{"paper_id":"c791c0f3-4fb9-4b34-926b-6578d21d94e4","body_text":"Gonadotropin-releasing hormone (GnRH), the\nmain stimulator of gonadotropes and secretion of\ngonadotropins, was first purified from the pig and\nsheep hypothalami in the 1970s ( 1 ,  2 ). For years\nGnRH has been considered the only regulator of the\nhypothalamic-hypophyseal-gonadal axis. Gonadal\nsteroids and inhibin regulate gonadotropin secretion\nvia negative/positive feedback mechanisms.\nAlthough existence of a hypothalamic inhibitor\nof gonadotropin secretion was suspected earlier\n( 3 ), in 2000 researchers discovered a 12 amino\nacid peptide (SIKPSAYLPLRFamide) in the quail\nbrain which could directly inhibit GnRH release. It\nwas subsequently named the gonadotropin inhibitory\nhormone (GnIH) ( 4 ). During the last 13 years,\navian homologues of GnIH have been identified in\nseveral mammalian species and named argininephenylalanine-\namide (RFamide)-related peptides\n(RFRP). In this review we describe the chemical\nstructure, biosynthesis and functions of RFRPs related\nto mammalian reproduction and their possible\nroles in other physiologic events.\nThe RFRPs are a family of peptides with an arginine-\nphenylalanine (RF-NH 2 ) sequence at their\ncarboxyl terminals. Researchers have discovered\nthe first peptide of this family in shell ganglions\n(FMRFamide) ( 5 ). The first RFRP in vertebrates\nwas discovered in the avian brain (LPLRFamide)\n( 6 ). In 2000 researchers reported that one of the\nRFRPs inhibited the secretion of gonadotropins.\nSince then, GnIH homologues have been identified\n( Table 1 ) in several species of mammals, including\nhumans ( 7 ), monkeys ( 8 ), cattle ( 9 ), sheep\n( 10 ,  11 ), rats, mice ( 12 ) and hamsters ( 13 ).\nFollowing transcription and translation of the\n RFRP  gene, a prepeptide is synthesized which routinely\nseparates into two mature peptides, RFRP-\n1 and RFRP-3 ( Table 1 ). The carboxy terminals\nof RFRPs contain a sequence of leucine-proline-XXX-arginine-phenylalanine (LPXRF, X=leucine\nor glutamine) followed by glycine (G) as an amidation\nsignal, and arginine (R) or lysine (K) that act as\nendoproteolytic basic amino acids ( 14 ). However, in\nhumans, monkeys, and cattle, RFRP-2 is also built\nfrom a prepeptide which differs from LPXRF. This\nprepeptide contains RS-amide sequences or an RSamide\nin the carboxyl terminal. RFRP-1 and RFRP-3\nbind the same receptor, named GPR147 (also known\nas OT7T022 and NPFF-1) with similar affinity ( 15 ).\nAmino acid sequence of the RFamide-related peptide (RFRP) prepeptide in different mammalian species\nRFRP neuronal bodies have been detected in\nthe rat dorsomedial hypothalamus and proven\nby a number of research studies using different\nantibodies. These antibodies included an antibody\nproduced against the sparrow GnIH produced\nin rabbits ( 16 ) and an antibody produced\nagainst the sequence 119-132 of prepeptide\nRFRP ( 17 ). We also reported similar findings in\nrats ( 18 ) by using an antibody against the quail\nGnIH produced in rabbits (supplied kindly by\nProfessor K. Tsutusi). The same neuronal extensions\nwere also found in the brains of hamsters\n( 13 ,  16 ) and mice ( 16 ). In sheep, neurons\nthat expressed RFRP were identified in the dorsomedial\nhypothalamic area (DMH), paraventricular\nnucleus (PVN), the area between these\nnuclei ( 11 ) and the preoptic area (POA) ( 19 ).\nWe showed that agouti-related peptide (AgRP)\nand RFRP coexpressed in 19 to 32% of the arcuate\n(Arc) neurons during various phases of\nthe estrous cycle in the ewe ( 20 ). In addition,\nwe observed similar neuronal extensions in the\nbrains of native Fars goats ( 21 ). Positive cells\nwere found in the monkey periventricular nuclei\n( 8 ) and human DMH ( 7 ).