| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, British Columbia, Canada V6H 3V5
Correspondence: Address all correspondence and requests for reprints to: Peter C. K. Leung, Ph.D., Department of Obstetrics and Gynecology, University of British Columbia, 2H30-4490 Oak Street, British Columbia Women’s Hospital, Vancouver, British Columbia, Canada V6H 3V5. E-mail: peleung{at}interchange.ubc.ca
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
Until now, more than a dozen isoforms of GnRH sharing 10–50% amino acid identity have been found in vertebrates (14). It is generally thought that most vertebrate species possess at least two, and usually three, forms of GnRH, which differ in their amino acid sequences, localizations, and embryonic origins. In addition to GnRH-I, a second GnRH subtype (termed GnRH-II) that was originally identified from chicken hypothalamus has been found in humans (15, 16). This second GnRH form differs from GnRH-I by three amino acid residues at positions 5, 7, and 8 (His5Trp7Tyr8GnRH-I) and is conserved from primitive fish to humans (16, 17). One of the established biological functions specific to GnRH-II is to serve as a potent inhibitor of K+ channels in the amphibian sympathetic ganglion (17). Inhibition of these ion channels facilitates rapid excitatory transmission by conventional neurotransmitters and may provide a general neuromodulatory mechanism for GnRH-II in the nervous system. Recently, Temple et al. (18) have shown that GnRH-II, but not GnRH-I, activates mating in energetically challenged musk shrews, suggesting a role of the evolutionarily conserved GnRH form in coordinating energy and reproductive behavior. In humans, a growing number of extrapituitary GnRH-II actions, such as suppressing tumor proliferation (3, 7, 8, 19, 20, 21, 22, 23), have been demonstrated although a full-length type II GnRH receptor transcript has not yet been identified in any of the human tissues or cell types.
|
II. GnRH Isoforms in Humans: GnRH-I and GnRH-II |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
The human GnRH-I gene is composed of four exons separated by three introns and is present as a single gene copy on chromosome 8p11.2-p21 (Fig. 1
) (26, 27). The first exon of the gene is untranslated and consists of 61 bp in mRNA expressed in the hypothalamus. The second exon encodes the signal sequence, the GnRH decapeptide, the GKR processing signal, and the first 11 GAP residues. The third exon codes for the next 32 GAP residues. The fourth exon encodes the remaining GAP residues and contains the translation termination codon as well as the entire 3'-UTR (24, 26).
|
). However, the human GnRH-II gene (2.1 kb) is shorter than the GnRH-I gene (5 kb) because introns 2 and 3 of the latter are much larger (16). Moreover, although their corresponding precursor proteins are quite similar in length, the GAP is 50% longer in the GnRH-II precursor (84 vs. 56 amino acids). In fact, a similar disparity in GAP has also been reported in the placental mammal, tree shrew (76 vs. 56 amino acids) (28), suggesting that a relatively larger GAP may be a common characteristic among mammalian GnRH-II precursors.
B. Tissue distribution in humans
1. Brain.
The most prominent difference in the tissue distribution of GnRH-I and GnRH-II in humans is that the latter isoform is expressed at the highest level outside the brain (16). The levels of GnRH-II mRNA in the kidney are approximately 30-fold higher than in any brain region, whereas the expression in the bone marrow and prostate is about 4-fold greater than in the brain (16). Conversely, GnRH-I expression was not observed at a high level outside the brain (16). In humans, cell bodies of GnRH-I neurons are concentrated in the preoptic area and basal hypothalamus. However, they can also be found in the septal region and anterior olfactory area, as well as the cortical and medial amygdaloid nuclei (29). On the other hand, immunoreactive GnRH-I fibers are localized predominantly in the median eminence and infundibular stalk although substantial projections to the neurohypophysis can be detected (30, 31). Using in situ RT-PCR, expression of GnRH-I mRNA has been demonstrated in normal human pituitary and various types of pituitary adenomas (32, 33).
In human brain, immunopositive GnRH-II signals localize mainly in the periaqueductal region of the midbrain (34). However, expression of GnRH-II mRNA in the human brain was found to be most abundant in the caudate nucleus and, to a lesser extent, in the hippocampus and amygdala (16). Using RT-PCR and Southern blot analysis, two GnRH-II mRNA variants were identified in human fetal brain and adult thalamus but not in adult kidney. These transcripts differ in the size of their GAPs, which are predicted to contain 77 and 84 amino acid residues (16).
Coexpression of GnRH-I and GnRH-II has been demonstrated at both the mRNA and protein levels in certain human neuronal cell lines in which where the concentration of GnRH-I is 10- to 40-fold higher than that of GnRH-II (35).
2. Placenta.
It has long been shown that human placenta in vitro synthesizes and secretes GnRH-I that is immunologically indistinguishable from its hypothalamic counterpart (36, 37). Likewise, Siler-Khodr and Grayson (38) have shown that GnRH-II is released from human placenta in vitro in a pulsatile fashion and that this second GnRH form is more resistant than GnRH-I to degradation by placental enzymes. Examination of the spatiotemporal distribution of these hormones revealed that both GnRH-I and GnRH-II mRNAs are expressed in human first-trimester placenta (39). However, only GnRH-I is also expressed in tissues obtained at term (39). Using immunohistochemistry, both hormones were found to localize in the mononucleate villus and in distinct subpopulations of the extravillous cytotrophoblast. However, GnRH-I is also present in the outer multinucleated syncytiotrophoblast layer and in cultures of cytotrophoblasts allowed to undergo differentiation and fusion in vitro (39).
3. Uterus.
Expression of GnRH-I mRNA has been demonstrated in virtually all human uterine compartments (40, 41, 42, 43, 44, 45). Interestingly, a dynamic expression pattern is observed in the endometrium as well as in isolated endometrial cells such that a significant increase in transcript levels is detected in the secretory phase of the menstrual cycle (44, 45). In support of these observations, GnRH-I immunoreactivity has been found in all endometrial cell types, with the most intense staining being observed in the luteal phase (45).
The spatiotemporal expression of GnRH-II has also been investigated in human endometrium. Throughout the entire reproductive cycle, two splice variants of GnRH-II mRNA are expressed, with the shorter transcript carrying a 21-bp deletion, which reduces the length of GAP from 77 to 70 amino acids (46). Like GnRH-I, GnRH-II immunoreactivity is dynamically expressed in stromal and epithelial cells such that stronger signals are detected in the early and midsecretory phases than in the proliferative and late-secretory phases (46).
