Regulation of Expression of Mammalian Gonadotrophin-Releasing Hormone Receptor Genes

Authors


Dr J. P. Hapgood, Department of Biochemistry, University of Stellenbosch, Private Bag X1, Matieland, 7602, South Africa (e-mail: jhap@sun.ac.za).

Abstract

Gonadotrophin-releasing hormone (GnRH), acting via its cognate GnRH receptor (GnRHR), is the primary regulator of mammalian reproductive function, and hence GnRH analogues are extensively used in the treatment of hormone-dependent diseases, as well as for assisted reproductive techniques. In addition to its established endocrine role in gonadotrophin regulation in the pituitary, evidence is rapidly accumulating to support the expression and functional roles for two forms of GnRHR (GnRHR I and GnRHR II) in multiple and diverse extra-pituitary mammalian tissues and cells. These findings, together with findings indicating that mutations of the GnRHR are linked to the disease hypogonadotrophic hypogonadism and that GnRHRs play a direct role in neuronal migration and reproductive cancers, have presented new therapeutic targets and intensified research into the structure, function and mechanisms of regulation of expression of GnRHR genes. The present review focuses on the current knowledge on tissue-specific and hormonal regulation of transcription of mammalian GnRH receptor genes. Emerging insights, such as the discovery of diverse regulatory mechanisms in pituitary and extra-pituitary cell types, nonclassical mechanisms of steroid regulation, the use of composite elements for cell-specific expression, the increasing profile of hormones involved in regulation, the complexity of kinase pathways that target the GnRHR I gene, as well as species-differences, are highlighted. Although further research is necessary to understand the mechanisms of regulation of expression of GnRHR I and GnRHR II genes, the GnRHR is emerging as a potential target gene for facilitating cross-talk between neuroendocrine, immune and stress-response systems in multiple tissues via autocrine, paracrine and endocrine signalling.

Gonadotrophin-releasing hormone (GnRH), in conjunction with the GnRH receptor (GnRHR) is the primary regulator of reproduction in vertebrates. It is well-established that GnRH is released from the hypothalamus in a pulsatile fashion, and travels to the pituitary gland via the portal hypophyseal vasculature (1). Upon GnRH binding to its G protein-coupled receptor (GPCR) on the plasma membranes of pituitary gonadotroph cells, a range of intracellular signalling pathways are activated that ultimately regulate the synthesis and secretion of the gonadotrophins, luteinising hormone (LH) and follicle-stimulating hormone (FSH). In turn, gonadotrophins stimulate sex hormone synthesis and gametogenesis in the gonads to ensure reproductive competence. At least two forms of the decapeptide hormone (i.e. GnRH I and GnRH II), as well as the receptor (i.e. GnRHR I and GnRHR II), have been found in most vertebrates, including mammals, increasing the potential for diverse physiological actions (2–6). GnRH II was originally isolated from chicken brain and its precise role in mammals remains to be elucidated (7). However, both GnRH peptides can bind to and activate both receptor forms, with GnRH I exhibiting a greater affinity and potency for GnRHR I and GnRH II exhibiting a greater affinity and potency for GnRHR II (8). GnRH I released from the hypothalamus is the hormone that appears to be sufficient for gonadotrophin regulation in the mammalian pituitary (4).

Besides the well-established role for GnRH I and GnRHR I in gonadotrophin regulation in the pituitary, the detection of both forms of the hormone and receptor (Table 1) (4) in multiple mammalian nonpituitary tissues and cells suggests numerous and diverse autocrine, paracrine and endocrine extra-pituitary roles for GnRHs and GnRHRs (2–5, 9). These include neuronal migration during development (10), neuromodulation in the brain to affect sexual behaviour (4), possible modulation of visual processing in the eye (11), digestive tract function (12), inhibition of gastric acid secretion (13), adhesion chemotaxis and homing in T cells (14), human chorionic gonadotrophin (hCG) release in the placenta (15), steroidogenesis in ovarian cells (16–19), proliferation in melanoma cells (20), sperm function and sperm–oocyte interactions (21) and growth inhibition in reproductive tumours (22–26). However, the specific roles of each form of the hormone and receptor in these various tissues and cells remain to be elucidated. Recently, there has been an upsurge in the available literature on GnRHRs. GnRH and its analogues are extensively used in the treatment of hormone-dependent diseases, as well as for assisted reproductive techniques (6). More recently, they have been proposed as novel contraceptives in men and women (6, 27). The finding that naturally occurring mutations of the GnRHR are linked to the disease hypogonadotrophic hypogonadism, which results in delayed puberty, has also recently stimulated interest in GnRHR function (28). Thus, the presence of multiple forms of GnRH and its receptor in mammals, as well as the emerging multiple roles thereof, have presented new therapeutic targets and intensified the search for novel interventive GnRH analogues.

Table 1.  Summary of Mammalian Tissues and Cell Lines Where GnRHR I mRNA and/or Protein have been Detected.
TissueSpeciesmRNAProteinReference
  1. +, Expression in specific tissues. –, expression investigated and found not to occur. Numbers refer to detection method: (1) = in situ hybridisation ; (2) = cDNA cloned ; (3) = RT-PCR ; (4) = Northern Blot; (5) = Xenopus oocyte expression; (6) = ligand binding assay; (7) = hormone response; (8) = Western Blot; (9) = Immunohistochemistry or immunocytochemistry.

