| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute of Reproductive Medicine of the University, Domagkstraße 11, D-48129 Münster, Germany
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
After two decades of investigations using classic biochemical approaches, the FSH receptor cDNA was finally cloned in 1990, the last in the group of closely related receptors for the glycoprotein hormones (6). Thereafter the first mutations were described, with impressive impact on the reproductive phenotype (7, 8). The new knowledge emerging from naturally occurring mutations and from in vitro molecular work provides important new insights into FSH physiology and pathophysiology. Unlike the cognate LH and TSH receptor, the functional properties of which have been recently reviewed by several authors (9, 10, 11, 12, 13), the large body of FSH receptor research has not yet been comprehensively considered. Feeling the need for an integrated view of the relevant knowledge about biochemical and molecular properties of the FSH receptor at this stage, we compiled this article with the aim of providing both a state-of-the-art review and a stimulating springboard for further pertinent research.
|
II. Biochemical Properties of the FSH Receptor: A Historical Prelude |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
).
|
After the initial attempts to isolate the rat and calf FSH receptor (17, 18, 22, 37), two classes of FSH-binding sites were detected, with high affinity and low capacity and with low affinity and high capacity, respectively (14, 16, 17, 37, 38). The low-affinity component, however, turned out to be artifactual (38), and only one class of high-affinity binding sites can be demonstrated in purified preparations (22). The specific binding of FSH is rapid at physiological temperatures and reaches saturation within 4 h in all systems investigated (14, 16, 17, 21, 22, 29, 30, 31, 39).
FSH receptors are particularly abundant in the immature bovine testis, and the receptor concentration is higher in the bovine calf compared with the mature rat and human testis (22, 29, 40). The testicular content of FSH-binding sites increases with age in the rat and the bovine, where the appearance of FSH receptors precedes the increase in plasma FSH, testicular growth, and the increase of LH receptor concentration at puberty (41). There is very little, if any, species specificity in the FSH-receptor interaction, which shows similar characteristics in homologous as well as in heterologous systems (32, 37).
Concerning the precise localization of the FSH receptor, autoradiographic studies showed that labeled FSH was selectively localized on the surface of the Sertoli cells, outside the tight junctions (19). After hyperosmotic fixation, which produces preferential shrinkage of the cells of the basal compartment, FSH-binding sites were also evident on spermatogonia (42), but the spermatogonial localization was never confirmed (43). A recent immunocytochemical study with an antibody directed against the human FSH receptor showed uniform labeling of granulosa cell membranes and of the basal pole of Sertoli cells around the spermatogonia (44). Interstitial macrophages were shown to bind and accumulate FSH (19, 33) and to respond to FSH administration (45, 46, 47). The endothelial cells of the small vessels in the interstitial space were also stained (44). Whether these localizations are related to the transport of the large glycoprotein hormone to the target site remains to be determined (48, 49, 50).
Earlier experiments showed that the FSH receptor is a glycoprotein (17, 21, 30) and that FSH-receptor interaction is partially dependent on the presence of phospholipids when membrane preparations are used (17, 22, 30, 51). In fact, G protein-coupled receptors are anchored to plasma membranes by fatty acylation or protein lipidation, which stabilizes protein conformation and possibly plays some role in signal transduction (52, 53). Moreover, FSH receptor binding is dependent on the integrity of disulfide bonds (17, 30), which stabilize the receptor conformation, whereas a postulated role in maintaining a subunit structure (54) has not been confirmed.
The purification studies yielded widely variable models of the size and structure of the receptor (22, 24, 28, 34, 54, 55, 56). Even after the cloned cDNA predicted a single peptide chain with a molecular mass of 75 kDa (6), the mature FSH receptor from rat Sertoli cell membranes is occasionally found to have a much larger size, and the controversy about the receptor structure is still not definitely resolved (57).
The involvement of guanine nucleotides in FSH receptor function had already become evident at a time when G proteins were not yet known (17), and the stable solubilization of the bovine FSH receptor demonstrated its physical and functional coupling to Gs protein (26, 27). While indirect effects on the phosphoinositide pathway are possible (54), the FSH-dependent increase in intracellular cAMP was soon recognized to be the main signal transduction mechanism in Sertoli and granulosa cells (58) and FSH induces receptor down-regulation (59, 60, 61, 62, 63). The most recent knowledge about receptor coupling, signal transduction, and receptor desensitization derives, however, from studies with recombinant receptors transfected in cell lines and will be reviewed in Section VII.
|
III. Molecular Structure of the FSH Receptor |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
In 1989 the rat LH receptor was cloned by Seeburg’s and Segaloff’s groups using a classic molecular biology approach for the isolation of target cDNA as follows (65). First, Segaloff’s group purified rat LH receptor protein by affinity chromatography. Based on the aa sequence of the N-terminal side, degenerate oligonucleotide primers allowed the PCR amplification of a specific cDNA product, which was then used to isolate the corresponding full-length cDNA from an ovarian cDNA library. Sequence analysis of the cDNA revealed that the rat LH receptor is a single potentially glycosylated protein containing an unusually large, predicted extracellular domain.