4. Ovary.
Expression of GnRH-I and GnRH-II mRNAs identical to their brain counterparts has been demonstrated in various human ovarian tissues including granulosa-luteal (GL) cells, ovarian surface epithelial (OSE) cells, and ovarian carcinoma (19, 20, 47, 48). In addition, expression of GnRH-I mRNA and protein has been found in the tubal epithelium of the fallopian tube during the luteal phase of the reproductive cycle (49).
5. Other tissues.
Both forms of GnRH are expressed in normal human breast tissue and are overexpressed in breast cancer (50, 51). Moreover, certain immune cell lineages such as T lymphocytes and peripheral blood mononuclear cells have been found to produce GnRH-I or both GnRH-I and GnRH-II (7, 52, 53). In Jurkat leukemic T cells, the concentration of GnRH-I is higher than that of GnRH-II, as determined by RIAs (7). Expression of GnRH-I protein has also been demonstrated in human seminiferous tubular cells (54) and preimplantation embryos, in which immunoreactive signals are localized in all the blastomeres as well as the trophectoderm and inner cell mass of the blastocyst (55).
C. Regulation of gene expression in humans
1. GnRH-I and GnRH-II.
In human OSE, GL, and OVCAR-3 ovarian cancer cells, treatment with GnRH-I analogs produces a biphasic effect on its mRNA levels such that high concentrations decrease whereas low concentrations increase the expression (20, 47, 48). In contrast, GnRH-I suppresses its mRNA levels in peripheral blood mononuclear cells in a dose-dependent manner (53). Homologous down-regulation of GnRH-I mRNA levels has also been demonstrated, in a dose- and time-dependent fashion, in rat hypothalamus in vivo and in GT1-1 cells (56, 57). On the other hand, heterologous regulation of GnRH-I expression has been studied only in human GL cells, in which GnRH-II or its analog causes down-regulation of GnRH-I mRNA levels at a wide range of concentrations (20).
2. Gonadal steroid hormones.
There are substantial lines of evidence indicating that the expression of GnRH-I and GnRH-II is differentially regulated by gonadal steroids. In human GL, OVCAR-3, and TE-671 neuronal cells, treatment with 17ß-estradiol (E2) down-regulates the steady-state GnRH-I mRNA levels (58, 59, 60, 61). This E2 action is believed to be mediated via the nuclear estrogen receptor (ER) because cotreatment with the antiestrogen tamoxifen can abolish the inhibitory effect. Using the ER-negative Chinese hamster ovary-K1 cell line as a model, Chen et al. (62) demonstrated that E2 can repress the human GnRH-I promoter when ER
is overexpressed. Also, they found that the estrogen response area lies between 169 and 548 bp 5' of the upstream transcription start site of the GnRH-I gene. Similarly, E2 has been found to suppress the mRNA expression and promoter activity of the GnRH-I gene in mouse GT1-7 neurons, possibly via an ER-mediated mechanism (63). On the contrary, our laboratory has recently shown that E2 increases GnRH-II mRNA levels in a dose- and time-dependent manner in human GL cells (61). Likewise, a stimulatory effect of E2 on GnRH-II expression has been reported in TE-671 cells (60).
The role of progesterone in regulating GnRH-I and GnRH-II expression has been investigated in human GL cells. Whereas treatment with the progesterone receptor (PR) antagonist RU486 does not affect GnRH-I mRNA levels, the levels of GnRH-II transcript are stimulated by the antagonist in a dose and time-dependent manner (61), suggesting that the gonadal steroid has an inhibitory role in GnRH-II expression in the ovary.
Regulation of GnRH-I gene expression by androgen has been examined in the androgen receptor-expressing GT1-7 cell line. In these cells, treatment with 5-dihydrotestosterone causes a time-dependent reduction in GnRH-I mRNA levels, and this repression can be blocked by the androgen receptor antagonist hydroxyflutamide (64). However, no significant changes in GnRH-I expression can be observed when the hypothalamic neurons are treated with cell-impermeable testosterone-BSA conjugates (65), indicating that the androgen action is mediated via classical nuclear receptor activation.
3. Gonadotropins.
Further evidence that the expression of the two forms of GnRH is differentially modulated comes from studies on their regulation by gonadotropins, which mediate their actions by stimulating intracellular cAMP production and activating the protein kinase A signaling pathway. In human GL cells, treatment with FSH or human (h) chorionic gonadotropin (CG) up-regulates the mRNA levels of GnRH-II but down-regulates those of GnRH-I in a dose-dependent manner (20). Consistently, an increase in GnRH-II mRNA and protein levels in response to cAMP has been observed in TE-671 cells (66). This cAMP-activated GnRH-II gene expression is thought to occur at the transcriptional level because mutation of a putative cAMP-responsive element (CRE) in the human GnRH-II 5'-flanking region causes a reduction in both the cAMP-induced and basal promoter activities (66).
4. Other physiological regulators.
It has been shown recently that IL-1ß can up-regulate GnRH-I mRNA levels in human endometrial stromal cells in vitro in a dose-dependent manner (67). In addition, an increase in human GnRH-I gene transcription has been observed in NLT neuronal cells following IGF-I treatment (68). This stimulation is likely mediated via a consensus activator protein-1 (AP-1) motif in the proximal promoter region of the gene (68). Moreover, certain odorants have been found to induce a dramatic increase in GnRH-I mRNA levels and protein release in human olfactory cells (69), which share a common origin with GnRH-I neurons during organogenesis (70, 71). Using the immortalized GT1-7 neurons as a model, Roy et al. (72) and Roy and Belsham (73) have demonstrated that melatonin significantly down-regulates GnRH-I mRNA levels in a 24-h cyclical manner and that this regulation may involve the protein kinase C (PKC) and the ERK1/2 pathways.
5. Basal transcriptional regulation.
The molecular mechanism underlying neuron-specific expression of the human GnRH-I gene has been explored. By means of deletion analysis, a region between –992 and –795 of the human GnRH-I 5'-flanking region was found to be essential and sufficient for targeting luciferase expression in the hypothalamus and olfactory tissue in vivo (74). This region contains two specific DNA-binding sites for the POU homeodomain transcription factors Brn-2 and Oct-1. Functional studies revealed that overexpression of Brn-2, but not Oct-1, can transactivate both the human and mouse GnRH-I promoters (74). These findings thus indicate a role of Brn-2 or Brn-2-related proteins in regulating neuron-specific GnRH-I gene transcription.
