than in animals kept under SD (Fig. 3F). Compared with the Syrian hamster, the staining intensity of individual cells in the Siberian hamster appeared stronger, and the SD-induced decrease in the number of immunoreactive cell bodies was even higher (Fig. 3G; LD, 345.3 ± 28.8 cells per brain; SD, 41.3 ± 12.5 cells per brain; P < 0.001). Also, the morphology of the RFRP-immunoreactive neurons in the Syrian hamster was slightly different from the Siberian hamster exposed to LD (Fig. 3, B and E). The positive neurons contained relatively little immunoreactive material in the cell soma in the Syrian hamster compared with the Siberian (Fig. 3, B and E). However, the presence of RFRP-positive nerve fibers was found in about the same quantity and regions in the two species (not shown).
Modulation of photoperiodic RFRP expression by gonadal steroids
The role of gonadal steroids in the photoperiodic modulation of the RFRP gene in Syrian hamsters was examined. There was no statistical difference in RFRP expression between LD Syrian hamsters that were castrated (LD-cast) or sham-operated (LD-sham), although we observed a nonsignificant tendency for increased expression in castrated animals (Fig. 4A; P = 0.053). In addition, two groups of SD-adapted hamsters were implanted for 4 wk with empty (SD-empty) or testosterone-filled (SD-testo) capsules to restore blood concentrations similar to those observed in LD. The plasma testosterone level in SD animals receiving an empty capsule was 0.33 ± 0.13 ng/ml, whereas that of SD animals receiving testosterone was 2.17 ± 0.52 ng/ml. This level in the SD-testo group was not statistically different from LD hamsters (3.24 ± 0.42 ng/ml; P = 0.219), whereas both were significantly different from SD-empty animals (LD vs. SD-empty, P = 0.003; SD-testo vs. SD-empty, P = 0.028). The expression of RFRP in SD hamsters treated with testosterone for 4 wk was not statistically different from control hamsters receiving an empty capsule (Fig. 4B; P = 0.663), demonstrating that testosterone treatment alone was not sufficient to restore RFRP expression to LD levels. Both SD-empty and SD-testo animals had reduced RFRP mRNA level compared with LD control animals (P < 0.001). It is concluded from these experiments that the photoperiodic change of RFRP expression is not directly modulated by sex steroids.
|
; P < 0.001). Importantly, the magnitude of difference in RFRP expression between Pin-X and Pin-sham animals (Fig. 5A) was comparable to that observed between LD and SD animals.
|
). In melatonin-treated LD animals, testicular weight was not changed significantly after 7 or 21 d of treatment (Fig. 5B; P > 0.05), whereas it was significantly reduced in hamsters treated for 60 d (P < 0.001). We observed that RFRP expression was significantly inhibited in LD hamsters treated with melatonin for 21 d (P = 0.02 vs. vehicle 21 d; P = 0.03 vs. melatonin 7 d) and 60 d (P < 0.001 vs. vehicle; P < 0.001 vs. melatonin 7 and 21 d) but not 7 d (P > 0.05 vs. vehicle at 7, 21, and 60 d) (Fig. 5C). The decrease of RFRP mRNA level in melatonin-treated animals was progressive, and 60 d of treatment were necessary to have the RFRP expression level comparable to that observed in SD animals. In contrast, vehicle injections did not alter RFRP mRNA levels.