The similarity of the cloned LH receptor and TSH receptor and the
observation that all three glycoprotein hormones act on their
respective receptors via the cAMP pathway led to the assumption that
the structural design of the FSH receptor should display similar
characteristics. Since Sertoli cells are the sole specific target of
FSH action and do not bind LH, cDNA probes corresponding to selected
regions of the LH receptor were used to screen a rat Sertoli cell cDNA
library. The isolated cDNA had an approximate size of 2.3 kb, and its
specificity was investigated by functional expression studies. Human
embryonic kidney cells transfected with the putative receptor cDNA
displayed an FSH-dependent and saturable increase in intracellular
cAMP. In contrast, no cAMP stimulation was observed when using human
(h) CG or hTSH, indicating the successful cloning of the rat FSH
receptor (6) (Fig. 1).
|
Several sequences of the human FSH receptor have been reported, differing at several nucleotide positions and resulting, in some cases, in aa substitutions (77, 78, 79). Some of these discrepancies were revised (77) but others, such as the presence of amino acid Thr or Ala at position 307 and Ser or Asn at position 680, are not due to inaccurate sequencing. Since the different cDNA sequences were generated from testicular (67, 78) or ovarian tissue (66), it was originally postulated that these differences might be related to sex-specific changes. It is now clear that the observed substitutions in the FSH receptor cDNA are due to two polymorphic sites in the FSH receptor gene (see Section X).
B. Predicted primary structure of the FSH receptor
The predicted human FSH receptor protein (66, 67, 68) is composed of
695 aa (692 aa in the rat and 691 aa in the equine), including the
first 17 aa, which encode a hydrophobic signal peptide (Fig. 2: aa numbering maintained thoughout this
article). Therefore, the mature protein is likely to consist of 678 aa
(675 aa in the rat and 677 in the equine). Depending on the species,
the calculated molecular mass based on the cDNA sequence for the mature
receptor protein ranges between 75 and 76.5 kDa. Further
characterization of the aa sequence and hydropathy plot analysis
revealed that the FSH receptor consists of a huge hydrophilic domain
followed by hydrophobic segments spanning the membrane seven times,
with a length of 21–24 aa. At the C terminus the sequence predicts a
highly basic cytosolic segment.
|
).
1. Extracellular domain. The extracellular domain of the FSH
receptor displays several significant primary and secondary structure
features. It is composed of several imperfectly replicated units of
approximately 24 residues each. This characteristic motif is also
present in the LH and in the TSH receptors and, in part, even in the
so-called remainder forms of the recently cloned glycoprotein hormone
receptor ancestor (80, 81, 82). Similar motifs, termed leucine-rich repeats
(LRR), are found in proteins involved in cell-specific adhesion and
protein-protein interaction in species extending from yeast to man
(Refs. 65 and 83 and references therein). The crystal structure of the
porcine ribonuclease inhibitor, containing LRR, has been resolved
recently (83). The individual repeats constitute structural units of
alternating ß-sheets and 
-helices, probably occurring in the
gonadotropin receptors as well. The nonglobular shape of the structure
and the exposed face of the parallel ß-sheet may explain the
involvement of LRR in strong protein-protein interactions (83) (Fig. 1).
Alignment of exons 2–9 (see also Section IV) in
the extracellular domain of the FSH receptor reveals at least 10
imperfect LRR motifs (Fig. 3). The
conserved positions are occupied by Ile, Leu, Val, Ala, and Phe, aa
belonging to the aliphatic group. The LRR pattern is highly conserved
in exons 2–8, less so in exon 9. Exon 1 and the C-terminal part of the
extracellular domain of the FSH receptor, encoded by exon 10, do not
conform to the consensus motif. The consensus sequence of the LRR in
the FSH receptor is homologous to the motif found in the LH receptor
(83) and in the TSH receptor (12). Due to their amphiphatic nature, the
repeats might confer the dual property of interacting both with the
hormone and the transmembrane domain (84). Within the FSH receptor,
repeats 1–10 participate in FSH binding (85), and the binding
specificity is probably localized between LRR 5 and LRR 10 (see
Section VII.B) (86).
|
), whereas the
others cannot be aligned with the remaining six potential glycosylation
sites of the LH receptor and five potential glycosylation sites of the
TSH receptor. Although it was suggested that the glycosylation pattern
of the FSH receptor might affect hormone binding, recent studies
indicate that glycosylation is rather required for proper folding of
the receptor protein and trafficking to the membrane (see
Section VIII) (85).