Normal tissues, primary cells
 Pituitary
 Mouse+ (1)+ (5)(173)
Rat+ (2)+ (6)(156, 174)
Guinea pig+ (2)+ (7)(175)
Bovine+ (2)+ (6)(176, 177)
Sheep+ (2)+ (6)(133)
Wallaby+ (2) (78)
Possum+ (2)+ (6, 7)(178)
Human+ (2)+ (6)(85)
Bonnet monkey+ (3)+ (8)(179)
Marmoset monkey+ (1, 3)+ (7)(180, 181)
Macaque monkey+ (1) (180)
Pig+ (2) (182)
  GH, TSH cellsHuman + (9)(183)
 Hypothalamus
 Rat+ (3)+ (6)(185, 186)
Bovine− (3) (176)
Wallaby– (3) (187)
 Brain
 Sheep– (3) (133)
Human+ (3) (38)
Wallaby– (3) (187)
  Placode-derivedRat+ (3)+ (7)(50)
  Pre-optic areaRat+ (3)+ (6)(185, 188)
  HippocampusBovine– (3) (176)
  Enteric neuronesRat+ (3) (189)
 Placenta
 Human+ (1, 3)+ (6)(190, 191)
  IEVTHuman+ (3)+ (7)(163)
  Primary trophoblastsHuman+ (3)+ (7)(163)
 EndometriumHuman+ (3) (15)
Human+ (3)+ (6)(192)
Human+ (3) (193)
 Myometrium
 Human+ (3)+ (6, 7)(194, 195)
Bovine– (3) (176)
Human+ (3)+ (7)(194)
  DeciduaHuman + (7)(196)
 Ovary
 Wallaby+ (3) (78)
Bovine– (3) (176)
Human+ (3, 4)+ (7)(85, 197)
  Granulosa-lutealHuman+ (3)+ (7)(38, 197)
  Surface epitheliumHuman+ (3)+ (7)(159, 198)
Rat+ (3)+ (6)(199, 200)
  InterstitialRat+ (1) (79)
  GranulosaRat+ (1) (79)
  Corpus luteumRat+ (1)+ (6, 7)(79, 201)
Human + (6)(202)
  ThecaRat– (1) (79)
  Luteal cellsBovine– (3) (176)
 Breast/mammary gland
 Rat+ (3)+ (6)(203)
Mouse– (3) (81)
Human+ (3, 4)– (6)(85, 204)
 Testis
 Wallaby+ (3) (78)
Human+ (3, 4) (85)
Rat+ (3)+ (6)(199, 205)
Bovine– (3) (176)
  InterstitialRat + (6)(205)
  LeydigRat + (6)(206)
  SpermHuman + (9)(207)
 Prostate
 Human+ (3, 4) (85)
Rat+ (3) (208)
 Immune system
  LymphocytesMouse+ (3) (87)
  T cellsHuman + (7)(14)
  Mononuclear blood cellsHuman+ (3)+ (7)(88)
  SpleenHuman– (3) (85)
Bovine– (3) (176)
 LiverHuman+ (3) (209)
Wallaby– (3) (187)
Bovine– (3) (176)
Sheep– (4) (133)
 PancreasHamster – (6)(210)
Human – (6)(211)
 KidneyHuman+ (3)+ (6)(209, 212)
Bovine– (3) (176)
Sheep– (4) (133)
 AdrenalBovine– (3) (176)
Sheep– (4) (133)
 HeartHuman+ (3) (209)
Sheep– (4) (133)
 Skeletal muscleHuman+ (3) (209)
Human– (3) (38)
Wallaby– (3) (187)
Rat – (8)(213)
 Submaxillary glandsRat+ (1, 3)+ (9)(214)
 Digestive tractRat+ (1)+ (9)(12)
  Gastric parietal cellsRat+ (1)+ (9)(13)
 Spinal chordSheep+ (3) (86)
 RetinaVole+ (3) (11)
Cancer tissues, cancer cell lines
 Pituitary
  Pituitary adenomaHuman+ (1, 3) (215)
  αT3-1Mouse+ (4)+ (6)(184)
  LβT2Mouse+ (4)+ (7)(96)
 Hypothalamus
  GT1-7Mouse+ (4)+ (6)(39)
  FNC-B4Human+ (3)+ (7)(10)
 Liver
  HepG2 (hepatocarcinoma)Human+ (3)+ (7)(89, 216)
 PancreasHamster + (9)(90)
Human + (6)(211)
 Skin
  BLM, Me15392 (melanoma)Human+ (3)+ (6–8)(217)
 Placenta
  JEG-3 (choriocarcinoma)Human+ (3)– (7)(218, 219)
  JEG-3Human+ (3)+ (7)(163)
 Endometrium
 Human + (6, 7)(220)
Human+ (3)– (6)(219)
Human+ (3)+ (6)(195)
  HEC-1AHuman+ (3)+ (7)(221)
Human+ (3)– (7)(219)
  IshikawaHuman+ (3)– (7)(219)
 Uterus
  LeiomyosarcomaHuman+ (3)+ (6)(222)
  MyomaHuman+ (3)+ (6)(195)
  CervicalHuman– (3)– (6)(195)
 Ovary
 Human+ (3)+ (7)(223, 224)
Human+ (3)+ (6)(225)
  EpithelialHuman+ (3)+ (6)(195)
  StromalHuman+ (3)+ (6)(195)
  Germ-cell derivedHuman– (3)– (6)(195)
Rat + (6)(200)
  OVCAR-3Human+ (3)+ (7)(159)
  Caov-3Human+ (3)+ (7)(226)
  SK-OV3Human + (7)(227)
  EFO-21, EFO-27Human + (6, 7)(228)
  SVOG-4o, SVOG-4 mHuman+ (3)+ (8)(111)
 Breast
 Human+ (3)+ (6)(204, 223)
  MCF-7Human+ (3)+ (6, 7)
(6)
(229)
(23)
  MDA-MB-157, MDA-MB-231Human + (6, 7)(229)
  ZR-75–1Human + (6, 7)(229)
  Sk Br 3Human + (6, 7)(229)
  MXTMouse + (9)(90)
 Prostate
 Human+ (3)+ (6)(223, 230)
Rat+ (3)+ (6)(208, 231)
  Dunning R3327Rat+ (3) (208)
  DU145Human+ (1) (232)
  LNCaPHuman+ (1)+ (8)(232, 233)
  PC-3Human+ (3)+ (6, 7)(90, 223)
  TSU-Pr1Human+ (3)+ (7)(234)

A central issue in the field is to understand the extracellular signals and the intracellular mechanisms that regulate expression of GnRHRs in these diverse tissues and cells. Responsiveness to GnRH depends on the number of GnRHRs on the cell surface. In turn, GnRH appears to be an important regulator of receptor levels on the gonadotroph cell surface (29). Several lines of evidence indicate that the number of GnRHRs is partially dependent upon the level of GnRHR mRNA, which appears to be regulated at least in part at the transcriptional level in gonadotrophs, including by GnRH itself (29–33). New insights have recently been obtained about other hormones [melatonin (34), adrenal and sex steroids (35–37), activin (20), hCG (38, 39)], as well as the intracellular signalling pathways (40–44) and transcription factors involved (45–49) in regulating mammalian GnRHR transcription in diverse tissues (10, 50, 51) from different species (35, 40, 41, 45, 52). Knowledge of these mechanisms is important to fully understand both the physiological and the therapeutic actions of GnRH and GnRHRs, given the central role of GnRHRs in reproductive endocrinology and the widespread use of GnRH analogues in endocrine and anticancer therapy. The present review focuses on transcription of mammalian GnRHR genes, including the hormones that are involved, the signalling pathways that are activated and the promoter elements and transcription factors that mediate responses to multiple signals in various tissues.

Mammalian GnRHR genes

The structures of the mouse (53), rat (54), human (55), pig (56) and sheep (57) GnRHR I genes have been characterised. In these species, the genes exist as a single copy, and have a high degree of sequence homology in the coding regions. They are structurally similar and consist of three exons separated by two introns. The exon–intron boundaries are conserved between the species, but the genes differ with regard to the size of the introns, as well as the sequence and length of the 5′-and 3′ untranslated regions (UTR) (Fig. 1). Exon 1 encodes the N-terminal tail as well as transmembrane helices (TM) 1, 2, 3 and part of TM 4 of the GnRHR I protein. Exon 2 encodes the rest of TM 4 and the whole of TM 5 whereas exon 3 encodes TM 6 and 7 (58).

Figure 1.

Structural organisation of the the gonadotrophin-releasing hormone receptor I gene in human, mouse, rat and sheep. Exons are represented by blocks, with portions of exons containing coding sequences shown as dark areas, and untranslated regions (UTR) shown as light areas. Sizes of exons are shown as the sum of the sizes of the coding and noncoding portions of each exon. The size of the sheep 3′ UTR has not been established. Introns are represented by solid lines, with sizes as indicated. All sizes are indicated in kilobasepairs. For the human gene, the size of the 5′ UTR is given relative to the most-3′ transcription start site as identified by Kakar et al. (98) for human pituitary tissue, and the size of the 3′ UTR is as established by Fan et al. (55) for human brain tissue. For the mouse gene, the 5′ UTR is relative to the major transcription start site as identified by Albarracin et al. (97) in αT3-1 cells and by Sadie et al. (49) in mouse pituitary tissue. For the rat gene, the size of the 5′ UTR is given relative to the major transcription start site as identified by Reinhart et al. (54) for rat pituitary tissue. For the human gene, the correlation of coding regions with protein structure is indicated, as adapted from (58), and is identical for the other species shown. The figure was adapated from (53–55, 57, 58, 98, 99, 172). Note that the figure is not drawn to scale.