In addition to the putative CRE described (66), we have recently uncovered a novel function of the untranslated first exon of the human GnRH-II gene in mediating full expression of GnRH-II promoter activity (75). Although this exon can work as a stand-alone enhancer element, its enhancer activity is strictly dependent on its position and orientation relative to the target sequence (75). Two putative E box binding sites and one Ets-like element are present juxtaposed to each other within the exon, and site-directed mutagenesis indicated that these motifs function in a cooperative manner to stimulate basal GnRH-II gene transcription (75). Detailed characterization of the E box binding factors revealed that the basic helix-loop-helix transcription factor AP-4 (76), the expression of which correlates well with that of GnRH-II, is an enhancer protein for the human GnRH-II promoter (75).
|
III. Molecular Characterization of Human Type I GnRH Receptor Gene |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
B. Genomic organization and chromosomal localization
In contrast to the genes of many other GPCRs, which are intronless and believed to have arisen by retroposition (89), the human GnRH receptor gene is composed of three exons separated by two introns and spans more than 15 kb along the chromosome (Fig. 1) (90, 91). Exon 1 contains the 5'-UTR and the first 522 nucleotides of the open reading frame, which encode the first three TM domains and a portion of the fourth TM domain. Exon 2 encodes the next 220 nucleotides of the reading frame, which encompass the remainder of the fourth TM domain, the fifth TM domain, as well as part of the third intracellular loop. Exon 3 contains the rest of the coding sequence and the 3'-UTR. Although the location of all the exon-intron boundaries of the human GnRH receptor gene is perfectly conserved in the rodent and ovine sequences, the first intron of the human gene is comparatively much smaller (92, 93, 94). Using genomic Southern blot and chromosomal in situ hybridization, the human GnRH receptor gene has been identified as a single copy on chromosome 4q21.2 (91, 95).
C. Untranslated and 5'-flanking regions
Five and 18 transcription start sites have been identified for the GnRH receptor gene in human brain and pituitary, respectively (90, 91). All these start sites are clustered into two regions, which are 579–819 and 1348–1751 bp upstream of the ATG initiation codon. Five typical polyadenylation signals residing within an 800-bp area in a cluster-like format are present in the 3'-UTR of the human GnRH receptor gene (90). Also, the 3'-UTR contains several ATTTA motifs, which are implicated in mRNA instability and are notably present in many RNAs that are rapidly degraded (96, 97). The size of the GnRH receptor mRNA predicted from the length of the 5'- and 3'-UTRs is about 5.5 kb, which is in close agreement with the reported size of the major transcript (4.7–5 kb) expressed in human pituitary.
Although the proximal 5'-flanking region of the human GnRH receptor gene shares a substantial homology with that of the rodent and ovine sequences (90, 92, 93, 94), the human gene possesses some distinctive features that are not observed in other species. One significant difference between the human and rodent genes is the location of their transcription start sites. Thus, whereas the start sites for the rodent genes are within 100 nucleotides from the initiation codon (92, 94), those for the human gene are no less than 703 bp (90). Another difference is that the human sequence contains multiple canonical TATA and CAAT boxes residing in close proximity to each other near the transcription start sites (90, 91). The presence of consensus TATA boxes is unusual among all the GPCRs sequenced to date, as many of these genes contain GC-rich promoter regions (98, 99, 100, 101).
D. Tissue distribution in humans
1. Pituitary.
Northern blot analysis has revealed a predominant GnRH receptor transcript of 4.7–5 kb as well as two fainter bands of 2.5 and 1.5 kb in human pituitary (91, 102). All these mRNA species contain the full-length coding sequence and are correctly spliced (91). Additionally, two splice variants of the human GnRH receptor, termed sb2 and sb3, have been identified in normal pituitary and pituitary adenoma (103, 104). The shorter transcript sb3 contains a 220-bp deletion in exon 2 such that it codes for a protein of only 177 amino acids, lacking the last four TM domains, the second and third extracellular loops, as well as the third intracellular loop. On the other hand, the sb2 variant carries a shorter deletion of 128 bp and arises from alternative splicing by accepting a cryptic acceptor site in exon 2. This deletion generates a truncated protein in which the glutamine residue at position 174 is followed by a stretch of 75 new amino acids (104). Interestingly, when coexpressed with the full-length receptor cDNA, the sb2 variant exhibits a dominant-negative action on GnRH receptor signaling, potentially by impairing insertion of the wild-type receptor protein into the plasma membrane. This inhibitory effect is highly specific for the GnRH receptor as signaling via other GPCRs is not affected (103).
The distribution of GnRH receptor immunoreactivity in normal and tumorous human pituitary has also been determined. In normal adenohypophysis, immunopositive signals colocalize with -subunit-, FSHß-, LHß-, TSH ß-, and GH-producing cells (105), suggesting that the receptor is expressed in gonadotropes, thyrotropes, as well as somatotropes. Consistent with the mRNA expression pattern in tumorous pituitary, immunoreactive GnRH receptor signals are frequently detected in adenomas derived from gonadotropes, somatotropes, and -subunit/null-cells (33, 105).
2. Placenta.
Using in situ hybridization, expression of GnRH receptor has been demonstrated in the cytotrophoblast and syncytiotrophoblast cell layers of human placenta (106). The temporal expression of the placental receptor parallels with the time course of hCG secretion and peaks at 9 wk (106). The full-length GnRH receptor cDNA has been cloned from various human placental cell types, and their nucleotide sequences are identical to that of the pituitary receptor (107). Northern blot hybridization indicated that a 2.5- and 1.2-kb transcript, but not the major 4.7–5-kb one found in the pituitary, are expressed in the placental cells (107).