2. Transmembrane domain. The structural motif of the
heptahelical or serpentine transmembrane domain is typical of members
belonging to the superfamily of G protein-coupled receptors (87, 88)
(Fig. 1). In each member of this group the motif is characterized by
seven hydrophobic stretches of 20–25 aa predicted to form
transmembrane -helices, connected by alternating extracellular and
intracellular loops. Similar to members of this receptor family, the
FSH receptor contains two highly conserved Cys residues (positions 442
and 517) in the first and second extracellular loop, predicted to form
an intramolecular disulfide bridge (5), which constrains the
conformation of the protein (6). The highly conserved Asp-Arg-Tyr
triplet motif (5), believed to play a central role in the interaction
between receptor and G protein, is present in the FSH receptor in a
modified version in which Asp is replaced by Glu (positions 466–468).
The same substitution is observed in the corresponding triplet motif of
the LH and TSH receptor.
Comparison of the transmembrane domains of the FSH, LH, and TSH receptor reveals that transmembrane domains 2, 3, and 4 are highly conserved, whereas conservation in the other four transmembrane regions is lower. Proline residues, which may be necessary for proper insertion of the protein into the membrane, are homologous in the fourth, sixth, and seventh transmembrane segment of all glycoprotein hormone receptors. Between the cytoplasmic loops the highest aa homology can be noted in the first loop. The third cytoplasmic loop is significantly shorter compared with other members of the G protein-coupled receptor family, and the homology between the FSH, LH, and TSH receptor in this region is low.
3. C-terminal domain (aa 631–695). The intracellular domain of the glycoprotein hormone receptors displays some homology only in the N-terminal portion (aa 631–659 of the FSH receptor). Moreover, within different species the C-terminal domain of the FSH receptor is quite heterogenous. For example, there is a gap of two aa in the rat FSH receptor. The domain is rich in serine and threonine residues, which are potential phosphorylation sites (see Section VIII), but a typical consensus recognition site for the cAMP-dependent kinase (Arg-Arg-X-Ser/Thr) is lacking. The cysteine residues might be palmitoylated, thereby serving as an additional membrane anchor of the cytoplasmatic receptor tail. This fourth intracellular loop might be relevant for receptor coupling and regulation of signal transduction. Two of these cysteine residues (positions 646 and 672) are conserved within the different species, whereas the human and equine FSH receptors possess an additional cysteine, at position 644 in the human receptor and at position 671 in the equine receptor, respectively.
C. Molecular mass of the FSH receptor
The predicted glycoprotein nature of the FSH receptor, with the
inherent possibility of glycosylation variants, might be responsible
for the differences in its molecular mass reported in the literature.
Based on the ORF of the cDNA, a molecular mass of approximately 75 kDa
can be calculated. However, several groups have demonstrated precursors
of the glycosylated protein or isoforms of the mature receptor with
lower or higher molecular mass (44, 57, 89, 90).
Taken together, the reports on mass determination indicate that the
mature, glycosylated, recombinant FSH receptor is a protein of
approximately 80 kDa (78, 85). Furthermore, binding of FSH does not
require prior receptor oligomerization, although aggregation might be
necessary for stability and/or signal transduction. Cleavage of the
related LH and TSH receptor proteins, resulting in lower molecular mass
forms, has been described (11, 91). The FSH receptor sequence displays
five putative dibasic cleavage sites at positions 242, 253, 282, 572,
and 634, respectively, conserved within the different species. Some of
them are also found in the LH and in the TSH receptor at corresponding
sites (Fig. 2). To date, however, the studies on tissues naturally
expressing the FSH receptor did not reveal any evidence for spontaneous
cleavage of the FSH receptor protein (44).
|
IV. The FSH Receptor Gene |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
The large extracellular domain of the glycoprotein hormone receptors is a unique feature within the G protein-coupled receptor family. The similar genomic arrangement and the nearly identical exon/intron boundaries of the three receptor genes, together with the proximity of the two gonadotropin receptor genes, indicate a common ancestor. This ancestral gene could have evolved first by chromosomal duplication, followed by duplication of the gene (94). Recent studies on genome evolution have shown that human FSH, LH, and TSH receptors are located in a group of chromosomes with extensive paralogous connections, i.e., containing genes arising from gene duplication and subsequent divergence (97). A detailed aa sequence analysis of the three glycoprotein hormone receptors reveals a closer sequence similiarity between the FSH and LH receptor than between either to the TSH receptor. This finding and the lack of a duplication locus for the TSH receptor have led to the hypothesis that the LH receptor and TSH receptor have evolved by chromosomal duplication. Further duplication of the FSH receptor/LH receptor locus and subsequent functional divergence would then have resulted in the two gonadotropin receptors present in mammals today (94).