\nIn rodents, fibers and terminals of RFRP neurons\nwere observed in the middle areas of the\nbrain, limbic areas (POA, septal and amygdala),\nrostral hypothalamus and Arc ( 13 ,  16 ). In\nthe monkey brain, RFRP fibers were observed\nin most parts of the brain, including the hemispheres\nor telencephalon, septal nuclei and accumbens,\nhypothalamus and particularly POA,\nPe, PVN and ARC, habenular nuclei, thalamus,\nupper calculi of the midbrain, Raphe nuclei and\nthe pons ( 8 ).\nIn all vertebrate species studied from fish ( 22 )\nto humans, the RFRP/GnIH peptides decreased the\nsecretion of gonadotropins, particularly luteinizing\nhormone (LH), via actions on GnRH neurons and/\nor gonadotropes. This showed the possibility of a\nprotective role in various species ( 16 ,  23 ,  24 ). Recently\npublished reports indicated that these peptides\nin certain situations did not affect LH secretion\nor even have a stimulatory effect, which in the\nfollowing they also be explained.\nIn all vertebrate species studied from fish ( 22 ) to humans, the RFRP/GnIH peptides decreased the secretion of gonadotropins, particularly luteinizing hormone (LH), via actions on GnRH neurons and/ or gonadotropes. This showed the possibility of a protective role in various species ( 16 , 23 , 24 ). Recently published reports indicated that these peptides in certain situations did not affect LH secretion or even have a stimulatory effect, which in the following they also be explained.\nAny direct effect of RFRP on GnRH neurons necessitates a direct connection between RFRP neuronal terminals and GnRH neurons. In the POA of male rats, research has shown that RFRP fibers formed a close association with approximately 75% of GnRH neuronal bodies ( 24 ). Similar finding was reported in female hamsters (more than 40%), mice and rats ( 16 ). In another study, there was communication of RFRP fibers with GnRH neurons observed in the anterior hypothalamic area, MBH (approximately 30%) and POA of sheep ( 25 ). In sheep, co-expression of RFRP and GnRH during proestrus and estrus (follicular phase) and the luteal phase has been reported. During the luteal phase of sheep, more POA neurons expressed RFRP compared to the follicular stage, while there were no differences in the number of GnRH neurons in the hypothalamus, which indicated a direct effect of RFRP neurons in POA on GnRH neurons and an indirect effect on LH secretion ( 23 ).\nIn the POA of monkeys, 67.9% of GnRH neurons established connections with RFRP fibers ( 8 ), with similar connections observed in the human brain ( 7 ). More than 80% of GnRH neurons of POA in the Siberian hamster expressed GPR147 receptor ( 13 ). In adult male and female diestrus mice, there was a close relation between RFRP-3 neuron terminals with 25% of the body of GnRH neurons in the medial septum and 27% in the rostral part of the POA, and 33% of GnRH neurons which expressed  GPR147  mRNA ( 26 ). Addition of RFRP into GnRH neurons  in vitro  decreased the firing rate of 41% of the neurons. However, electrophysiologic evaluations in that study showed that RFRP treatment had a stimulatory effect on 12% of the neurons and no effect on 47% of neurons ( 27 ). In the same study, RFRP treatment caused hyperpolarization of more than 50% of GnRH neurons ( 28 ).\nIntraventricular administration of RFRP rapidly\ndecreased plasma LH concentration in male rats\n( 24 ), ovariectomized hamsters ( 16 ) and Siberian\nhamsters maintained on a long-day photoperiod;\nhowever, injection of RFRP in hamsters on a shortday\nphotoperiod stimulated LH release 30 minutes\nafter the injection ( 13 ). In another research, intraventricular\ninjection of RFRP-3 stimulated expression of c-Fos in GnRH neurons and increased both\nLH and testosterone secretion ( 29 ). In contrast,\nintraventricular injection of RFRP in ovariectomized\nrats had no effect on the mean plasma LH\nconcentration or frequency of LH pulses ( 30 ). Intraventricular\ninjection of RFRP in ovariectomized\nrats following induction of the GnRH/LH