3. Ovary.
High-affinity binding sites specific for GnRH-I have been detected in human corpus luteum, luteinized granulosa cells, epithelial ovarian carcinoma, and a number of ovarian cancer cell lines (108, 109, 110, 111). Interestingly, an additional type of GnRH-I binding site, which is of lower affinity but higher capacity, is found in EFO-21 and EFO-27 ovarian cancer cells (110). Using RT-PCR and Southern blot analysis, expression of GnRH receptor mRNA indistinguishable from its pituitary counterpart has been demonstrated in various ovarian compartments (Fig. 2) (19, 48, 112, 113).
|
5. Prostate gland.
The presence of specific binding sites for GnRH-I has been demonstrated in human prostate cancer and certain prostatic cancer cell lines (119, 120, 121). However, the affinity of these sites is generally lower than that of the pituitary receptor (120, 121). PCR products of the expected size for the GnRH receptor cDNA have been obtained from both normal and neoplastic prostate samples (Fig. 2) (112, 122, 123, 124, 125), whereas immunopositive signals have been detected in tumorous prostate tissue as well as intraprostatic lymphocytes (125). Expression of GnRH receptor in the prostate is further supported by the detection of a 64-kDa band, which corresponds well to the molecular mass of the pituitary GnRH receptor, in LnCAP and DU 145 cells (123).
6. Breast.
The existence of specific GnRH-I binding sites has been reported in breast carcinoma and MCF-7 mammary cancer cells (126, 127). Interestingly, the MCF-7 cells express two distinct types of binding sites, one of high affinity, which is specific for GnRH-I, and the other, which is only recognizable by GnRH-I antagonists (127). Expression of GnRH receptor immunoreactivity and mRNA with sequence identical to the pituitary counterpart has been demonstrated in both normal and malignant breast tissues (Fig. 2) (112, 128, 129). However, unlike its ligands (51), expression of GnRH receptor is not up-regulated in breast cancer cells (128).
7. Other extrapituitary tissues.
Multiple lines of evidence indicate that the expression of extrapituitary GnRH receptor is not limited to reproductive tissues. For instance, it has been demonstrated by RT-PCR and Southern blot hybridization that the receptor is also expressed in the liver, heart, skeletal muscle, kidney, and peripheral blood mononuclear cells (53, 130). Moreover, the receptor is expressed in melanoma cells at both the RNA and protein levels (131).
E. Regulation of gene expression in humans
1. GnRH-I and GnRH-II.
It has been well documented that pituitary GnRH receptor expression is dynamically regulated by GnRH-I such that subnanomolar concentrations up-regulate whereas high concentrations down-regulate receptor expression (132, 133, 134, 135). The extent of this up-regulation, however, is differentially controlled by varying GnRH-I pulse frequencies such that maximal stimulation is achieved at a frequency of every 30 min in cultured rat pituitary cells (136). Similarly, a biphasic effect of GnRH-I on GnRH receptor expression has been demonstrated in human GL, OSE, ovarian cancer, and peripheral blood mononuclear cells (20, 47, 48, 53). Conversely, a significant increase instead of decrease in receptor mRNA levels is observed in JEG-3 choriocarcinoma and extravillous trophoblast cells after chronic GnRH-I stimulation (107). The effect of GnRH-II on GnRH receptor expression has also been investigated in human GL cells. In contrast to the biphasic response induced by GnRH-I, treatment with GnRH-II or its analog significantly inhibits the mRNA levels of the receptor in the steroidogenic cells irrespective of the concentration used (20).
2. Gonadal steroid hormones.
The role of E2 in regulating GnRH receptor expression has been extensively studied at the pituitary level, where the gonadal effect is dynamic and apparently depends on the administration pattern (137, 138, 139, 140). In humans, modulation of GnRH receptor expression by E2 has been examined in extrapituitary tissues. Using semiquantitative RT-PCR, the steady-state mRNA levels of the receptor were found to be suppressed by E2 in GL and OVCAR-3 cells in a dose- and time-dependent manner (58, 59). This inhibitory effect can be abolished by cotreatment with tamoxifen, suggesting the mediation through the classical ER. Accordingly, E2 has been demonstrated to repress the human GnRH receptor promoter in ovarian cancer cells via an ER-dependent mechanism (141). In addition to modulating gene transcription, prolonged E2 treatment has been shown to increase glycosylation of the ovine GnRH receptor to generate a 43-kDa protein (142, 143). Although the biological significance of this estrogen-induced hyperglycosylation is unclear, it appears that this posttranslational modification is not associated with pituitary desensitization of LH response to GnRH-I (143).
Several lines of evidence indicate that progesterone directly inhibits GnRH receptor expression in the pituitary (144, 145, 146, 147). Intriguingly, our colleagues have revealed that the gonadal steroid has a dual role in controlling human GnRH receptor gene transcription such that the hormone suppresses the GnRH receptor promoter in gonadotropes but stimulates it in placental cells (148). The molecular mechanism underlying these opposing effects of progesterone will be discussed in detail below.
3. Gonadotropins.
In human GL cells, treatment with hCG for 24 h induces a dose-dependent inhibition of GnRH receptor mRNA levels (113). Accordingly, a similar effect has also been demonstrated in rat granulosa cells, rat testis, and GT1-7 neurons (149, 150, 151). However, contradictory results have been obtained from JEG-3 cells, in which the gonadotropin stimulates the receptor expression at the transcriptional level (107, 152). Thus, it is conceivable that the effect of gonadotropins on GnRH receptor gene expression may be tissue specific.
4. Melatonin.
It has become increasingly evident that melatonin can modulate ovarian functions in an autocrine manner (153, 154, 155, 156, 157). In human GL cells, transcripts encoding two melatonin receptor subtypes MT1 and MT2, which are homologous to their brain counterparts, have been identified (154, 157). Accordingly, treatment of the steroidogenic cells with melatonin significantly decreases the steady-state mRNA levels of the GnRH receptor and GnRH-I but increases those of the LH receptor in a dose-dependent manner (157). Because GnRH-I has been implicated as a luteolytic factor in the ovary (5), it is postulated that this melatonin-induced down-regulation of GnRH receptor expression may interfere with corpus luteum regression during the mid- and late luteal phases of the reproductive cycle (157).
5. Activin.
It has been demonstrated that activin A can stimulate the synthesis of GnRH receptor in rat pituitary cells (158). In T3-1 cells expressing the inhibin ßB-subunit, activin A increases GnRH receptor mRNA levels and promoter activity in a dose- and time-dependent manner, and these effects can be abolished by the activin antagonist follistatin (159). On the contrary, treatment with follistatin alone decreases the basal transcription of the gene, suggesting a potential autocrine and/or paracrine role of endogenous activin B in GnRH receptor expression in the gonadotropes (159). The biological significance of this activin-stimulated GnRH receptor gene transcription is confirmed by the observation that activin A pretreatment can enhance the GnRH-I responsiveness of the human glycoprotein -subunit promoter (159).