Candidate genes for GnRHR II can be found at two different loci in the human genome (2, 4). The first is located on chromosome 1 and overlaps in the antisense orientation with the gene encoding the RNA-binding motif protein-8A (RBM-8A) (59). The GnRHR II gene has the same exon–intron structure as GnRHR I, except that exon 3 includes a cytoplasmic C-terminal tail, which is absent in GnRHR I. A premature stop codon (UAA) is located inframe within exon 2 in the human gene, suggesting that the gene products are nonfunctional. A second human locus containing a pseudogene for GnRHR II and RBM-8A is on chromosome 14. GnRHR II genes have also been detected in other mammalian genomes. The premature stop codon found in the human gene is conserved in the chimpanzee GnRHR II gene (7), whereas a fully functional GnRHR II gene is present in other primates, such as the marmoset monkey (8), rhesus monkey and African green monkey (60). GnRHR II genomic sequences for some nonprimate mammals have also been identified, and potentially encode functional proteins in pigs and dogs, but not in sheep and cows. The gene is completely deleted from the mouse genome (4), but a gene remnant is present in the rat genome (5).

Expression and physiological roles of mammalian GnRHRs

Both GnRHR I (Table 1) and GnRHR II (4) transcripts and, in some cases, GnRHR I protein (Table 1) have been detected in pituitary as well as extra-pituitary tissues and cell lines from several mammalian species. However, expression of endogenous GnRHR II protein has not been conclusively shown in any mammalian species. One study shows immunodetection of putative GnRHR II extracellular loop 3 in mammalian pituitary tissues, however, positive signals were also detected in mouse pituitary tissue (8), raising doubts about the specificity of the antibody. Expression of a full-length human GnRHR II protein would appear unlikely because, apart from the premature stop codon within exon 2, the human transcript lacks a methionine translation initiation codon. Several as yet unsubstantiated theories have been postulated as to how such a transcript could result in the synthesis of a functional receptor protein (7, 61). However, a truncated form of the human GnRHR II may play a modulatory role in GnRHR I expression by perturbing normal processing of GnRHR I (62). Although recent indirect evidence suggests that a functional human GnRHR II protein may be expressed in cancer cells (63), no definitive functional role has been established to date for any mammalian GnRHR II (3).

Pituitary

It is well established that GnRHR I protein is expressed in mammalian pituitary gonadotrophs, where its primary role is to regulate LH and FSH synthesis and release (1). GnRHR I expression in the pituitary is regulated during foetal development, during sexual maturation (showing differential patterns between sexes) and during the reproductive cycle and pregnancy in the adult (51, 64, 65). In adult mammals, activation of the GnRHR I results in stimulation of diverse intracellular signalling pathways in the anterior pituitary, the nature of which depends on the cellular context (the relative concentrations of receptors and G-proteins vary during the oestrous cycle) (66, 67). In mouse gonadotroph cell lines, GnRHR I has been shown to couple to Gq/G11 in αT3-1 cells and to Gq/G11 and Gs in LβT2 cells (42). In rat primary pituitary cell cultures, the rat GnRHR I can couple to Gs, Gi/Go and Gq/G11, to modulate the activity of both the protein kinase A (PKA) and protein kinase C (PKC) pathways (41). Several recent reports have further unravelled the downstream kinase pathways in gonadotroph cell lines showing that GnRH can activate ERK, JNK and p38 MAPK in both αT3-1 and LβT2 cell lines, in various ways including via both PKA- and PKC-dependent and -independent pathways (41, 68). These pathways then differentially regulate synthesis and release of the gonadotrophins, via mechanisms that are not well defined. A feedback mechanism whereby PKC regulates the affinity of the GnRHR for GnRH has been reported, which suggests a novel form of ‘inside-out’ signalling (44). Furthermore, it has recently been shown in αT3-1 cells that the appropriate organisation of the GnRHR I into low-density membrane microdomains on the cell surface appears critical in mediating GnRH I-induced intracellular signalling (43). The important role played by the GnRHR in gonadotrophin regulation is illustrated by the findings that several naturally occurring mutations in the human GnRHR I result in hypogonadotrophic hypogonadism, with symptoms of delayed sexual development, low or apulsatile gonadotrophin and sex steroid hormone levels, in the absence of abnormalities in the hypothalamic-pituitary axis (6, 69). The majority of these mutated receptors are mislocalised proteins, exhibiting altered membrane trafficking (70) and endoplasmic reticulum retention that can be restored to function by pharmacological chaperones (71).

Multiple GnRHR I mRNA transcripts coding for full-length protein are detected in normal mammalian pituitary tissue and cell lines. In addition, several species, including humans, express splice variants that may code for functionally relevant truncated GnRHR I proteins (9, 72). In the gonadotrophs, ligand-mediated GnRHR I activation also leads to an increase in the expression and enzyme activity of nitric oxide synthase (NOS) I (73), the enzyme responsible for producing the signalling molecule nitric oxide (NO). In particular, this up-regulation occurs during pro-oestrus (74), and a role for NO in gonadotrophin release, fertility (75) and mating behaviour (76) has been suggested although the link between GnRHR I activation and NOS remains unclear. In some mammals (4), functional GnRHR II mRNA has been detected in pituitaries. Because stimulation with GnRH II has been shown to result in preferential FSH release in sheep (8), it is tempting to speculate that both GnRHR II and GnRH II may be involved in regulation of gonadotrophin synthesis and release in the pituitary gonadotrophs. GnRHR I immunoreactivity has also been detected in human thyrotrophs and somatotrophs, suggesting additional roles for GnRHR I in the pituitary other than gonadotrophin regulation in gonadotrophs (9).

Extra-pituitary

Reproductive tissues

In female reproductive tissues, paracrine/autocrine actions of GnRH via GnRHR I play a role in normal breast (77) and ovarian (78, 79) development, regulation of the menstrual cycle, early establishment and maintenance of pregnancy (80) and in lactation (81, 82). GnRHR I mRNA and/or protein has been detected in normal human breast tissue and several ovarian compartments, in endometrial tissue and in placental trophoblasts, cytoblasts and syncytiotrophoblasts (Table 1) (9). Endometrial GnRH may play a paracrine/autocrine role in the early stages of implantation by modulating placental hCG secretion (15), which is involved in establishment and maintenance of pregnancy. The ovarian levels of GnRHR I and GnRH I mRNA vary during the oestrous cycle in the rat, where they are thought to play a local role in preparing the ovary for ovulation (83). Recent findings have also revealed additional novel functions for GnRHRs in the human ovary, including inhibition of gonadotrophin-regulated steroidogenesis and suppression of hCG-stimulated progesterone production in granulosa-luteal cells (16–19).

In male reproductive tissue, paracrine/autocrine actions of GnRH I via GnRHR I play a role in both testis and sperm development (78), as well as sperm motility and sperm–oocyte interactions (21). Although GnRHR II transcripts have been detected in mature human sperm and postmeiotic testicular cells, these appear not to be functional (61). GnRH or GnRH-like peptides produced in the human testis and prostate, and detected in human seminal plasma (84), may all be part of complex autocrine and paracrine regulatory circuits, because the GnRHR I is expressed in the human testis and prostate (85).

Non-reproductive tissues

Hypothalamic GnRH neurones have been found to express GnRHR I, which is proposed to function in an autocrine fashion to regulate GnRH release (50). Furthermore, recent findings in rat primary GnRH neuronal cells suggest that the GnRH-activated Ca2+ signalling and autocrine regulation of GnRH release could provide a mechanism for regulated GnRH I secretion during embryonic neuronal migration (50). Support for a role for GnRHRs in neuronal differentiation and migration also comes from work on human foetal olfactory epithelial cells (10). In addition, the detection of both GnRHR I (Table 1) and GnRHR II (4) transcripts in many mammalian brain tissues has supported a role for GnRH I and/or GnRH II as a neurotransmitter or neuromodulator. This hypothesis is supported by the expression of both GnRH I and GnRHR I in the spinal cord of sheep (86), as well as functional evidence in vivo for an integral role for GnRH in sexual behaviour in mammals via several different brain tissues (4). In addition, the detection of GnRH and GnRHR I transcripts and/or protein in T-cells (14), spleen (87) and gastric parietal cells (13), combined with functional evidence, suggests other autocrine/paracrine roles for GnRHRs in immunomodulation (87, 88), such as adhesion chemotaxis and homing in T cells (14), and inhibition of gastric acid secretion (13).