surge by\nestradiol (E 2 ) and progesterone decreased the activity\nof GnRH neurons (evaluated based on  c-Fos \ngene expression) by 50 to 60% ( 31 ). However, in\nthat study, central injection of RFRP in ovariectomized\nrats treated with E 2  implant had no effect on\nLH pulse and amplitude or mean concentration of\nLH. Recently, it was observed that intraventricular\nRFRP-3 injection in ovariectomized ewes had no\neffect on plasma LH concentration ( 32 ). Intraventricular\nadministration of RF9, a potent and specific\nantagonist of the RFRPs receptor ( 33 ), resulted\nin a rapid, dose-dependent increase in gonadotropin\nsecretion in male and female rats ( 34 ). Collectively,\nthese finding suggested that RFRP could\nchange GnRH secretion via a direct action on the\nGnRH neuronal system [for more information see\nthe review by Anderson ( 35 )].\nIn order to generate a physiologic effect on gonadotropin secretion, the hypothalamic RFRP neuronal terminals must either form a close association with GnRH neurons in the median eminence (ME) and/or RFRP receptors must be located on gonadotropes. The RFRP neuronal terminals are found in the external layer of the ME in hamsters ( 13 , 16 , 36 ), sheep ( 11 ), monkeys ( 8 ) and humans ( 7 ). GPR147 expression is reported in the hypophysis of hamsters ( 36 ), rats ( 37 , 38 ) and humans ( 7 ). The presence of RFRP in the hypothalamic-hypophyseal portal vein of sheep has been reported by Smith et al. ( 39 ). In rats, while some researchers did not observe RFRP fibers in ME ( 17 , 24 ), others reported the presence of RFRP fibers in male ( 40 ) and female ( 18 ) Sprague-Dawley rats.\nFluorogold is a retrograde tracer that does not cross the blood-brain barrier but can be absorbed from portal arterioles of hypophysis by neurons terminals in the external area of ME. Intraperitoneal injection of this tracer has been used to detect central hypophysiotropic cells. The results indicated that more than 90% of GnRH neurons and only 3 out of 234 RFRP neurons in the POA of rats stained with Fluorogold ( 17 ).\nIntravenous injection of RFRP decreased LH secretion in several mammals; however, the mode of action might differ in various species. Intravenous injection of RFRP in ovariectomized ewes decreased the amplitude of LH pulses; but had no effect on pulse frequency ( 11 ). Intravenous injection of RFRP in castrated bulls decreased the frequency of LH pulses, however a single injection had no effect ( 41 ). Intravenous injection of RFRP in ovariectomized rats ( 30 ) and ovariectomized hamsters decreased mean concentrations of LH ( 16 ).\nThe addition of RFRP to cultures of hypophyseal cells of rats ( 30 ), cows ( 41 ) and sheep decreased GnRH-induced LH secretion. Interestingly, the addition of GnRH to hypophyseal cells increased expression of  LHβ  mRNA in rams (4 times) and ewes (2.5 times), but RFRP inhibited LHβ subunit expression ( 42 ). On the other hand, it was also reported that treatment with RFRP ( 31 ) or RF9 (an RFRP receptor antagonist) ( 34 ) had no effect on GnRH-induced LH secretion in a hypophyseal cell culture in rats.\nIn addition to expression in the brain, expression of RFRPs and their receptors in mammalian gonads have been reported. In male hamsters, cells that expressed RFRP were observed in the seminiferous tubules. Its receptor, GPR147, was observed in spermatocytes and spermatids ( 43 ). In the monkey, RFRP and its receptor were expressed in Leydig cells, spermatogonia and spermatocytes ( 44 ). RFRP was also found in the granulosa and luteal cells of mice ovaries ( 45 ).\nRFRP peptides in granulosa cells of preovulatoy follicles and corpus luteum along with GPR147 receptors in granulosa cells, theca cells and the corpus luteum have been observed in women. RFRP-3 could inhibit the effect of gonadotropins on progesterone production and expression of StAR protein ( 46 ). Thus, it was postulated that RFRP might have autocrine/paracrine roles in gametogenesis and steroidogenesis ( 44 ).