Further evolutionary insights have recently been obtained from the identification of glycoprotein hormone receptor-like receptors in invertebrates. In the mollusc Lymnea stagnalis, a G protein-coupled receptor with a very large extracellular domain has been recently cloned (81). The N-terminal portion of the extracellular domain consists of several Cys-containing repeats, a motif present also in low-density lipoprotein receptors, whereas the second part of the extracellular domain contains six LRR. Thus, this receptor, in a phylogenetically very old species, might be a remainder of the common ancestor from which the genes encoding the mammalian glycoprotein hormone receptors have evolved through duplication of the LRR and removal of the Cys-containing repeats. Another G protein-coupled receptor, recently cloned from sea anemones, displays striking similarity to the glycoprotein hormone receptors. In fact, it possesses a huge extracellular domain, shows alternative splicing of the primary transcript, and encloses two introns with position and intron phase identical to those of introns 7 and 8 of the glycoprotein hormone receptor genes (80). These characteristic similarities have been shown also in a G protein-coupled receptor cloned from Drosophila melanogaster, which appears to be involved in developmental processes of insects (82). These invertebrate receptors might reflect steps in the evolutionary process of defining and remodeling the structure of glycoprotein hormone receptors.
B. Structure and organization of the FSH receptor gene
The structure and organization of the FSH receptor gene have been
investigated in humans and rats (98, 99). The FSH receptor gene is a
single-copy gene and spans a region of 54 kbp in the human and 84 kbp
in the rat, as judged from restriction analysis of genomic clones and
size determination of PCR products. It consists of 10 exons and nine
introns (Fig. 4). The extracellular
domain of the human receptor is encoded by nine exons ranging from
69–251 bp. The C-terminal part of the extracellular domain,
transmembrane and the intracellular domain, is encoded by exon 10 with
more than 1234 bp (99). Overall, the human gene encodes 695 aa,
including a signal peptide with 17 aa. The nine introns vary greatly in
their corresponding sizes from 108 bp for intron 7 to 15 kbp for intron
1. The exon-intron boundaries correspond to a canonical splice
consensus sequence conserved in all exons. The introns are in phase 2,
and the aa that resides at nearly each exon/intron junction is either
Leu or Ile.
|
With 70 kbp for the LH receptor (100), 60 kbp for the TSH receptor (101), and 54 kbp for the FSH receptor, the three human glycoprotein hormone receptor genes are huge. The FSH and TSH receptors consist of 10 exons, while the LH receptor has 11 exons. The similarity between the genes is high. The sizes of several exons of the extracellular domain are identical in the three genes. The other exons differ only by three bases; the exception is the additional exon 10 of the LH receptor, which is unique as it contains three putative N-linked glycosylation sites (102). Furthermore, the intronic sequences of the 5'-end of exon 11 of the LH receptor correspond to promoter and regulatory regions of the intronless genes of other G protein-coupled receptors (103, 104). Probably this last intron was lost during the evolution of the cognate FSH and TSH receptor, resulting in the structural arrangement of ten exons and nine introns.
C. The promoter of the FSH receptor gene
The FSH receptor gene expression is highly tissue-specific and
strictly dependent upon different hormonal stimulation. To elucidate
the regulatory mechanisms of the expression, several groups of
investigators tried to characterize the promoter response elements and
corresponding factors in the 5'-flanking region of the FSH receptor
gene. The putative promoter regions of the human, rat, and mouse FSH
receptor have been cloned, the sequences were analyzed, and different
promoter constructs were investigated in functional studies using
different cell types (98, 105, 106, 107). Using ribonuclease (RNAse)
protection assay or primer extension analysis, a major transcriptional
start site has been located at position -99 in the human (104) and at
-534 in the mouse (107) receptor, relative to the translational start
site. In the rat, two major transcriptional start sites at positions
-80 and -98 were found (98). In all species, additional, alternative,
less marked transcriptional initiation sites have been observed. The
transcriptional start sites of the human FSH receptor mRNA are
identical in the testis and ovary, thereby excluding the possibility
that the sex-specific regulation of gene expression makes use of
different transcriptional start sites (106).