Cancer cells

It is widely accepted that continuous administration of GnRH analogues inhibits growth of several reproductive tissue-derived tumours and that this effect may be mediated via GnRHRs expressed on these cells (20, 22–26). However, the antiproliferative effects of GnRH analogues on human melanoma cells (20) suggest that such GnRHR-mediated growth effects are not unique to reproductive tissue-derived cancer cells. The GnRHR-mediated intracellular pathways involved may include nuclear GnRH binding sites (89, 90) and/or interaction and interference with epidermal growth factor receptor mitogenic signalling (91, 92). Some investigators have recently provided indirect evidence to support a functional role for a putative human GnRHR II in mediating the antiproliferative effects of GnRH analogues in human endometrial, leukaemic and prostate cancer cells (63).

Regulation of GnRHR gene transcription in different mammalian tissues and cell lines

Regulation of expression of GnRHR numbers has been shown to occur at the transcriptional, translational and post-translational level. A well-known mechanism for ligand-mediated post-translational down-regulation of GPCR numbers on the cell surface involves desensitisation, internalisation and degradation. Whereas type I mammalian GnRHRs have been shown to internalise slowly due to the lack of a C-terminal tail (40), a recent study showed that the marmoset monkey GnRHR II, which has a C-terminal cytoplasmic domain, internalises rapidly (93). Homologous regulation of translation efficiency from GnRHR mRNA has also been shown to occur in αT3-1 cells (94). However, very little research has been reported on post-transcriptional regulation of GnRHR I gene expression. The present review will thus focus on transcriptional mechanisms.

Work carried out in animals, in primary cells, as well as in several model cell lines, has contributed towards an emerging understanding of the complex transcriptional regulatory pathways by which mammals regulate transcription of GnRHR I genes. Much of the detailed molecular mechanisms of gene regulation of the mouse, rat and human GnRHR I genes have been investigated in mouse pituitary cell lines. The αT3-1 cell line is a precursor gonadotroph cell line that retains several differential functions of gonadotrophs, such as gonadotrophin α-subunit expression, synthesis and secretion, as well as expression of GnRHR I and receptor-dependent responsiveness to GnRH I. However, these cells differ from mature primary gonadotrophs in that they do not express or secrete the gonadotrophin beta-subunits LHβ and FSHβ (95). The LβT2 mouse pituitary cell line is more differentiated, exhibiting more pronounced gonadotroph-like characteristics, such as expression and secretion of the gonadotrophin α-subunit and both gonadotrophin-specific β-subunits (96). Extensive characterisation of the human GnRHR I gene promoter has been performed in human reproductive tissue-derived cell lines.

Promoter characterisation, basal and cell-specific expression (Fig. 2)

Figure 2.

Functional elements in the gonadotrophin-releasing hormone receptor I promoter regions of human, mouse and rat. Shaded boxes and striped boxes represent TATA and CCAAT elements, respectively. Black boxes represent elements that have been functionally characterised. White boxes represent putative elements that have been identified through promoter sequence analysis. Transcription start sites are indicated with arrows, and the translation start site with ‘ATG’. For the human gene, the most-3′ transcription start site, as identified by Kakar et al. (98) for human pituitary tissue, is indicated. Other transcription start sites, as identified for human brain (55), pituitary (98) and placental tissues (109), are not indicated. Hormone responses and their corresponding cis-elements established in functional studies in pituitary cell lines are indicated. The mouse promoter has not been functionally characterised upstream of the GRAS element. This figure was adapted from (9) and (72), and other references quoted in the text. *GRAS contains binding sites for SMAD, AP-1 and FoxL2 proteins (45, 47). #Several functional elements overlap in this region. The positions for DARE and SURG-1 are indicated. LHX2 was specifically shown to bind DARE (154), but LHX3 was shown to bind an overlapping site (104). SF-1, Steroidogenic factor-1 binding site; PRE, progesterone response element; hPRE, PRE half-site; CRE, cAMP response element; AP-1, activator protein 1 binding site; C/EBP, CCAAT/enhancer binding protein motif; GRE/PRE, glucocorticoid response element/progesterone response element; PEA-3, phorbol ester response element; Pit-1, Pit-1 transcription factor binding site; Oct-1, octamer transcription factor-1 binding site; GATA, GATA transcription factor binding site; LIM, LIM-homeodomain factor binding site; GRAS, GnRH receptor activating sequence; DARE, downstream activin response element; SURG, sequence underlying responsiveness to GnRH; GnSE, GnRHR-specific enhancer; NF-Y, nuclear factor-Y binding site; NRE, negative regulatory element; GL-specific, granulosa-luteal cell-specific; GC, glucocorticoid. Note that the figure is not drawn to scale.

To date, the 5′ flanking regions of the mouse (97), rat (54), human (55, 98) and sheep (57) GnRHR I genes have been characterised. Although the mouse and rat promoters share > 80% homology over 1.9 kb, the rat promoter shares 55% homology with the human promoter over 2.2 kb, and 63% homology with the sheep promoter over 0.9 kb (99). There are several highly homologous regions within the proximal 500 basepairs of the mouse, rat, human and sheep promoters (99). A number of cis-elements have been conserved, in sequence as well as position, supporting their role as important functional elements. No functional characterisation of mammalian GnRHR II promoters has as yet been published, and therefore this section will focus on regulation of transcription of the mammalian GnRHR I gene.

The mouse GnRHR I proximal promoter was the first to be isolated and characterised (97). The major transcription start site in primary pituitary tissue (49) and αT3-1 cells (49, 97), is located at −62 (all numbering is relative to the translation start site) and is not associated with a consensus TATA box. In addition to this site, Clay et al. (100) identified other pituitary transcription start sites at −90 and −200 bp in αT3-1 cells. Gonadotroph-specific activity of the mouse promoter in αT3-1 cells is conferred by a tripartite basal enhancer, which includes binding sites for steroidogenic factor-1 (SF-1) at −244/−236, and activator protein-1 (AP-1) at −336/−330, respectively, as well as an element originally termed GnRHR-activating sequence (GRAS) at −391/−380 (102). The pan-pituitary homeobox transcription factor Pitx-1 has been shown by chromatin immunoprecipitation assay to interact with AP-1 in intact LβT2 cells, and functional evidence in other cell types indicate that this interaction might be important for GnRHR I gonadotroph-specific, basal promoter activity (103). In addition, the promoter region around −360, shown to bind LHX3 homeodomain protein in vitro and in intact cells, was recently demonstrated to be important for mouse GnRHR I basal promoter activity in αT3-1 cells (104). Experiments with transgenic mice suggest tissue-specific promoter usage for the mouse GnRHR I gene, because 1900 bp of mouse GnRHR I 5′ flanking sequence can drive reporter expression in pituitary, brain and testis, but not in the ovary, indicating an essential requirement for promoter elements located further upstream for ovary-specific expression in vivo (105).

In the rat proximal GnRHR I promoter, the transcription start site in αT3-1 cells was initially found to be 103 bp upstream from the start codon, with a putative TATA box 23 bp upstream from the transcription start site (54). A different study group later identified five major transcription start sites in αT3-1 cells, four of which are clustered around −103, and one situated at −30, along with several minor start sites (99). Maximal gonadotroph-specific expression of the rat GnRHR I is conferred by multiple regulatory domains within 1260 bp of 5′ flanking region. The proximal 183 bp constitutes a self-sufficient, but fairly weak promoter which confers basal but not gonadotroph-specific activity. A distal GnRHR-specific enhancer (GnSE), located between −1135 and −753, contains binding sites for GATA-related and LIM homeodomain-related factors, and facilitates gonadotroph-specific expression through functional interaction with an SF-1 site at − 245 (99, 106) (Fig. 2). The mechanisms involved in gonadotroph-specific expression of the mouse and rat GnRHR I are therefore clearly different, although both involve SF-1 sites. An AP-1 site in the rat promoter is also involved in basal promoter activity, but has no influence on the GnSE function. The function of the proximal rat promotor and the GnSE is supported by results obtained in transgenic mice, showing that the proximal 1.1-kb rat GnRHR I promoter is sufficient to drive gonadotroph-specific expression. Furthermore, 3.3 kb of the rat promoter was found to drive cell-specific expression of the transgene in gonadotrophs and certain areas of the brain (51).