F. Pathophysiology of human GnRH receptor mutations
Idiopathic hypogonadotropic hypogonadism (IHH) is a clinical disorder characterized by delayed sexual development and inappropriately low gonadotropin and sex steroid levels in the absence of any anatomical or functional abnormalities of the hypothalamic-pituitary axis (160). Patients with IHH exhibit a wide spectrum of phenotypes, ranging from partial to complete hypogonadism even among affected kindred. In addition, this disorder is genetically heterogeneous and may be sporadic or familial. Mutations of two distinct genes located at the short arm of the X chromosome, KAL-1 and DAX-1, are responsible for the X-linked forms of IHH, which are accompanied by anosmia and adrenal insufficiency, respectively (161, 162, 163). In contrast, mutations of the GnRH receptor gene cause IHH without anosmia or adrenal failure and are responsible for autosomal inheritance of the disorder. To date, a total of 15 naturally occurring mutations have been identified along the entire sequence of the human GnRH receptor gene. Of these, one is a truncation mutant, nine are compound heterozygotes (164, 165, 166, 167, 168, 169, 170, 171, 172), and five are compound homozygotes (168, 173, 174). Functional studies in heterologous cell systems demonstrated that the naturally occurring GnRH receptor mutants have impaired cellular expression, ligand binding, and/or signal transduction such that 10 of them are totally nonfunctional (E90K, A129D, R139H, S168R, A171T, C200Y, S217R, L266R, C279Y, and L314X), whereas others retain a modest degree of receptor function (N10K, T32I, Q106R, R262Q, and Y284C). However, there are emerging data suggesting that misrouting of these mutant receptors contributes to the molecular etiology of normosomic, adrenal-sufficient IHH (175, 176, 177).
|
IV. Transcriptional Regulation of Human Type I GnRH Receptor Gene |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
). Using T3-1 cells as a model, the proximal 173-bp flanking region was found to be important for directing GnRH receptor gene expression in gonadotropes (178). This regulatory region contains two putative gonadotrope-specific elements (GSEs) with the core sequence 5'-TGA/TCC-3' at –143/–135 and –13/–5. Such regulatory elements have been shown to confer cell-specific expression of the glycoprotein hormone -subunit (179, 180) and LHß (181) genes in pituitary gonadotropes. Site-directed mutagenesis revealed that the upstream GSE (i.e., at –143/–135) is essential for gonadotrope-specific transcription of the GnRH receptor gene because mutation of this element selectively impairs the promoter function in T3-1 cells (178). EMSAs indicated that the orphan nuclear receptor steroidogenic factor-1 (SF-1) binds specifically to the upstream GSE, of which the second, fifth, sixth, and the ninth nucleotides are crucial for the interaction (178). The functional significance of SF-1 in regulating human GnRH receptor gene transcription in gonadotropes is confirmed by the findings that overexpression of sense and antisense SF-1 mRNAs can stimulate and repress the native promoter, respectively (178).
|
T3-1 cells (184). Despite sharing a high degree of homology with the mouse sequence, the human gene possesses neither the AP-1 nor the GRAS site in the corresponding positions along the proximal promoter region (178). These observations thus indicate that differential transcriptional apparatus may be involved in gonadotrope-specific expression of the human and rodent GnRH receptor genes. Many studies suggest that tissue-specific gene expression can be mediated via differential promoter usage in various cell types (185, 186, 187). Accordingly, Cheng et al. (188) have identified a novel human GnRH receptor promoter residing between –1737 and –1346, which is highly active in JEG-3 and extravillous trophoblast cells. The usage of this distal promoter is supported by the identification of five transcription start sites at –1629, –1608, –1416, –1391, and –1379 in the placental cells. Four putative cis-acting regulatory motifs termed human (h) hGR-Oct-1 (–1718/–1711), hGR-CRE (–1650/–1642), hGR-GATA (–1603/–1598), and hGR-AP-1 (–1519/–1513), which can interact specifically with transcription factors Oct-1, CRE-binding protein (CREB), GATA-2, GATA-3, and c-Jun/c-Fos heterodimer, were identified in the distal promoter (188). Mutational analysis indicated that the hGR-Oct-1 and hGR-AP-1 motifs act in a ubiquitous manner. Conversely, the hGR-CRE and hGR-GATA motifs appear to play a role in placenta-specific gene transcription because mutations of these elements result in a selective loss of promoter activity in JEG-3 cells (188).
Using a similar deletion approach, we have identified a new upstream GnRH receptor promoter that is primarily used by human GL cells (189). This novel promoter resides between –1300 and –1018 and contains two putative CCAAT/enhancer binding protein motifs and one GATA motif. The usage of this promoter in the GL cells is confirmed by the detection of a major transcription start site at –769, which is shortly downstream of a canonical TATA and CAAT box (189). Site-directed mutagenesis revealed that the CCAAT/enhancer binding protein and GATA binding sites work cooperatively to regulate the GnRH receptor promoter in the GL cells because simultaneous mutations of all these elements are required to cause a drastic abolishment of promoter function (189). Most importantly, these observations strengthen the notion that tissue-specific expression of the human GnRH receptor gene is mediated, at least partly, via differential promoter usage in various cell types.
B. Transcriptional regulation by GnRH-I
Previous studies on homologous activation of the mouse GnRH receptor promoter in T3-1 cells have revealed an integral role of a consensus AP-1 motif as well as the PKC and ERK1/2 signaling pathways (190, 191). Interestingly, this GnRH-I-stimulated effect can be augmented by activin A pretreatment, which is inhibited by follistatin (192). Deletion analysis of the mouse GnRH receptor promoter indicated that the region between –387 and –308, which contains two overlapping cis-acting regulatory motifs (the GRAS at –329/–318 and a SMAD-binding element at –331/–324), is responsible for the augmented response (192, 193). Competitive EMSAs showed that AP-1 and SMAD protein complexes bind respectively to –327/–322 and –329/–328, and disruption of either motif can eliminate both the GnRH-I and activin A responsiveness of the mouse GnRH receptor promoter (193). The functional significance of the SMAD-binding element is further supported by the observation that overexpression of SMAD2 and SMAD3 along with SMAD4 can increase the transcription of the GnRH receptor gene (192).