\nA low concentration of E 2  secreted during the\nmajority of the ovarian cycle in most mammals exerts a negative feedback effect on GnRH neurons\nby keeping GnRH/LH secretion at a basal level.\nDuring the preovulatory period, high levels of estrogen\nsecreted from mature follicles results in a\nGnRH/LH surge via a positive feedback effect.\nThe GnRH neurons do not express alpha E 2  receptors\n(ER α ) which are essential for the positive and\nnegative feedback effects of E 2  ( 47 ). Therefore, it\nseems that other steroid sensitive neurons are intermediaries\nof the estrogen effect on regulation of\nGnRH (and LH) secretion.\nApproximately 40% of RFRP neurons in the\nbrain of female hamsters ( 16 ) and 18% in ovariectomized\nmice expressed ER α  ( 48 ). Therefore,\nRFRP neurons might intermediate the E 2  feedback\neffect. We studied the expression of  RFRP  mRNA\nand peptides during the estrous cycle of rats. Expression\nof  RFRP  mRNA in proestrus was less\nthan in diestrus and the numbers of neurons that\nexpressed the RFRP peptides during proestrus and\nearly estrus was less than during estrus and diestrus.\nIncreased secretion of E 2  in the evening of\nproestrus from dominant follicles in addition to the\npositive feedback effect on GnRH/LH surge might\nfacilitate GnRH/LH secretion by exerting an inhibitory\neffect on RFRP expression in DMH ( 18 ).\nIn another study we evaluated the numbers of neurons\nthat expressed RFRP in DMH/PVN during\nthe follicular and luteal phases of goats. The numbers\nof positive cells in the follicular phase (periovulatory\nperiod) was less than in the luteal phase\n( 21 ). Consistent with this finding, the numbers of\nRFRP neurons decreased during the preovulatory\nperiod in hamsters ( 36 ). It was also reported that\nthe number of POA neurons that expressed RFRP\nwas greater during the luteal phase compared with\nthe follicular phase in sheep ( 19 ).\nE 2  implants (100 μg/ml) for 4 days in ovariectomized\nmice decreased the number of RFRP cells\nand the expression of  RFRP  mRNA per cell ( 48 )\nas determined by in situ hybridization. Possibly,\nonly high or long term levels of E 2  could decrease\nthe expression of RFRP in rodents because only\nonce subcutaneous injection of E 2  in ovariectomized\nhamsters sufficiently increased the activity\nof RFRP neurons (evaluated by c-Fos expression)\nat 3 and 6 hours after injection ( 16 ). In contrast to\nthese findings, during the breeding and non-breeding\nseasons for sheep, there was no difference\nin the numbers of neurons that expressed  RFRP \nmRNA and  RFRP  mRNA levels per cell between\novariectomized ewes and ovariectomized ewes\nthat received E 2  implants ( 25 ).\nRFRP neurons in the brain of male hamsters\nexpressed an androgen receptor ( 16 ), however\ncastration of male hamsters or treatment with testosterone\nimplants for 4 weeks had no significant\neffect on the number of cells that expressed RFRP\n( 49 ). Therefore, more studies should be conducted\nto clarify the mechanism of sex steroid action on\nRFRP neurons.\nReproductive activity in several mammalian species shows salient seasonal alterations due to basal alterations in secretion of reproductive hormones. In compliance with the action of RFRP mammalian reproduction, it is logical that RFRP expression in seasonal breeders will be harmonized with changes in the photoperiod. Contrary to the expectation in Syrian and/or Siberian hamsters (long-day breeders), there were fewer neurons that expressed RFRP mRNA and RFRP peptide during the short-term photoperiods (8 hours light) compared with the long-term (16 hours light) photoperiods ( 13 , 49 , 50 ). These findings were not related to the specific time of day since gene expression was the same during 24 hours ( 49 ).