The 5′ flanking regions of the human and sheep genes are much more complex than that of the mouse and rat genes, with the presence of multiple transcription start sites and CAP sites (57, 107). Although the sheep proximal 5′ flanking region is structurally similar to the mouse promoter, it has greater sequence homology to the human promoter (57). No further functional characterisation of sheep promoter elements has been performed. In stark contrast to the single start site identified in mouse pituitary tissue (49), 18 transcription start sites have been identified for the human GnRHR I gene in human pituitary tissue (98). These start sites are located between −1748 and −577 and are well dispersed among several TATA and CCAAT boxes. The proximal 173 bp of the human 5′ flanking region, although not a self-sufficient promoter, is critical for basal promoter activity in αT3-1 cells (108). However, characterisation of the human pituitary promoter has been hampered by unavailability of human gonadotroph cell lines, and the results in mouse cell lines may not be physiologically relevant. The mouse, rat and human promoters all contain several SF-1 sites, with at least one site in each promoter occurring in the 5′ untranslated region. For the human promoter, this site is situated at −140/−134 and is primarily responsible for mediating high cell-specific expression in αT3-1 cells (108), whereas the same function has not been assigned for similar sites in the mouse and rat promoters (situated at −15/−7 in both species) (49). An upstream Oct-1 binding at −1718 is also required for basal activity of the human promoter in αT3-1 cells (109).

The regulatory elements involved in expression of the mouse, rat and sheep GnRHRs have not been characterised in cells other than pituitary cell lines. However, cell-specific cis and trans elements have recently been identified for the human promoter in ovarian, placental and neuronal medulloblastoma cell lines. Expression of the human GnRHR I gene in both αT3-1 mouse gonadotroph cells and OVCAR-3 human ovarian carcinoma cells requires two promoter regions, located between −771/−557 and between −1351/−1022 (110). However, different trans-acting factors appear to bind to these regions in the different cell-types, possibly providing a mechanism for cell-selective expression (110). Two additional upstream promoters are responsible for high expression levels in human placental and ovarian granulosa-luteal cells, respectively (111). The granulosa (Fig. 2) cell-specific promoter is situated between −1300 and −1018, and contains a GATA element and two putative CCAAT/enhancer binding protein (C/EBP) motifs that were shown to be crucial in regulating GnRHR I transcription in the human ovarian granulosa-luteal cell lines SVOG-4o and SVOG-4m (111). GnRHR I expression in human placental cells requires a distal promoter region, located between −1737/−1346, in combination with a proximal region, between −707 and −167 (109). At least five placental transcription start sites were identified within the distal promoter region (109). A strong negative regulatory element is located between −1018 and −771, with a strong positive regulatory region between −771 and −577 (109). The distal placenta-specific promoter also contains an Oct-1 and an AP-1 binding site, required for basal expression in placental cells and other cell-types, as well as a cAMP response element (CRE) and a GATA element, essential for placenta-specific expression (109). Taken together, these studies indicate that various reproductive tissues differentially utilise downstream and upstream promoter elements and transcription factor binding sites for tissue-specific transcription of the human GnRHR I gene (111).

The transcription factor Oct-1 appears to regulate basal GnRHR I gene expression both positively and negatively, depending on the species and cell-type. As already mentioned, Oct-1 is required for basal expression of the human GnRHR I gene in several cell types, including placental, ovarian and gonadotroph cell lines, via an Oct-1 binding site at −1718 (109). On the other hand, in placental JEG-3 cells, ovarian OVCAR-3 cells and αT3-1 cells, Oct-1 acts as a potent repressor of the human GnRHR I promoter via a negative regulatory element (NRE) at position −1017 (112). Oct-1 is also involved in basal and GnRH-stimulated activity of the mouse GnRHR I promoter in αT3-1 cells via the SURG-1 (Sequence Underlying Responsiveness to GnRH) element (48).

The mouse CRE has been found to be essential for basal promoter activity in some pituitary cell lines, such as LβT2 gonadotroph cells (Sadie et al., unpublished data) and GGH3 somatolactotroph cells (113), but the rat CRE does not appear to be involved in basal promoter activity in αT3-1 cells (114). A CRE at position −1650 is required for placenta-specific expression of the human GnRHR I gene (109). These findings indicate a cell- and/or species-specific contribution of CREs to basal GnRHR I expression levels.

Transcriptional regulation of GnRHRs in the pituitary and in gonadotroph cell lines by physiological signals

GnRH

Homologous regulation of the GnRHR I is a physiologically relevant mechanism for increasing pituitary sensitivity to GnRH during ovulation (31). GnRH I activation of GnRHR I is thus a potent stimulus for increased expression of multiple genes including the gene encoding the GnRHR itself. GnRH I regulates the GnRHR I in a biphasic manner, with initial (short-term) exposure to hormone leading to an increase in receptor expression, whereas prolonged exposure leads to receptor down-regulation (32). It is widely accepted that pulsatile GnRH I stimulation is essential for appropriate GnRHR I expression levels, at the same time avoiding receptor down-regulation due to continuous hormonal stimulation (115). GnRH I pulse frequency and amplitude vary with physiological state, during the oestrous cycle in mammals and the menstrual cycle in humans, as well as during puberty and menopause (32). Regulation of pituitary GnRHR I mRNA levels and receptor numbers by GnRH I also differs between sexes (29). Recent experiments in transgenic mice show that mutation of the AP-1 site at −336 leads to a loss of GnRH I regulation of the mouse GnRHR I promoter (105).

The effects of GnRH I on GnRHR I protein and/or mRNA levels in primary pituitary cultures and cell lines suggest a direct mechanism of GnRH on pituitary cells, with a combination of both transcriptional and post-transcriptional mechanisms regulating GnRHR I expression levels. In attempts to mimic the situation in vivo, rat pituitary cultures were stimulated with GnRH I in a pulsatile fashion, resulting in increased GnRHR I mRNA levels (30). The mechanism appears to involve MAPK and possibly also cAMP/PKA (116, 117). Different pulse frequencies were found to have different effects on GnRHR I mRNA, with higher pulse frequencies causing maximal stimulation (118).

In αT3-1 cells, continuous stimulation with GnRH I appears to decrease endogenous GnRHR I levels via post-transcriptional mechanisms, although transcriptional mechanisms also contribute (94, 119). By contrast, the expression levels of mouse GnRHR I promoter–reporter constructs transfected into αT3-1 cells increase in response to 100 nM GnRH I after 4–6 h of continuous stimulation (32). This GnRH I responsiveness was mapped to two regions, designated SURG-1 and SURG-2 (32). SURG-1 contains binding sites for nuclear factor Y (NF-Y) and Oct-1, and it was shown by chromatin immunoprecipitation assays that GnRH increased binding of these factors to SURG-1 in intact cells (48). SURG-2 contains the AP-1 site described earlier. GnRH I responsiveness via SURG-2 appears to be mediated by PKC-induced activation of JNK which increases expression, activity and binding of AP-1 proteins to SURG-2 (46). SURG-1 and SURG-2 can respond to GnRH I independently, but the AP-1 element is critical for conferring maximal GnRH I responsiveness (32). These findings are in agreement with the results obtained in transgenic mice (46). However, in the mouse promoter, responsiveness to GnRH I also involves binding of Smad and AP-1 factors to another composite element called GRAS, which occurs further upstream at position − 391/−380 (Fig. 2) (45, 52). This is discussed in more detail below.