Expression of RFRP in photorefractory Syrian hamster
We examined RFRP expression in SD-R Syrian hamsters. Animals kept in SD for 28 wk became refractory to SD and underwent gonadal recrudescence. Accordingly, testicular weight in SD-R hamsters was not significantly different from that of LD hamsters (PTW = 3.58 ± 0.24 g for SD-R hamsters vs. 3.78 ± 0.11 g for LD animals). RFRP expression remained low in the DMH of SD-R hamsters (Fig. 6A), with mRNA level being 43% of that in LD animals (Fig. 6C; P = 0.002). It tended to be higher than in SD animals, although this was not significant (Fig. 6C; P = 0.45). Examination of immunoreactivity to RFRP revealed that the number of stained cell bodies in SD-R animals (256.7 ± 31.0 cells per brain) was not significantly different from that in LD hamsters (Fig. 6, B and C). This number tended to be lower than in LD animals, being intermediate between that of LD and SD hamsters (one-way ANOVA: LD vs. SD-R, P = 0.078; LD vs. SD, P = 0.004; SD vs. SD-R, P = 0.205).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In mammals, only five members of the RFamide family have been identified so far (30, 31, 32, 33), and for each member, a specific gene encodes one or more RFamide peptides (30, 31). The RFRP (also found as NPVF) (32, 33) gene encodes two or three putative peptides (RFRP-1, -2, and -3), depending on species (32, 33, 38). In mammals, only two peptides are processed, RFRP-1 and -3, whereas the predicted amino acid sequence for the C terminus of RFRP-2 that is seen in birds is, instead, included at the N terminus of RFRP-3 (32). Furthermore, in rodents, the RFRP gene lacks the sequence for RFRP-2 (32, 36, 38). Several reports have shown that polyclonal antisera raised against members of the RFamide family of peptides are likely to cross-react with other members (36, 45, 47, 50, 51, 52). This is also the case in this study, where an antiserum raised against hKp10 cross-reacts with both RFRP-1 and RFRP-3. We have shown that the detected immunoreactivity can be completely abolished after preincubation with RFRP-1 and -3, and the antiserum binds to both RFRP-1, RFRP-3 and mKp10 in dot-blots (unpublished data). This concords with the data published by Brailoiu et al. (46), who reported immunoreactive cells in the dorso-hypothalamic hypothalamus using the same anti-hKp10 antiserum. Furthermore, Kiss1 mRNA has not been detected in the region of the dorsomedial hypothalamus where RFRP neurons are present but in the arcuate nucleus and in the anteroventral periventricular nucleus (24, 27, 28, 37). Based on these data, we can conclude that the immunoreactive neurons very likely contain RFRP-1/RFRP-3 and are accordingly termed RFRP immunoreactive.
So far, there has been limited information on the biological roles of RFRP peptides in mammals. Both peptides may possess feeding modulatory actions (30). In rat, RFRP-1 indirectly stimulates prolactin secretion (33, 59) and modulates nociceptive responses (38, 39). RFRP-3 increases GH secretion and food intake but, importantly, strongly inhibits LH (but not FSH) release and sex behavior (35, 36). Interestingly, the mammalian RFRP gene has been found to be orthologous to the GnIH gene in birds. This gene encodes for three putative peptides termed GnIH, GnIH-related peptide-1, and GnIH-related peptide-2, designated according to the inhibitory activity of the GnIH peptide on gonadotropin secretion (34, 53, 54, 55).
The negative effects on LH release of RFRP-3 in mammals (35, 36) and GnIH in birds (34, 53, 54, 55) has suggested that RFRP-3 and GnIH are functional homologs. But in contrast with birds where GnIH affects gonadotropin release directly at the pituitary level, RFRP-3 in mammals may modulate gonadotropin secretion indirectly, via GnRH neurons. Several studies in rodents (including the Syrian hamster) have reported few RFRP-immunoreactive fibers in the outer layer of the median eminence, whereas RFRP-immunoreactive fibers contact a large percentage of GnRH cells (35, 36). Clearly, further studies are needed to understand how RFRP-3 modulates LH release in mammals, in particular to determine whether it affects LH secretion at the pituitary level and/or via GnRH release. In the latter possibility, if RFRP peptides modulate the activity of GnRH neurons, one can speculate that it occurs via direct neuronal connection between RFRP and GnRH neurons, and/or indirectly via contacts to Kp neurons in the arcuate nucleus.