\nAggregation of RFRP fibers in POA and rostral hypothalamus (aggregation area of GnRH neurons) and the percent of GnRH neurons that established connections with RFRP fibers were less during the shortterm compared to the long-term photoperiod ( 13 , 50 ). On the other hand, pinealectomy prevented a decrease in RFRP expression during short days ( 13 , 49 ). A 60day melatonin injection administered to hamsters maintained under long-term photoperiods remarkably decreased  RFRP  mRNA expression and produced the same response as in the short-term photoperiod ( 49 ). Administration of melatonin for 13 weeks to pinealectomized hamsters kept under a short-term photoperiod decreased  RFRP  gene expression ( 13 ). Therefore, melatonin appeared to decrease the activity of RFRP neurons during short-term photoperiods.\nCoordination of these findings with the inhibitory role of RFRP was difficult because when the lowest level of expression was seen, the reproductive system was inactive. In Siberian hamsters, the relative expression of  RFRP  mRNA during average days (13.5 hours light) was more than 40 times the long days (16 hours light) ( 51 ). Therefore, it was possible that a considerable increase in RFRP expression during the early period of reproductive system regression (average days) would be necessary to inhibit the reproductive axis. However, this level of expression in hamsters whose reproductive axis did not completely regress was not necessary. Because intraventricular injections of RFRP in hamsters maintained under short-term photoperiod conditions had a stimulatory effect ( 13 ) it was possible that the decrease in RFRP expression during short-term photoperiods was important for inhibition of reproduction.\nUnlike hamsters, sheep and goats are shortday breeders. The number of RFRP that expressed neurons during the non-breeding season in sheep (long-term photoperiod) was approximately 40% more than during the breeding season, but there was no difference in the number of  RFRP  mRNA per cell ( 25 ). However, in another study,  RFRP  mRNA expression was highest during the long days ( 10 ). Communication of RFRP fibers with GnRH neurons in POA and rostral hypothalamus was highest during the non-breeding season ( 25 ). In addition to the seasonal change of RFRP expression in DMH/PVN nuclei,  RFRP  mRNA expression in epithelial or ependymal cells around the ventricle was seen only in long days ( 10 ).\nRecently we evaluated RFRP expression in DMH/PVN nuclei in goats during the breeding season (follicular and luteal phases) and anestrus. In both nuclei, the number of cells that expressed RFRP was higher in during anestrus compared to the follicular phase. However, there was no difference between the anestrus and luteal phases. We also determined the number of positive neurons in the rostral, middle and caudal parts of the DMH/PVN. In the rostral areas, more RFRP neurons were observed during anestrus than during the follicular phase; however, there was no effect of the reproductive stage recorded in middle and caudal parts of these nuclei ( 21 ). These results in sheep and goats were in accordance with the inhibitory role of RFRP on the reproductive axis.\nThe diffuse distribution of RFRP neuronal processes in the brain is suggestive of additional roles for this neuropeptide in physiology. The RFRP neuronal terminals in sheep brain are extended to neurons of orexin, melanin, proopiomelanocortin and neuropeptide Y; therefore, RFRP neurons may have a role in the regulation of appetite and energy balance, and possibly function as a link between nutrition and reproduction ( 52 ). Long term malnutrition (2 weeks) has been shown to increase  RFRP-3  mRNA expression in DMH of the hypothalamus in ovariectomized female rats ( 53 ).