Down-regulation of the transcriptional activity of the transfected human GnRHR I promoter-reporter construct by 24 h of continuous GnRH agonist treatment in αT3-1 cells is also mediated via an AP-1 element in a PKC-dependent fashion (120). In LβT2 cells, endogenous GnRHR I mRNA and protein levels are up-regulated upon long-term pulsatile GnRH I stimulation (96, 121), whereas long-term continuous stimulation down-regulates receptor levels (121). By contrast, both continuous and pulsatile stimulation induced only a small increase in the activity of a transfected 1.2 kb mouse GnRHR I promoter-reporter construct in LβT2 cells (121). Conn et al. (122) studied the regulation of mouse GnRHR I promoter activity in the GGH3 cell line, which was engineered by stably transfecting GH3 rat somatolactotroph cells with rat GnRHR I cDNA. Several intracellular signalling pathways were found to be involved in mediating the up-regulation of the mouse GnRHR I promoter activity by GnRH I in these cells, such as PKA (123), PKC and the Ca2+ signalling pathway (124). However, unlike the results in αT3-1 cells (31, 32), the AP-1 site does not appear to be involved (35). Although the PKA pathway mediates homologous regulation of the mouse GnRHR I promoter in GGH3 cells, this is not the case for the mouse or human promoters in αT3-1 cells, most likely reflecting differences in GnRHR I G-protein coupling between the cell lines. However, functional studies do indicate a role for the PKA pathway and cAMP response elements (CREs) in regulating GnRHR I mouse, rat and human promoter activity. These promoters all contain functional CREs and are up-regulated by activators of the PKA pathway in αT3-1 cells (49, 114, 125). It is thus likely that in pituitary or extra-pituitary cells in which the GnRHR I can couple to Gs, homologous regulation will involve the PKA pathway. Other factors likely to be involved in mediating PKA responses, as shown in αT3-1 cells, are CREB (94) for the rat and SF-1 (49, 114) for the rat and mouse promoters.

Steroids

Studies in rat, sheep and cow conclude that oestradiol increases the level of GnRHR I mRNA and protein in pituitary (126, 127) consistent with a requirement for a strong, prolonged LH surge for ovulation during the preovulatory phase of the reproductive cycle. Experiments in ovariectomised transgenic mice harbouring a sheep GnRHR I promoter-reporter construct, as well as experiments in sheep primary pituitary cells (128–131), suggest that transcription is the predominant mechanism of oestradiol up-regulation of GnRHR I numbers in the pituitary. However, oestradiol stimulation of αT3-1 cells was found to down-regulate GnRHR I numbers (132), whereas oestradiol stimulation of LβT2 cells had little effect on endogenous GnRHR I gene expression (96, 132). These conflicting results highlight the apparent discrepancies that may occur when using transformed cell lines compared to primary cells that contain mixed cell populations. In addition, one group reported that the GnRHR I mRNA levels increase before an increase in circulating concentration of oestradiol (133), leading them to postulate that a decrease in progesterone, rather than an increase in oestradiol, is required for up-regulation of GnRHR I numbers.

In most mammals, high levels of progesterone correlate with reduced GnRHR I protein levels in pituitary and reduced pituitary responsiveness to GnRH I, such as that occurring during the luteal phase of the menstrual cycle and during pregnancy (126, 134, 135). In sheep pituitary cells, progesterone was found to dramatically down-regulate GnRHR I numbers within 48 h (129, 136), consistent with a direct effect of progesterone on the pituitary. Progesterone was also able to prevent oestradiol- and inhibin-induced increases in GnRHR I mRNA levels in these cells. Recent results with the human GnRHR I promoter in αT3-1 cells, showing that progesterone administration and overexpression of progesterone receptor (PR) isoforms inhibited GnRHR I promoter activity (137), suggest that, at least for the human promoter, repression by progesterone occurs via direct transcriptional effects on the GnRHR I promoter in gonadotrophs. Furthermore, this negative effect was shown to occur via a glucocorticoid response element (GRE)/progesterone response element (PRE) at −535/−521, which has 75% homology to a consensus progesterone response element (Fig. 2), and to which PR isoforms were shown to bind in vitro (137). In the same study, a half-PRE binding site was shown to be located at −402/−397. However, this site did not play a role in the progesterone-mediated transcriptional effects. Interestingly, another putative GRE/PRE is located further upstream (55), but its function remains unknown.

In male rats, pituitary GnRHR I mRNA levels appear to be repressed by testosterone because a negative correlation exists between mRNA levels and testosterone concentrations in serum (30, 64). GnRHR I numbers in primary pituitary cultures from male rats decreased after treatment with α-dihydrotestosterone (138), consistent with in vivo results and suggesting direct actions of α-dihydrotestosterone on the pituitary. By contrast, α-dihydrotestosterone up-regulated GnRHR I mRNA levels in LβT2 cells (139).

It is well documented that chronic or prolonged stress results in inhibition of gonadotrophin secretion and inhibition of reproduction in mammals, whereas the effects of acute stress are less clear and can even stimulate reproduction (140). Although the mechanisms whereby stress regulates reproduction in mammals are not well defined, there is evidence that glucocorticoids play an important role in modulating pituitary responsivess to GnRH I, as part of a feedback mechanism from adrenal to pituitary (140, 141). Further evidence for direct actions of glucocorticoids on pituitary is provided by findings that cortisol inhibits GnRH-induced LH release from bovine and porcine primary pituitary cells (142, 143). One mechanism whereby glucocorticoids may regulate GnRH responsiveness in pituitary may be via modulating GnRHR levels. Rosen et al. (144) showed that glucocorticoids augmented GnRH I-induced increase in GnRHR I numbers in castrated testosterone-replaced male rats. However, earlier studies in rats did not show a change in GnRHR I levels after treatment with corticosterone (138, 145). In sheep, administration of cortisol led to a decrease in GnRHR protein, but did not reduce GnRHR I mRNA levels (146). These experiments suggest that varying effects of glucocorticoids on GnRHR I levels may depend on species, the cellular milieu, and the dose, type and duration of glucocorticoid administration. However, a direct positive transcriptional effect of glucocorticoids on the mouse GnRHR I promoter has been established. Glucocorticoids increased endogenous GnRHR I mRNA levels in LβT2 cells, whereas pretreatment with GnRH I further augmented this increase (96). Glucocorticoids can also directly up-regulate activity of the mouse GnRHR I promoter in GGH3 cells (35). Although the tested 1.2 kb of 5′ flank of the mouse gene does not contain a classical GRE (Fig. 2), the glucocorticoid-responsive region of the mouse GnRHR I promoter was mapped to the AP-1 site at −336 in GGH3 cells (35). The results from this study suggest that liganded glucocorticoid receptor interacts directly or indirectly with AP-1 proteins, such as c-Jun, to increase GnRHR I transcription (35).