We observed that RFRP expression is reduced in SD hamsters, both at the mRNA and peptide level, and this result may appear surprising, for two reasons. First, if the major role of RFRP was to inhibit the gonadotropic axis, we would expect RFRP expression to increase in SD hamsters, so that the negative drive on reproduction is increased, in parallel to the decrease of the positive drive represented by Kp (24, 25). One can suggest that the role of RFRP in hamsters would be to fine-tune the activity of the hypothalamo-pituitary-gonadal axis. This may occur directly on GnRH neurons, and/or via Kiss1 neurons or other intermediates. This possibility would be particularly relevant, given the extreme potency of Kp as secretagogue of GnRH release. Second, our observation is reversed compared with the results reported for GnIH in quails (see below) (60). Both Syrian and Siberian hamsters and quails are long-day breeders; however, the photoperiod controls reproduction differently in birds and hamsters. Finally, the down-regulation of RFRP expression in SD hamsters also raises a number of questions on the role of RFRP as a gonadotropin-inhibitory factor. The biological action of RFRP peptides may extend far beyond the modulation of reproductive activity, and RFRP-1 and RFRP-3 peptides may fulfill different biological roles (33, 35, 38, 39, 59). RFRP-immunoreactive nerve fibers are broadly distributed in the brain (35, 36, 41, 42), suggesting that RFRP peptides may control a large palette of biological functions. The receptor for RFRP peptides, termed RFRP-R (also found as FF1 or OT7T22), is a G protein-coupled receptor that is widely expressed in the forebrain (38). Importantly, because the RFRP and NPFF peptides (another member of the RFamide peptide family) share significant homology, they have been reported to interact with each other’s receptor (38, 39, 54). Consequently, biological actions such as central anti-opioid effect initially attributed to NPFF may occur via RFRP receptor, and may thus reflect endogenous RFRP activity (38). If RFRP peptides possess true pain-modulatory actions, then interesting consequences exist in term of photoperiodic variations.
In addition, the regulation of RFRP mRNA level, RFRP propeptide, and its maturation as RFRP-1 and RFRP-3 should be carefully dissociated. Further experiments using antisera specific for either RFRP-1 or RFRP-3 (35, 41, 42, 43, 44) should allow one to distinguish putative differential processing of the two peptides by melatonin. It was recently reported that photoperiod affects processing of the neuropeptide precursor proopiomelanocortin in the hypothalamus of Siberian hamsters by affecting the expression of prohormone convertases 1 and 2 (61). Accordingly, independent from RFRP mRNA levels, photoperiodic-dependent regulation of RFRP-1/3 processing may occur as well as differential processing of RFRP-1 and RFRP-3. In SD-R Syrian hamsters, we observed that RFRP expression was still reduced compared with LD hamsters, although it was higher than in SD animals. In contrast, the number of RFRP-immunoreactive cell bodies in SD-R hamsters was higher than in SD animals. This differential regulation between mRNA and peptides requires further investigations but suggests that although RFRP gene expression might continue to be partially inhibited by melatonin, translation and processing of RFRP peptides may become refractory independently.
The photoperiodic change for RFRP expression did not result from a difference in circadian expression profiles between LD and SD, because we did not detect significant 24-h expression rhythms in the Syrian hamster, RFRP mRNA level being constantly elevated and reduced in LD and SD animals, respectively. In contrast to hamsters, RFRP expression was not influenced by day length in the laboratory Wistar rats that do not display photoperiodic variation in testicular function. These results suggest that only in seasonal species is RFRP expression modified by photoperiod. One can therefore assume that further investigations of this differential regulation may help understanding the difference between photoperiodic and non-photoperiodic species.
Hamsters maintained in SD have fully regressed gonads and consequently reduced levels of circulating sex hormones. Given that sex steroids control the expression of a wide range of genes, we verified that the photoperiodic change of RFRP expression was not a secondary consequence of the reduction in circulating testosterone in SD animals. Castrated LD hamsters had high RFRP mRNA level that was comparable to that of intact animals, whereas RFRP expression remained low in SD hamsters implanted with testosterone.