The 5'-flanking regions of the different genes lack canonical TATA or CCAAT promoter elements. Furthermore, GC box motifs, binding sites for the promoter-specific transcription factor SP-1, are present in the LH receptor (100, 104) but not in the FSH receptor promoter. Searching for further transcription binding sites revealed an activator protein 1-binding site at position -214 in the rat FSH receptor gene (98). Although the treatment of cultured rat Sertoli cells with phorbol esters results in a decreased response of the cells to FSH (108), the importance of this activator protein 1-binding site is questionable since it is not present in the human or mouse promoter. Similarly, a consensus estrogen-responsive element found in the human promoter at positions -217 to -221 has no correspondent in the rat or mouse promoter (106). A described initiator region (InR), encompassing a transcriptional start site, is conserved only in mouse and rat, whereas an E box element, interacting with a family of basic helix-loop-helix transcription factors, is conserved in the promoter region of all species. Promoter studies in primary rat Sertoli cells or a mouse Sertoli cell line displayed lower activities when the E box was mutated (109). Based on the presence of a cAMP-regulatory element (CRE)-like element at around position -115 in the rat FSH-receptor promoter, Monaco et al. (110) proposed a role for the inducible cAMP early repressor (ICER) in the regulation of the FSH receptor gene expression. Functional studies in primary rat Sertoli cells showed that the cAMP responsive element modulator (CREM) isoform ICER increases rapidly upon FSH stimulation, indicating that it might be involved in the rapid down-regulation of the FSH receptor transcripts and long-term receptor desensitization. This repression was ascribed to the binding of ICER to the CRE-like sequence in the FSH receptor promoter, since ICER could repress expression of a transcriptional reporter gene containing this CRE-like site in transfected primary rat Sertoli cells. However, the CRE-like motif ATTAGTCA is present neither in the human nor in the mouse FSH receptor promoter, and other rat promoter studies could not demonstrate direct interactions between the CRE-like sequence and rat Sertoli nuclear proteins (109, 111).
The promoter activity was investigated by transfection studies using primary cells and cell lines in which a reporter gene expression vector was driven by different 5'-extensions of the FSH receptor promoter. The highest activity was obtained using constructs of the FSH receptor promoter from -1 to -286 bp relative to the translational start site (105, 106). The activity was markedly reduced when longer DNA constructs were used, indicating the presence of repressor elements. In all cases the FSH receptor promoter was constitutively active in the absence of hormone, and the basal activity of a construct ranging from -847 to +114 was stimulated 4-fold by (Bu)2cAMP treatment (110). The core promoter region could be allocated to the first 286 bp, a tract that includes the major transcriptional start sites and shows the highest homology among the 5'-flanking regions of the human, rat, and mouse FSH receptor (106). These findings indicate that repression and derepression of cis-acting elements upstream of the core promoter region, which mediate constitutive transcriptional activity, is a potential mechanism to modify expression of the FSH receptor promoter activity.
Human FSH receptor promoter activity was observed in cells naturally expressing the receptor, such as human granulosa cells and rat Sertoli cells, but also in the nonexpressing Chinese hamster ovary (CHO) cell line (106). Mouse promoter activity was detected in Sertoli cells but not in CHO cells (107), and rat promoter activity was detected in a mouse Sertoli cell line (MSC-1) and a Leydig cell line (MA-10), but none of the different constructs tested so far was active in COS-7 cells (105). Although the experiments might indicate a cell-specific expression of the FSH receptor gene, which sequences do confer cell specificity have not yet been identified. In transgenic mice carrying a 5-kbp FSH receptor promoter/ß-galactosidase fusion gene, expression of ß-galactosidase transcripts was detected only in the testis and the ovary. Thus, elements within this region are able to direct the expression specifically in testicular and ovarian cell types (105). Analysis of cell-specific transcription factors by electrophoretic mobility shift assays revealed several DNA-protein complexes in cells expressing the FSH receptor gene and an additional specific DNA-protein complex in the nonexpressing COS-7 cell line (105), indicating inhibition of FSH receptor expression in this cell type.
Promoter methylation might be a mechanism involved in the inhibition of gene expression in cell types other than the cells naturally expressing the FSH receptor. Studies using methylation-sensitive enzymes indicated that DNA methylation of the rat promoter is involved in the suppression of transcription in cells lacking detectable FSH receptor mRNA (105). However, the human and mouse FSH receptor promoters contain neither GC-rich islands nor methylation consensus sequences (CCGG), indicating that methylation events do not play a significant role in the modulation of transcriptional activity in these species.