Other physiological regulators

Activin and inhibin, both members of the transforming growth factor-β family of proteins, are produced by primary gonadotrophs (147), αT3-1 (148) and LβT2 cells (149), and exert autocrine/paracrine effects on pituitary cells. Activin-A stimulates the rate of synthesis of new GnRHRs in rat pituitary cell cultures (150), and decreases receptor numbers in sheep pituitary cultures (151). Inhibin was found to prevent the stimulation of receptor synthesis by GnRH I in rat pituitary cultures (152), but increases GnRHR I mRNA levels (129) and receptor numbers (151) in sheep pituitary cultures. Whether these differences are species-specific or due to different experimental conditions is not known. In αT3-1 and LβT2 cells, long-term stimulation with activin-A up-regulates endogenous GnRHR I mRNA synthesis and mouse GnRHR I promoter-reporter activity (148, 149), and pretreatment of αT3-1 cells with activin enhances the response of the mouse GnRHR I promoter to GnRH I (45). Follistatin blocks the activin-mediated stimulation at both mRNA and promoter level. In addition, follistatin decreases the basal activity of the mouse GnRHR I promoter in αT3-1 and LβT2 cells, indicating that endogenous activin maintains basal GnRHR I expression levels in these cells (45, 148, 149). Activin responsiveness of the mouse GnRHR I promoter was mapped to the GRAS element (153) described earlier, together with a region immediately downstream from GRAS, termed DARE (down-stream activin regulatory element) (154) (Fig. 2). The mouse GRAS element is a composite regulatory element for which the functional activity in αT3-1 cells depends on the proper organisation and assembly of a multiprotein complex, which includes Smad, AP-1 and FoxL2 proteins (47). Basal GnRHR I promoter activity, as well as responsiveness to GnRH I and to activin require binding of Smad factors to the Smad binding element, as well as binding of AP-1 to a novel AP-1 element contained within GRAS (Fig. 2) (45, 52). The LIM-homeodomain protein LHX2 was shown to bind the DARE sequence in vitro (154). It has been postulated that activin responsiveness requires a specific configuration of multiple transcription factors on these distinct elements, to form a complex activin-responsive ‘enhanceosome’ (154). Interestingly, the sequence of the corresponding GRAS element in the rat GnRHR I promoter differs from the mouse GRAS by only one base-pair, but does not confer activin responsiveness to the rat promoter (106, 154), suggesting that the rat DARE sequence is nonfunctional for activin responsiveness.

Pituitary adenylate cyclase activating polypeptide (PACAP) is a hypothalamic peptide hormone that modulates pulsatile GnRH I release from the hypothalamus and responsiveness to GnRH I, as well as regulates gonadotrophin subunit expression (155). The mouse, rat and human GnRHR I promoters have all been shown to be regulated by PACAP in αT3-1 cells (49, 114, 125). For the rat and human promoters, this has been shown to involve PKA (114, 125). Two promoter elements, designated PARE (PACAP response element) I and PARE II, are required for the PACAP response of the rat GnRHR I promoter. PARE I includes the SF-1 binding site at position −245/−237, along with binding sites for additional factors, whereas PARE II contains an imperfect cAMP response element (CRE) at position −110/−103 that can bind CREB (114). Both the SF-1 site and the imperfect CRE are conserved in relative position in the mouse GnRHR I promoter (Fig. 2), raising the possibility that a similar mechanism could be responsible for the PACAP response of the mouse promoter.

Transcriptional regulation of GnRHRs in extra-pituitary tissues and cell lines by physiological signals

Ovary and placenta

By contrast to results obtained in primary rat pituitary cultures (30, 156), homologous regulation of GnRHR I by GnRH I has not been consistently observed in rat primary granulosa cells (157, 158). Treatment of human granulosa cell lines (SVOG-4o and SVOG-4m) with high and low doses of GnRH II induced a significant decrease in GnRHR I mRNA levels, whereas GnRH I induced a down-regulation at high and an up-regulation at low doses, showing that the two ligands regulate GnRHR I transcription differentially (16). Responsiveness of GnRHR I transcription to oestradiol appears to vary between ovarian cell types. In human primary ovarian surface epithelial cells, as well as in OVCAR-3 ovarian cancer cells, treatment with oestradiol caused a significant down-regulation of GnRHR I mRNA (37, 159, 160). In human granulosa-luteal cells, short-term oestradiol treatment (6 h) increased GnRHR I mRNA levels, whereas long-term treatment (48 h) decreased GnRHR I mRNA levels (161). This observation is consistent with what was found in vivo (30, 162). A recent study in ovarian cancer cells demonstrated that oestradiol represses GnRHR I transcription in a ERα-dependent and ERβ-independent way, via an AP-1-like motif at −130/−124 (9, 37). Repression of GnRHR I promoter activity by oestradiol did not involve direct binding of the ER to the AP-1 site, suggesting that the ER interacts with other proteins bound to this motif, such as c-Jun or c-Fos. Other physiological modulators of GnRHR I expression in human primary granulosa-luteal cells are hCG (38) and melatonin (34), although the mechanisms are not well defined. hCG down-regulates GnRHR I mRNA levels in primary ovarian granulosa-luteal cells without changing GnRH I expression (38). Until recently, regulation of reproductive function by melatonin was assumed to be restricted to the level of the pituitary and the hypothalamus. However, the presence of melatonin in the follicular fluid and of melatonin binding sites in the ovary suggests a role for this hormone in the ovary. In support of this, melatonin reduces both GnRH I and GnRHR I mRNA levels in human primary granulosa-luteal cells (34), suggesting that melatonin directly regulates ovarian function.

In the human choriocarcinoma JEG-3 and the immortalised extravillous trophoblast IEVT placental cell lines, the human GnRHR I mRNA is up-regulated after 24 h of continuous stimulation with GnRH I (163). This may be a tissue-specific mechanism to help maintain GnRH I-stimulated hCG secretion throughout pregnancy. The kinase pathways and transcription factors mediating this response have not been determined, but may involve the PKC pathway, as shown in αT3-1 cells, and/or perhaps the PKA pathway, because the human gene is up-regulated by activators of the PKA pathway, via binding of CREB to two AP-1/CRE elements (9). Progesterone has a positive effect on GnRHR I promotor activity in the JEG-3 placental cell line in contrast to the repression observed in αT3-1 cells (137). The GRE/PRE at position −535/−521 was shown to mediate PR regulation in both αT3-1 gonadotroph and JEG-3 placental cells (137). Furthermore, it was shown that both PR-A and PR-B isoforms bound to the PRE in vitro and that the balance between PR-A and PR-B overexpression in the different cell lines can determine the response to progesterone (137). Whereas PR-A inhibits transcription in both placental and pituitary cells, PR-B activates transcription in placental cells, and inhibits transcription in pituitary cells.

Brain

In the GT-17 hypothalamic GnRH neuronal cell line hCG was found to down-regulate the expression of the GnRHR I as well as of GnRH I (39). By down-regulating the GnRH/GnRHR system, hCG may disrupt the autocrine regulation in hypothalamic GnRH neurones. Results obtained in the TE671 neuronal medulloblastoma cell line showed that progesterone has a potent negative effect on human GnRHR I promoter activity, and up-regulates GnRH I expression (36). Overexpression of PR-A increased sensitivity towards progesterone-mediated repression of the GnRHR I gene, whereas PR-B reversed the PR-A-induced repression, suggesting that negative regulation occurs in the absence of overexpression via endogenous PR-A (36).

Immune cells

Consistent with the idea that immune function is differentially regulated during the reproductive cycle, expression of GnRHR I in lymphocytes was shown to vary throughout the oestrous cycle and parallels expression in pituitary (87). In vitro administration of GnRH and oestradiol increased GnRHR I mRNA levels in immune cells (87) but the mechanisms remain unclear.