In mammals, the changes in photoperiod are translated by variations in the melatonin daily profile, and in Syrian hamsters, suppression of melatonin through ablation of the pineal gland renders the animals blind to photoperiod, so that exposure to SD is no longer able to inhibit sexual activity. Accordingly, we observed that Pin-X hamsters failed to show a SD-induced decline in RFRP mRNA level relative to sham-operated control animals. Altogether, these data strongly suggest that melatonin mediates the photoperiodic regulation of RFRP expression in Syrian hamsters without direct influence from gonadal hormones. In addition, we performed melatonin injection experiments in LD Syrian hamsters. Animals injected daily for 60 d with melatonin during the late light phase displayed reduced RFRP expression, comparable to SD hamsters, confirming the results of the Pin-X experiment. A 7-d treatment was not sufficient for changing RFRP expression in LD hamsters, and a partially reduced mRNA level was noted after 21 d of treatment. Thus, the melatonin-dependent inhibition of RFRP expression does not appear to be a fast, early event in the physiological adaptation to short photoperiod. This contrasts with the type 2 deiodinase (Dio2) gene, which is expressed in the blood-brain barrier-free zone of the arcuate nucleus. Dramatic inhibition of Dio2 expression occurs in SD Syrian hamsters, and 7 d (or less) of melatonin treatment is sufficient to suppress Dio2 mRNA level in LD Syrian hamsters (11). Thus, regulation of Dio2 and RFRP expression by melatonin presumably involves different mechanisms. Whereas Dio2 may be very sensitive to changes in daily melatonin patterns, inhibition of RFRP expression presumably necessitates integrating recurrent long SD melatonin profiles.
In birds, the GnIH gene is expressed in the paraventricular hypothalamic nuclei (53, 62) and, interestingly, also displays seasonal-dependent variations (60, 62, 63). In sparrows, the GnIH-immunoreactive neuron area was reported to be greater at the onset of photorefractoriness to LD compared with photosensitive (SD) or photostimulated (LD) birds (63). In addition, GnIH expression and GnIH cell content both increase in quails under SD compared with LD photostimulated birds (60). Importantly, Ubuka et al. (60) demonstrated that this photoperiodic modulation of GnIH expression is fully dependent upon the melatonin released from the pineal gland and the eyes, and they showed that GnIH-immunoreactive neurons in the paraventricular hypothalamus coexpress the Mel1c melatonin receptor (63). This suggested that melatonin acts directly on GnIH-expressing neurons through Mel1c-mediated mechanisms to induce GnIH expression in SD. Similar to GnIH in birds, we demonstrate here that melatonin also controls RFRP expression in hamsters (but not in Wistar rats), although the direction of this regulation appears to be opposite, possibly due to several reasons (see above). Whether melatonin acts directly on RFRP-expressing cells of hamsters will necessitate further investigations. Interestingly, the location of these neurons strikingly corresponds to the site that in Syrian hamsters mediates the action of melatonin on reproduction (16, 17, 18, 19, 20). Not only have melatonin-binding sites been reported in the dorsomedial hypothalamus, but also bilateral lesion of this area results in the loss of gonadotropic response to melatonin (19, 20). Thus, it is likely that the photoperiodic change in RFRP expression is mediated by those melatonin receptors described in the same zone (18, 19, 20).
One can further speculate that in cells expressing RFRP, additional genes could be regulated, either activated or inhibited, by melatonin, with a dynamic possibly different from that of RFRP. So far, the identification of the melatonin-responsive cells controlling reproduction has been remarkably challenging (4, 5, 6, 9). Our results now make it possible to spot a defined population of cells in the medial hypothalamus that respond to melatonin, thus opening new tracks for investigating the molecular bases of melatonin action in the brain.
Finally, RFRP expression may be modulated by factors other than melatonin. One study reported that RFRP-immunoreactive cells express estrogen receptor-
and exhibit c-Fos expression shortly after estradiol injections (36). In our experiments, however, we did not observe long-term effects of gonadal steroids on RFRP expression. Clearly, further investigations should be carried out to examine whether and how additional factors can influence RFRP expression.