Transcriptional regulation of the human GnRH receptor gene by GnRH-I has also been investigated. In T3-1 cells, continuous administration of [D-Ala6]-GnRH-I represses the human GnRH receptor promoter in a dose- and time-dependent manner via a PKC-dependent pathway (194). Subsequent experiments indicated that a 248-bp region between –1018 and –771 is sufficient for mediating the suppression and that mutation of an AP-1-like motif (–1000/–994) can abolish the sensitivity of the promoter to both the GnRH-I analog and phorbol ester (194). EMSAs revealed that the AP-1-like motif binds c-Jun homodimer under nonstimulated conditions. However, an additional complex that is recognized by both anti-c-Jun and anti-c-Fos is formed when nuclear extracts from GnRH-I-stimulated cells are used (194). Therefore, it is apparent that homologous repression of the human GnRH receptor promoter may involve induction of c-Fos DNA binding activity at the AP-1-like site. Significantly, this GnRH-I-mediated down-regulation of GnRH receptor gene transcription may serve as a putative mechanism for pituitary desensitization to prolonged ligand stimulation. However, it is important to note that under conditions that can produce the maximal stimulatory response of the rodent GnRH receptor promoter (190, 191), a significant inhibition is observed for the human counterpart (194). These results thus further highlight the potential existence of species-specific mechanisms in transcriptional regulation among the GnRH receptor genes.
C. Transcriptional regulation by the cAMP-dependent signal transduction pathway
The cAMP signaling pathway is known to enhance the responsiveness of gonadotropes to GnRH-I by up-regulating GnRH receptor expression (195, 196, 197, 198). Moreover, an increase in GnRH receptor mRNA levels has been demonstrated in placental cells after forskolin treatment (107). These stimulatory effects are thought to occur at the transcriptional level because forskolin can activate the human GnRH receptor promoter in a dose- and time-dependent manner in T3-1 and JEG-3 cells (152, 199). Similar responses are also observed with other physiological regulators that activate the cAMP-dependent signaling pathway (152, 199). Using progressive deletion analysis, the forskolin response area has been mapped to a region between –577 and –167, within which two potential AP-1/CREB binding sites termed hGR-AP/CRE-1 (–569/–562) and hGR-AP/CRE-2 (–341/–334) partly contributing to the forskolin effect were identified (152, 199). Although both the hGR-AP/CRE-1 and hGR-AP/CRE-2 sites interact specifically with CREB in forskolin-stimulated cells, a differential binding of transcription factors to hGR-AP/CRE-2 was observed such that the motif interacts primarily with AP-1 when nuclear extracts from nonstimulated T3-1 cells were used (152, 199).
D. Transcriptional regulation by gonadal steroid hormones
A study from Cheng et al. (148) has shown that progesterone can repress the human GnRH receptor promoter in T3-1 cells in a dose- and time-dependent fashion. In contrast, the steroid exerts a stimulatory effect in JEG-3 cells as blockade of endogenous progesterone production silences the GnRH receptor promoter. Deletion and mutational analysis indicated that an imperfect progesterone-response element at –536/–522 is responsible for mediating the responses in both the T3-1 and JEG-3 cells (148). Using EMSAs, a specific binding of PRs to the response element has been demonstrated (148), thus indicating a direct involvement of the nuclear receptors in conferring the transcriptional effects. Overexpression of the two human PR isoforms (PR-A and PR-B) indicated that PR-B plays a predominant role in mediating the down-regulatory effect in T3-1 cells. On the contrary, a differential action of PR-A and PR-B is observed in JEG-3 cells such that PR-B stimulates whereas PR-A suppresses the GnRH receptor promoter (148). In concert with these findings, PR-B has been identified as the major PR subtype in the placental cells (148), thus supporting a positive role of progesterone in controlling human GnRH receptor gene transcription in the placenta.
The mechanism by which estrogen regulates GnRH receptor gene transcription in ovarian and breast cancer cells has only been recently elucidated. In these cells, E2 can repress the human GnRH receptor promoter via a nonconsensus AP-1 motif and ER, of which the DNA-binding domain and the ligand-binding domain are indispensable for the repression (141). Interestingly, the same cis-acting motif is also important for both the basal activity as well as phorbol 12-myristate 13-acetate (PMA) responsiveness of the GnRH receptor promoter. Multiple transcription factors including c-Jun and c-Fos, but not ER, bind to the AP-1 site, indicating that the E2-induced repression occurs independently of direct ER binding to the promoter (141). This observation may be supported by the fact that no estrogen-response elements can be identified in GnRH receptor 5'-flanking regions sequenced so far (90, 92, 93, 94). Intriguingly, the repressive effect of E2 on the human GnRH receptor promoter can be antagonized by cotreatment with PMA, which stimulates c-Jun phosphorylation at serine 63 (141), a process prerequisite for recruitment of the transcriptional coactivator CREB-binding protein (200, 201). Concomitantly, overexpression of the coactivator can reverse the suppression in a dose-dependent manner (141), suggesting that E2-bound ER represses human GnRH receptor gene transcription via an indirect mechanism involving competition for a limiting amount of CREB-binding protein.
E. Transcriptional repression
Several reports from our laboratory have consistently suggested the presence of a very strong negative regulatory element (NRE) (–1017/–771) in the human GnRH receptor 5'-flanking region (188, 189, 202). Although this repressive element can work ubiquitously in a heterologous environment, its silencing activity is dependent on its orientation relative to the target promoter sequence (203). Progressive deletion analysis revealed that most of the NRE silencing effect resides in an evolutionarily conserved octamer sequence (–1017/–1009), which can suppress the native promoter activity by almost 90% in JEG-3 cells (203). Results from EMSAs and Southwestern blot analysis have convincingly shown that the ubiquitously expressed POU homeodomain transcription factor Oct-1 is the repressor protein binding to the powerful NRE (203).