Discussion and future perspectives

GnRHRs are expressed in multiple mammalian tissues and cell types (Table 1) and have diverse functional roles, in addition to the established endocrine role for gonadotroph GnRHRs in regulation of reproductive physiology. Furthermore, the profile of hormones that regulate expression of the GnRHR I is expanding and now includes adrenal glucocorticoids, gonadal sex steroids, as well as hCG, activin, inhibin, follistatin, PACAP, melatonin and GnRH. It is well-established that sex steroids and glucocorticoids feedback and inhibit the HPG axis at the hypothalamic, pituitary and gonadal levels, providing mechanisms for fine-tuning reproductive function (140). Studies also suggest additional connections between physiological processes, demonstrating that the neuroendocrine, immune, inflammatory and stress-response systems are functionally integrated and bidirectionally regulated (164–169). Although the target genes and intracellular mechanisms of such feedback regulation are poorly defined, the GnRHR is emerging as a potential target gene for facilitating cross-talk between these systems in multiple tissues via autocrine/paracrine and endocrine signalling. The identification of two forms of both the mammalian hormone and receptor, expressed in multiple tissues, and in many cases coexpressed, further increases the diverse signalling potential of the GnRH/GnRHR system in mammals. However, further work is needed to clearly establish the presence of GnRHR I and/or GnRHR II in the various tissues and cells, using receptor-subtype specific functional and morphological methods.

Studies to determine the promoter elements, transcription factors and detailed molecular mechanisms of transcription regulation have focused on the mouse, rat and human GnRHR I promoters. No characterisation of the mammalian GnRHR II promoters has been published to date, most likely due to the absence of suitable model systems where the GnRHR II is known to be endogenously expressed. Most of the GnRHR I work has been performed using transient transfections in cell lines, using the human, mouse and rat GnRHR I promoters transfected into αT3-1 and LβT2 mouse pituitary gonadotroph cell lines, and, more recently, with the human GnRHR I promoter transfected into human ovarian, placental and neuronal medulloblastoma cell lines. Some work has also been performed in primary cells, which usually contain mixed cell populations. It is clear that there are many apparent discrepancies between the results obtained in the various model systems, highlighting the difficulties inherent in finding suitable model systems that are physiologically relevant. These discrepancies may be due in part to the absence of native chromatin structure when comparing responses of transfected promoter constructs versus endogenous genes, or due to cell-specific differences between cell lines. In addition, indirect effects of one cell type on other cell types in primary cultures, or the use of heterologous expression systems may result in discrepancies. Finally, variations in experimental procedures such as doses and times and method of administration (e.g. continuous versus pulsatile) of hormonal stimulation, and culture conditions may lead to different results. Some experiments in transgenic mice have been particularly helpful in confirming the results obtained in vitro, such as the finding that an AP-1 site in the mouse promoter is necessary for homologous regulation (46). Two elegant studies have also recently addressed the issue of protein–DNA interactions in intact cells by employing the technique of chromatin immunoprecipitation assays to identify factors binding to the endogenous mouse GnRHR I promoter, to confirm the results obtained in vitro (48, 103). Certainly, future studies will be helpful in determination of protein–DNA interactions in intact cells for other factors.

Despite the limitations of the present model systems, several interesting insights into the mechanisms of GnRHR I regulation have emerged. Experiments performed in transgenic mice indicate that the mouse gene does not exhibit multiple, widely spaced, tissue-specific promoter usage in the pituitary, brain and testis, although they suggest that a different promoter that is upstream of −1900 in the mouse GnRHR I gene is used in the ovary (105). Similarly, experiments with transgenic mice harbouring the rat GnRHR I promoter found that a single 3.3-kb promoter is capable of driving transcription in gonadotrophs and multiple brain tissues (51). The human promoter uses different promoters in the pituitary compared to the placenta, and even within different ovarian cell types, although these regions all occur within the first 2 kb of 5′ flanking region (109–111). Evidence is emerging from work on the mouse and human GnRHR I promoters that mechanisms of regulation of transcription in the ovary appear to differ substantially as compared to other tissues investigated to date (105, 110, 111).

Mouse, rat and human promoters also appear to exploit the concept of multiple commonly expressed transcription factors binding to a composite element in a particular structural organisation, to achieve tissue-specific expression. In addition, the mouse GnRHR I promoter uses the complex SURG and GRAS composite elements to achieve homologous regulation in αT3-1 cells (32, 45, 48, 52, 102, 154). Although the components of these complex elements differ between species, they appear to often rely on some common factors. The human, mouse and rat promoters all contain several SF-1 sites, with at least one being involved in gonadotroph-specific expression (102, 106, 108), although SF-1 is expressed in several other tissues. Interestingly, the mouse and human promoters contain at least one AP-1 site shown to be involved in mediating homologous regulation via the PKC pathway in the pituitary (46, 120). However, the apparent lack of an AP-1 site in the sheep promoter (57) suggests that the PKC pathway may not be involved in homologous regulation in this species. It appears likely that both the rat and mouse promoters could employ similar mechanisms for PACAP regulation (49, 114), based on the presence of conserved promoter elements, whereas species-specific differences appear to exist for their regulation by activin (106, 154). Mechanisms of homologous regulation also appear to be cell-specific, and are likely to depend on the physiological state of the cells, as shown by different findings for GnRHR I regulation in rat primary granulosa cells (157, 158) versus primary rat pituitary cultures (30, 156).

In the light of the potential role of GnRHR genes as targets for cross-talk between various physiological systems, it is of great interest to determine whether adrenal and sex steroids directly regulate expression of GnRHR genes, and whether these mechanisms are transcriptional or post-transcriptional. Studies in some mammals show repressive effects with progesterone (126, 134, 135) and stimulatory effects with oestrogen (29, 126, 170, 171) on GnRHR I expression in the pituitary. However, experiments in primary cells and cell lines from pituitary and extra-pituitary origins reveal no consistent picture, suggesting species-, cell- and/or promoter-specific differences in response to these steroids (37, 96, 129, 132, 136, 137, 161). To date, no classical oestrogen or androgen response elements have been identified in the mouse, rat, human or sheep GnRHR I promoters. Although the same applies for glucocorticoid and progesterone response elements in the mouse, rat and sheep promoters, several PRE/GREs have been identified for the human promoter (137). Data are accumulating to suggest that steroid regulation of these promoters may occur via nonclassical pathways other than up-regulation via binding of homodimers of steroid receptors to steroid response elements. Several recent studies pinpoint a direct transcriptional effect of some steroids on mammalian GnRHR I promoters. A direct effect of oestradiol on the human GnRHR I promoter in human ovarian and breast cancer cell lines has been established, mediated via an AP-1 site (37). Interestingly, different ER isoforms exhibited different effects on the promoter, with the oestradiol response being ERα-dependent, but ERβ-independent (37). An AP-1 site is also involved in glucocorticoid regulation of mouse GnRHR I promoter activity in the GGH3 rat somatolactotroph cell line (35). However, in this case, an up-regulation is observed in response to glucocorticoids. Because it is well established that steroid receptors can inhibit transcription of several target genes via interference with the actions of AP-1, the up-regulation of mouse GnRHR I promoter activity via AP-1 may represent a novel mechanism. Progesterone has been shown to both up-regulate [in a human placenta cell line (137)] and down-regulate [in the mouse αT3-1 gonadotroph (137) and human medulloblastoma (36) cell lines] transcription of the human GnRHR I promoter, via binding of the PR to a PRE at −535/−521. Reminiscent of the situation with the ER described above, different isoforms of the PR have differential cell-specific effects on the human promoter (36, 137). Taken together, these ER and PR results suggest that, for the human promoter, variations in the levels of receptor isoforms may be a widely used mechanism for differential tissue-specific regulation in pituitary and extra-pituitary tissues. This could allow differential expression levels of the GnRHR I by varying the relative concentrations of receptor isoforms in response to different signals and thereby integrate connections between multiple physiological processes.

Continued research on the regulation of expression of mammalian GnRHRs is important for understanding reproductive endocrinology and could lead to novel insights on receptor function.

Acknowledgements

This work was supported by grants to J.P.H. from the Medical Research Council and National Research Foundation (NRF) in South Africa, and Stellenbosch University. Any opinion, findings and conclusions or recommendations expressed in this material are those of the author(s) and therefore the NRF does not accept any liability in regard thereto.

Ancillary