In summary, we show that in both Syrian and Siberian hamsters, the RFRP gene is modulated by photoperiod, with low mRNA levels and fewer immunoreactive cell bodies in SD compared with LD animals. This photoperiodic variation in the Syrian hamster is independent from secondary changes in gonadal steroids. Rather, RFRP expression is inhibited by SD melatoninergic signals. In contrast, there was no evidence for photoperiodic modulation of RFRP in the Wistar rat, a strain of rat showing no photoperiodic variation in testicular function. These results show for the first time in mammals that RFRP is a gene regulated by photoperiod, via melatonin. Identification of such melatonin-responsive cells located in the dorsal and medial hypothalamus of hamsters opens exciting perspectives for understanding how melatonin controls seasonal processes at the molecular level.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: DMH, Dorsomedial hypothalamic nucleus; GnIH, gonadotropin-inhibiting hormone; GPR54, G protein-coupled receptor 54; hKp10, human kisspeptin 10; Kp, kisspeptin; LD, long-day photoperiod; mKp10, mouse Kp10; Pin-X, pinealectomy; PTWs, paired testicular weights; RFRP, RFamide-related peptide; SD, short-day photoperiod; SD-R, short-day refractory; SSC, saline sodium citrate; VMH, ventromedial hypothalamic nucleus.
This work was supported in part by CIRIUS Denmark, the French Embassy in Denmark, the Centre National de la Recherche Scientifique, and the ACI "Plates-formes d’explorations fonctionnelles thématiques" from the Ministère Délégué à la Recherche.
Disclosure Statement: The authors have nothing to disclose.
Received June 25, 2007.
Accepted for publication December 5, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
with body weight. Endocrinology 145:13–20. Neurosci Lett 401:225–230[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  W. P. Williams III, E. M. Gibson, C. Wang, S. Tjho, N. Khattar, G. E. Bentley, K. Tsutsui, and L. J. Kriegsfeld Proximate mechanisms driving circadian control of neuroendocrine function: Lessons from the young and old Integr. Comp. Biol., November 1, 2009; 49(5): 519 - 537. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Z. Rizwan, R. Porteous, A. E. Herbison, and G. M. Anderson Cells Expressing RFamide-Related Peptide-1/3, the Mammalian Gonadotropin-Inhibitory Hormone Orthologs, Are Not Hypophysiotropic Neuroendocrine Neurons in the Rat Endocrinology, March 1, 2009; 150(3): 1413 - 1420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ubuka, N. L. McGuire, R. M. Calisi, N. Perfito, and G. E. Bentley The control of reproductive physiology and behavior by gonadotropin-inhibitory hormone Integr. Comp. Biol., November 1, 2008; 48(5): 560 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Smith, L. M. Coolen, L. J. Kriegsfeld, I. P. Sari, M. R. Jaafarzadehshirazi, M. Maltby, K. Bateman, R. L. Goodman, A. J. Tilbrook, T. Ubuka, et al. Variation in Kisspeptin and RFamide-Related Peptide (RFRP) Expression and Terminal Connections to Gonadotropin-Releasing Hormone Neurons in the Brain: A Novel Medium for Seasonal Breeding in the Sheep Endocrinology, November 1, 2008; 149(11): 5770 - 5782. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J Greives, L. J Kriegsfeld, G. E Bentley, K. Tsutsui, and G. E Demas Recent advances in reproductive neuroendocrinology: a role for RFamide peptides in seasonal reproduction? Proc R Soc B, September 7, 2008; 275(1646): 1943 - 1951. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. J. P. Ebling and S. M. Luckman RFAmide-Related Peptide: Another Sexy Peptide? Endocrinology, March 1, 2008; 149(3): 899 - 901. [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 |
) and SD (
) photoperiods. Regardless of the time of day, RFRP expression was constantly higher in LD compared with SD, with no pronounced 24-h rhythm (two-way ANOVA). Black and white bars represent the dark and light periods, respectively, with the LD photoperiod on top. Data are expressed as a percentage of the maximum LD value and represent the mean ± SEM; n = 4–5 per group.
0.001, one-way ANOVA. For RFRP immunoreactivity, bars with different letters differ significantly, one-way ANOVA. n = 4–6 per group.