The absence of usual TATA and CCAAT promoter elements, the presence of multiple transcriptional start sites, and the constitutive expression are features of housekeeping genes (112). Similar characteristics are shared by the promoters of the LH and TSH receptor gene (11, 104), but the overall homology between the core-regulatory sequences of the three receptors is low. The highest homology between the FSH and the LH receptor reaches 72% and is confined to a stretch of 58 nucleotides at position -298 to -352 (100). This is not unexpected, however, since physiological studies indicate that the expression of the two gonadotropin receptors is regulated differently.
Assuming a species-independent general mechanism of regulation of the FSH receptor expression, future studies should carefully elucidate common, non-species-specific sequence motifs in the promoter region of the FSH receptor gene. These elements should then be analyzed in vitro, either in primary granulosa and Sertoli cell cultures (keeping in mind the immature status of these cells) or in granulosa or Sertoli cell-derived lines stably expressing a recombinant FSH receptor, to enable FSH stimulation and to mimic, at least partially, the in vivo situation. These experiments are crucial to solving the enigmatic regulation of the FSH receptor expression and would have great impact on the targeting of genes specifically to granulosa and Sertoli cells in transgenic studies.
|
V. Expression of the FSH Receptor and Its Regulation |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
Studies on the hormonal regulation of the FSH receptor gene expression or during different stages of gametogenesis have shown that the transcripts are not differentially regulated. The ratio between alternative transcripts and the full-length mRNA remains constant (114). The presence of different transcripts has also been shown for the LH and the TSH receptor. However, a short LH receptor transcript (1.3 kb), encoding presumably only the extracellular domain, displays a different pattern of expression regulation compared with the other transcripts (115, 116).
The mechanism underlying the generation of different FSH receptor transcripts might be related either to different transcriptional start sites, or different polyadenylation sites, or to alternative splicing processes. When cDNA libraries were screened with FSH receptor cDNA probes or with RT-PCR, some of the transcripts were isolated and further characterized. However, the use of different transcriptional start sites is not the major mechanism responsible for the generation of transcripts ranging in length from 1.3–7 kb. Rather, the generation of different transcripts seems to originate from different polyadenylation sites giving rise to the long form (5–7 kb) and the normal form (2.5 kb) and, in addition, to alternative splicing of these primary transcripts. Analysis of the nucleotide sequences revealed four different possible mechanisms underlying this isoform heterogeneity:
1. Several isoforms lack one or more exons (79, 117, 118, 119). Interestingly, in principle the loss of an exon does not result in changes of the ORF. The isoforms, therefore, encode putative functional receptors. This splicing mechanism is known as cassette-exon-mode and reflects the module-like genomic structure, whereby processes such as insertion or excision of entire exons are enabled by the same exon phasing, leaving the ORF unchanged. This exon shuffling has also been shown for the LH receptor and TSH receptor (120, 121).
2. The second mechanism involves splicing events of the primary transcript through alternative internal 3'-acceptor sites. The presence of conserved splice acceptor sites, e.g., a CAGG nucleotide sequence stretch, can shorten exons if it is located within the exons or lead to incomplete intron splicing (67).
3. Some of the isoforms represent a combination of the cassette-exon-mode and usage of alternative 3'-acceptor sites (68).
4. Another mechanism is represented by the partial retention of intronic sequences. This incomplete splicing generally results in larger transcripts (67, 122, 123).
Splicing events such as described under 2, 3, and 4 in all cases result in a change of the ORF, starting with the branch point. In some cases the new aa sequences show a very basic pattern (122), and the aa sequence is terminated by new termination codons. If the retained introns contain polyadenylation sites, this results in the generation of smaller transcripts. Alternatively, the transcripts are larger than usual.
Apparently alternative splicing processes affect only the extracellular domain of the receptor, encoded by exons 1–9, since no splicing events involving the transmembrane domain encoded by exon 10 have been observed. As most isoforms lack the transmembrane domain whereas the high-affinity hormone-binding site encoded by the extracellular domain is still present, it is speculated that they might give rise to soluble and secretable receptor fragments. These isoforms could potentially act as hormone-binding proteins and thereby antagonize FSH action by sequestering it in the circulation. Similar hormone binding-proteins deriving from the GH receptor (124), and presumably the TSH receptor (121), have indeed been described and result from receptor shedding or alternative splicing events. However, there is no evidence yet of secreted FSH receptor isoforms. Studies on the LH receptor have shown that isoforms lacking the transmembrane domain are able to bind LH with high affinity but are trapped within the cell (125). A short, truncated TSH receptor form might be expressed and secreted, thereby acting as an nonfunctional autoantigen (121). One possible consequence of the simultaneous expression of the different mRNAs, encoding full and truncated receptor forms, might be a competition for the translation process and thereby regulation of the expression of the mature and functional FSH receptor protein (126).