\nFurthermore, in the monkey brain, RFRP fibers had a close relation with neurons of dopamine, beta-endorphin and GnRH-II. Since dopamine neurons express GPR147, it was suggested that RFRP might stimulate prolactin secretion by inhibition of dopamine neurons ( 8 ). Consistent with this idea, we showed that the numbers neurons that expressed RFRP in suckling rats (in which plasma prolactin is at its highest level) was higher than in non-suckling rats ( 54 ). Increased  RFRP-3  mRNA expressions in DMH of the hypothalamus while increasing milk production in rats might be the inhibitory factor for GnRH secretion ( 55 ).\nBased on the findings that prolactin ( 56 , 57 ) and oxytocin ( 58 ) secretion increased during the refractory period after ejaculation in men, we proposed a hypothesis that increased RFRP expression after ejaculation might be the cause of the post-ejaculation refractory period in men ( 59 ). Intracerebroventricular injection of RFRP in rats also increased the activity of oxytocin neurons in the hypothalamus and oxytocin concentrations in plasma. It was shown that the supraoptic and PVN nuclei of the hypothalamus expressed  GPR147  mRNA ( 60 ). Therefore, RFRP peptides might also participate in the regulation of oxytocin secretion.\nCoexpression of RFRP and AgRP in the Arc neurons of the ewe has been reported which indicated a probable role of these two peptides in control of the ewe reproductive cycle. This study also showed that ovarian steroids affected expression of these peptides in the Arc of the hypothalamus and might be a link between energy homeostasis and reproduction ( 23 ).\nAs mentioned before, RFRP peptides have an opposite effect against GnRH in numerous situations and inhibit secretion of gonadotropins. However in some cases they may have an effect on LH release and a stimulatory effect (please see the previous sections). GnRH analogs (agonists and antagonists) have been applied in the treatment of a wide spectrum of reproductive disorders, including precocious puberty, endometriosis, uterine fibroids, prostatic hyperplasia, prostatic and breast cancers. By 2000, more than 2 billion dollars in sales of these compounds was recorded ( 61 ). Therefore, considering the potential effect of RFRP in inhibition of gonadotropins, the use of these peptides in the future for the treatment of reproductive disorders would be expected ( 62 ).\nThe inhibitory effect of stress on reproductive performance has been demonstrated. Stress leads to activation of the hypothalamus-pituitary-adrenal axis which inhibits GnRH secretion. It seems that the effects of stress on the hypothalamuspituitary-gonad axis is mediated by adrenal steroid hormones (glucocorticoids). Since neurons of GnRH do not express glucocorticoid receptors, it is possible that these steroid hormones affect neurons upstream of GnRH neurons and change the release of GnRH. Reports have shown that RFRP neurons mediate the effects of stress on reduction of GnRH/LH secretion and stop of the reproductive axis ( 38 , 63 ). Therefore, it is possible that using RFRP antagonists or antibodies against RFRP safeguard reproductive performance in stressful situations. Also, as noted above, increase in RFRP expression may be involved in the post-ejaculatory refractory period ( 59 ) Hence disabling the RFRP system may shorten this period.\n\nBased on the findings in mammals, RFRPs are homologues of GnIH in birds and can inhibit LH secretion and the reproductive axis; however, their mode of action is not yet clearly established. For example it is not known whether they inhibit the GnRH system and/or have a direct effect on hypophyseal gonadotropes in preventing gonadotropin secretion. Amongst studied species, the most contradictory data have been reported in the rat. Based on extensive connection between the RFRP neurons and other neurons, more studies will be required to identify the exact role of these peptides in reproduction and other physiologic functions.","source_license":"CC-BY-4.0","license_restricted":false}