It is important to point out that the mouse gonadotrope-derived T3-1 cell line is employed as the model system, in most if not all studies, to investigate the transcriptional regulation of the human GnRH receptor gene in the pituitary (148, 178, 188, 194, 202, 203). Such cross-species studies may not be capable of accurately reflecting the mechanisms operating in the human counterpart, and the observable differences in transcriptional control between the human and rodent GnRH receptor genes may be due to the absence of certain species-specific transcription factor(s) in the mouse gonadotropes.
|
V. Signal Transduction Mechanism of the Mammalian Type I GnRH Receptor |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
q/11 in heterologous Chinese hamster ovary-K1 and COS-7 cells (204) but to Gs in the placenta (107). In contrast, others have reported that the receptor couples selectively to Gi in some reproductive tract tumors and their derived cell lines (2, 118, 123, 205). Interestingly, there is evidence showing that the rodent GnRH receptor couples to multiple G proteins in a single cell type (206, 207, 208). In GT1-7 neurons, high GnRH-I analog concentrations induce a ligand-dependent switch of G protein coupling from Gs to Gi, the activation of which inhibits episodic GnRH-I release, possibly via regulation of membrane ion channels (208). Such negative feedback action serves as an autocrine mechanism for the genesis of pulsatile GnRH-I secretion that is essential for the maintenance of normal gonadotropin release profiles and gonadal functions.
B. MAPKs
The MAPKs play an integral role in GPCR-mediated intracellular signaling (209, 210). In mouse pituitary gonadotropes, the GnRH receptor activates four MAPK cascades including the ERK1/2, the c-Jun amino-terminal kinase (JNK), the p38 MAPK, and the big MAPK (BMK1/ERK5) (211, 212, 213) to various extents by a PKC-, Ca2+-, and tyrosine kinase-dependent mechanism (214, 215). For ERK1/2, the activation is primarily PKC dependent and involves two distinct pathways that converge at Raf-1 (216, 217, 218). Also, this process requires Ca2+ elevation (216, 219) and sublocalization of the receptor to low-density membrane microdomains (220). On the other hand, activation of JNK is highly dependent on cytosolic Ca2+ and is mediated via a pathway requiring sequential stimulation of PKC, c-Src, CDC42/Rac1, and MAPK kinase (MEK)K1 (221, 222). Although the signaling pathways leading to p38 MAPK and BMK1 activation are less clear, it appears that the activation involves a PKC-dependent cascade (214, 218, 223). Stimulation of MAPK cascades by the GnRH receptor has also been investigated in other cell types, in which the intracellular mechanisms mainly involve transactivation of the epidermal growth factor receptor (224, 225).
C. Receptor desensitization and internalization
Activation of GPCRs is typically followed by their desensitization and internalization, and these processes involve rapid agonist-induced receptor phosphorylation by both second messenger-dependent protein kinases and G protein-coupled receptor kinases (81). Because the serine and threonine residues that are phosphorylated by G protein-coupled receptor kinases are often located in the carboxyl-terminal tail (81), which is uniquely absent in the mammalian GnRH receptor, a number of studies have revealed that the tailless GnRH receptor neither undergoes rapid homologous desensitization nor exhibits agonist-induced receptor phosphorylation (226, 227, 228, 229, 230). In addition, the receptor internalizes slowly via clathrin-coated vesicles, and this process occurs independently of ß-arrestin and dynamin (226, 229, 230, 231). The unusual resistance of the mammalian GnRH receptor to desensitization may be essential for mediating its direct antiproliferative effect (will be discussed in detail below), which requires sustained ligand stimulation and is shown to be ineffective by receptors having a carboxyl-terminal tail (230, 232).
|
VI. Biological Actions of GnRH-I and GnRH-II in Humans |
|---|
|
TopAbstractI. IntroductionII. GnRH Isoforms in...III. Molecular Characterization...IV. Transcriptional Regulation...V. Signal Transduction Mechanism...VI. Biological Actions of...VII. Type II GnRH...VIII. ConclusionsReferences |
|---|
). In tumors derived from various reproductive tissues, there is solid evidence showing that the GnRH receptor couples to a pertussis toxin-sensitive G protein (most probably Gi) and mediates its biological effects via pathways that are distinct from the classical cascade operated in gonadotropes. Extrapituitary actions elicited by GnRH-II have also been demonstrated in certain human peripheral tissues.
|
-subunit noncovalently bound to a specific ß-subunit (FSHß and LHß) (233, 234). Activation of the human -subunit gene transcription by GnRH-I is primarily Ca2+ dependent. Also, this process can be augmented by PKC and requires ERK1/2 and c-Src (235, 236). Although the nonreceptor tyrosine kinase c-Src has been demonstrated to mediate ERK stimulation by the GnRH receptor (217), the ERK and c-Src-response areas are located at different regions on the -subunit promoter (236), indicating that c-Src contribution is independent of ERK activation. The role of PKC, Ca2+, and MAPK signaling cascades in mediating GnRH-I stimulation of LHß gene transcription has not been clearly addressed. Although it was shown that the activation is Ca2+ dependent (237), others reported that PKC is mainly responsible for the effect (238, 239). Similarly, whereas it was found that both PKC and ERK1/2 are required for the stimulation (240), others demonstrated an essential role of JNK (216). These discrepancies are likely due to different experimental paradigms such as the use of different cell models or promoters from different species. Several cis-acting regulatory elements such as Sp1, CArG, and early growth response 1 transcription factor binding sites have been identified in the rat LHß promoter (241, 242, 243). It is suggested that these motifs may act in concert with different signal transduction cascades to confer GnRH-I sensitivity to the LHß promoter (237, 240, 244, 245).
Activation of FSHß gene transcription by GnRH-I requires the Ca2+, PKC, and MAPK signaling pathways in mouse LßT2 gonadotropes (246). The GnRH-I responsiveness of the ovine FSHß promoter involves at least two elements at the distal region, in association with one or several motifs within the proximal promoter sequence (246). Although previous studies have shown that two downstream AP-1-like elements are important for mediating the GnRH-I response in heterologous HeLa cells (247), these motifs are not functionally equivalent in the gonadotropes (246). On the other hand, Pernasetti et al. (248) have shown that GnRH-I stimulation of ß-subunit gene transcription in LßT2 cells is also dependent on an endogenous activin autocrine loop as follistatin treatment can block the GnRH-I induction.