B. Expression of the FSH receptor in the testis
1. Localization. Binding experiments have shown that FSH binds
specifically to receptors located on the membrane of Sertoli cells
(127, 128). No specific binding was observed in spermatogenic cells,
except for an isolated finding suggesting the presence of FSH receptors
in spermatogonia (42). Northern blot hybridization experiments in a
variety of tissues revealed a distinct signal in the testis, solely in
the Sertoli cells (114). In situ hybridization confirmed
that the Sertoli cells are the only cell type expressing the FSH
receptor in the testis (43, 129, 130).
A comprehensive study on FSH receptor expression was performed in a nonhuman primate in which 38 different tissues and organs were screened for the presence of FSH receptor transcripts, using the RNase protection assay technique (131). No transcript could be detected in organs or tissues other than the testis. Thus, unlike the LH receptor, the expression of the FSH receptor seems to be strictly gonad- and cell-specific. Quantification of the FSH receptor mRNA levels in the human and monkey testis showed that 0.05 to 0.1 pg/µg testis RNA encode the FSH receptor (132).
Interestingly, using monoclonal FSH receptor antibodies, Vannier et al. (44) reported a polar expression of the receptor protein at the basal part of the Sertoli cell and around the spermatogonia. The same group had previously reported that the LH receptor protein can be detected in the vascular endothelial cells of the testis by immunostaining (133). The authors proposed a model in which hCG is transported by receptor-mediated transcytosis from the blood vessel through the endothelium cells to the Leydig cells. This model requires the presence of the LH receptor in endothelial cells. In a recent report the FSH receptor was allocated to small vessels in the interstitial space of the testis, implying a similar transport mechanism for FSH (44). Whether these FSH receptor-like structures represent a fully active FSH receptor capable of signal transduction remains to be shown.
2. Ontogeny. Studies of the FSH receptor in the developing testis have demonstrated high-affinity binding for FSH starting from day 28 of gestation in pigs and during the first half and at the end of gestation in primates (134, 135). In the rat, ligand-binding experiments revealed the presence of the FSH receptor from fetal (f) day 17.5 onward. The content of FSH receptor increases between f day 20.5 and birth (136). By Northern blotting and RT-PCR, transcripts encoding the extracellular domain were detected from f day 14.5 onward, and full-length mRNA appeared around f day 16.5. The reason why the extracellular domain can be detected first might be due to differences in the onset of transcription of two mRNA species or, alternatively, the two transcripts may have different half-lives. The FSH receptor is present in the testis before significant concentrations of the cognate hormone appear in the fetal circulation. Furthermore, the fetal testes seem to lack a clear acute cAMP response to FSH despite the presence of the FSH receptor (137). This might be due to an immature signal transduction system or a different FSH receptor coupling in the fetus. From the physiological point of view, this nonresponsiveness might prevent premature activation of FSH-stimulated spermatogenesis (138).
After birth, FSH binding in the mouse testis reaches a peak between days 7 and 21 and then decreases rapidly between days 20 and 37 (139). In the rat, the FSH receptor mRNA increases until day 7, stays constant between days 10 to 20, and drops dramatically around day 40 (140). The initial increase is related both to the increase of receptor numbers per Sertoli cell and to the proliferation of Sertoli cells up to day 10 (141). The subsequent decrease of FSH receptor expression is related to the massive appearance of spermatocytes and spermatids, as indicated by the increasing weight of the testis. The FSH receptor mRNA expression seems to be comparable in adult and immature Sertoli cells (114). The increase of total number and density of FSH-binding sites in the initial phase of testis development has been shown in several mammalian species and in photoperiodic animals (142, 143). Since the receptor increase is parallel to the rise of circulating FSH levels, one might assume that the gonadotropin induces up-regulation of its own receptor in the developmental phase.