An essential role of intracellular Ca2+ in mediating GnRH-I-stimulated gonadotropin secretion has been established (77, 249, 250). In contrast, the initial phase of gonadotropin release is apparently independent of extracellular Ca2+ (249, 251, 252). The role of PKC activation in mediating the GnRH-I effect is less clear (77, 250). Whereas phorbol ester can stimulate LH secretion, GnRH-I-induced LH release is impaired but not abolished in PKC-depleted gonadotropes (253, 254). These observations thus indicate that PKC activation is not an absolute requirement for exocytosis. It is believed that PKC participates in the control of gonadotropin secretion through its actions on cytoskeleton elements, Rab proteins, and other elements involved in the exocytotic process (255, 256, 257). Although there is no evidence for rapid desensitization of the mammalian GnRH receptor, sustained ligand exposure causes down-regulation of inositol (1, 4, 5) triphosphate receptor in T3-1 cells (258), leading to a marked reduction in GnRH-I-stimulated Ca2+ mobilization and gonadotropin secretion (259, 260). This form of desensitization underlies the basis of hypothalamic-pituitary-gonadal axis suppression that is exploited in the major clinical applications of GnRH-I analogs (11).
Recently, GnRH-II has also been shown to be capable of stimulating LH and FSH release both in vivo (261) and in cultured pituitary cells (262). This stimulation is mediated via activation of the type I GnRH receptor because the effects can be blocked by antide (261, 262).
B. Ovarian steroidogenesis
There is a general consensus that GnRH-I analog treatment in vivo or in vitro exerts an inhibitory effect on gonadotropin-regulated steroidogenesis in human GL cells (20, 263, 264, 265, 266, 267). Exposure of the steroidogenic cells to [D-Ala6]-GnRH-I rapidly activates ERK1/2 and causes a drastic increase in c-Fos mRNA levels (4). This GnRH-I-induced ERK activation is mediated via Gq/11 and requires PKC because the effect can be mimicked by PMA and abolished by the PKC inhibitor GF109203X (4). Interestingly, pretreatment of GL cells with the MEK inhibitor PD98059 completely abrogates the down-regulatory effect of GnRH-I on steroidogenesis (267), suggesting that a PKC- and ERK-dependent cascade is involved in mediating the antisteroidogenic response.
Recent findings from our laboratory have revealed that treatment of human GL cells with GnRH-II or its analog in vitro can also suppress hCG-stimulated progesterone production (20). This inhibition can be blocked by antide, indicating the mediation via the type I receptor (20). Similar to the effects produced by GnRH-I, GnRH-II does not interfere with hCG-stimulated cAMP generation. Instead, these hormones down-regulate the steady-state mRNA levels of both the FSH and LH receptors in the steroidogenic cells (20). These observations thus support a notion that GnRH-I and GnRH-II exert their antigonadotropic effects at the receptor level but not at the cAMP level.
C. Cell proliferation
The role of GnRH-I as a negative autocrine growth factor has been well reported in cell lines derived from human malignant tumors including those of the ovary, endometrium, breast, prostate gland, and melanoma cells (2, 268, 269, 270, 271). It is generally thought that this antiproliferative action is mediated via high-affinity GnRH-I binding sites, as supported by the notion that the nucleotide sequence of the GnRH receptor is identical in tumor and pituitary cells (112, 118, 202). Nonetheless, in some systems, high doses of GnRH-I analogs (1–10 µM) are sometimes needed to demonstrate a significant but modest growth-inhibitory response (6, 21, 131, 272, 273). Although the intracellular mechanisms mediating the antiproliferative effect of GnRH-I analogs are not fully understood, several lines of evidence have suggested a role of the ERK1/2 signaling pathway. In ovarian carcinoma Caov-3 cells, the GnRH-I analog leuprolide induces phosphorylation of son of sevenless and Shc and causes a sustained stimulation of the MEK-ERK cascade (272). This process is mediated via a pertussis toxin-sensitive G protein and the Gß
dimer and occurs independently of PKC and extracellular Ca2+. Consequently, the prolonged ERK activation leads to hypophosphorylation of the retinoblastoma protein (272), a process known to prevent cell cycle progression from G1 to S phase. Similar observations have also been reported in other gynecological cancer cell lines, in which GnRH-I analog treatment in vitro blocks cell cycle transition and decreases DNA synthesis (274, 275). These growth-inhibitory effects of GnRH-I analogs may be mediated via stimulation of the DNA binding activity of JunD (275), which has been suggested as a negative regulator of cell proliferation (276).
The involvement of the ERK1/2 cascade in mediating the antitumor effect of GnRH-I analogs is further supported by the observation that the analogs can antagonize growth factor-induced mitogenic signaling via coupling to Gi proteins (2, 118, 123, 205). In primary ovarian and endometrial carcinomas as well as certain cancer cell lines, treatment with GnRH-I analogs in vitro activates phosphotyrosine phosphatase and causes a substantial loss of phosphotyrosines from cellular proteins such as the epidermal growth factor receptor (118, 277, 278, 279, 280). These effects are associated with a significant reduction in growth factor-induced ERK1/2 activation, c-Fos gene expression, matrix metalloproteinase (MMP) secretion, and cell proliferation (280, 281, 282). In some prostate cancer cells, GnRH-I analog treatment may reduce cellular tyrosine phosphorylation and proliferation index via down-regulation of growth factor receptor expression (279, 283). Thus, depending on the cell context, GnRH-I analogs may attenuate the mitogenic action of growth factors and suppress the ERK cascade to mediate their antitumor effects.
In addition to MAPK regulation, several potential mechanisms have been suggested to account for the growth-inhibitory effect of GnRH-I analogs in cancer cells. These putative mechanisms include inhibition of phosphatidylinositol kinase activity (284) and stimulation of lysophosphatidic acid hydrolysis (285), as well as down-regulation of telomerase reverse transcriptase, acidic ribosomal phosphoprotein, and prostate-specific antigen expression (51, 286, 287).
There are many reports showing that GnRH-I analogs can also inhibit cell proliferation in human uterine leiomyoma and endometriosis. Although GnRH-I-induced leiomyoma regression appears to occur predominantly through inhibition of gonadotropins and gonadal steroids (11), the suppression may involve alteration of growth factor (6, 288), cytokine (289), cell cycle regulator (43), and steroid hormone receptor (290) expression. Similarly, in endometriotic stromal cells, GnRH-I analog treatment in vivo reduces tumor necrosis factor -induced nuclear factor-
B (NF-B) activation and interleukin-8 expression (10, 291), which has been reported to promote endometriosis (292, 293).
The role of GnRH-II as an autocrine growth inhibitor has also been demonstrated. Like GnRH-I, treatment with GnRH-II in vitro inhibits the proliferation of both nontumorigenic and tumorigenic ovarian surface epithelial cells in a dose-dependent manner (19). In accord with the presence of two types of GnRH binding sites (