3. Regulation of the expression in vivo. In vivo studies have been performed mainly in hypophysectomized animals, thereby depleted of FSH action, followed by substitution with different hormones. The effects of this treatment on FSH receptor binding is different among rats, mice, and photosensitive animals, such as quails and the Djungarian hamster. In adult quails, FSH binding decreases remarkably in the absence of FSH, suggesting a mechanism of receptor up-regulation by FSH, whereas in mice and rats FSH deprivation leads to an increase in FSH binding, indicating instead a down-regulatory action of the hormone in these species (60). This down-regulation has recently been confirmed in the rat by Northern blot experiments (144, 145). In photoperiodic animals, the day length has a marked effect on the expression pattern of the FSH receptor (142, 143). Animals transferred from short-day to long-day conditions display rapid testicular growth sustained by a pronounced increase in FSH-binding sites induced by the elevation in gonadotropin levels. Since it is assumed that the number of Sertoli cells per testis remains constant, the number of FSH receptors per Sertoli cell must increase. The only study in humans so far was performed in transsexual men treated with estrogens for sex reversal over long time periods. High-dose estrogen treatment leads to a marked decrease in FSH receptor mRNA levels (132).
In the rat, spermatogenesis is organized into 14 stages, defined by
their different germ cell composition and present simultaneously in
different regions along the seminiferous tubules. Dissection of
segments containing individual stages or synchronization of
spermatogenesis by retinol deprivation and repletion enabled the
investigation of FSH receptor expression during the spermatogenetic
cycle. FSH binding and FSH receptor mRNA expression studies reach the
highest levels in stages XIII, XIV, and I, whereas lowest expression is
found in stages VII and VIII (108, 114, 128, 146) (Fig. 5). This stage-dependent expression of
the FSH receptor coincides with different maturation states of the germ
cells. The more advanced germ cells, e.g., those in stage VI
and VII when spermiation occurs, colocalize with reduced FSH receptor
expression, whereas stages containing less advanced germ cells, such as
early spermatids, are colocalized with an increased expression.
FSH-stimulated cAMP production in isolated sections of seminiferous
tubules is highest in stages II to IV (Fig. 5), indicating further
regulation of signal transduction by local factors (128, 147). Overall,
these data indicate that a stage-specific paracrine interaction between
spermatogenic cells and Sertoli cells regulates FSH receptor expression
(148).
|
C. Expression of the FSH receptor in the ovary
1. Localization. In the female, FSH binding has been localized
to the granulosa cells (149, 150). Several recent studies using
molecular biology techniques confirmed that the granulosa cells are the
only cell type expressing the FSH receptor (119, 151, 152, 153). Thus, as in
the male, the expression of the FSH receptor in the female is strictly
gonad- and highly cell-specific. This finding is in contrast to the
expression pattern observed for the LH receptor and TSH receptor. LH
receptor expression can be demonstrated in a variety of organs and
tissues (154), and the TSH receptor expression has also been shown in
extra/retro-orbital tissue (121), suggesting hitherto unknown or only
suspected physiological functions of LH and TSH in other tissues.
Similarly, a recent report indicated the presence of FSH receptor
protein and mRNA in cultures of human myometrial smooth muscle cells
(155). However, this isolated finding awaits further confirmation.
2. Ontogeny. The acquisition of FSH receptors is essential for granulosa cell differentiation and for follicle maturation (156). In the fetal rat ovary, expression of the extracellular domain of the FSH receptor is first detected on day 20.5. Full-length transcripts appear later, on day 1 post partum and more clearly from day 5 onward in the rat and mouse (130, 157). Similar to the ontogeny of the FSH receptor expression in the testis, this sequential appearance of short and full-length transcripts might reflect differences in mRNA half-lives and/or differences in the onset of transcription of the two RNA species. High-affinity binding sites are present on granulosa cells from day 3 onward, and a constant increase is observed until day 21 when the expression reaches a plateau (158, 159). In general, there is a strong parallelism in the developmental changes in FSH receptor mRNA and receptor protein (158, 159). The ovary does not respond to FSH between birth and day 3, whereas from day 4 to 7 ovaries show an acute, FSH-sensitive cAMP response coincident with the appearance of full-length FSH receptor mRNA (130, 149, 158).
3. Regulation of expression in vivo. In the immature rat ovary FSH receptor mRNA can already be localized in the granulosa cells of small follicles. Treatment with PMSG to stimulate follicle growth results in a marked increase of FSH receptor mRNA expression and FSH-binding sites, whereas subsequent administration of hCG to induce ovulation and luteinization significantly decreases FSH receptor expression (160, 161). In hypophysectomized, estrogen-treated rats, follicular development and ovulation can be induced by FSH alone (162), which increases both FSH binding and FSH receptor mRNA levels. In contrast, induction of ovulation with a surge dose of recombinant FSH suppresses FSH binding and FSH receptor gene expression. These data suggest a biphasic, homologous regulation of FSH receptor expression in the ovary. Low doses of FSH increase the number of FSH-binding sites parallel to the increase of FSH receptor mRNA levels. High doses of FSH down-regulate FSH receptor-binding sites and mRNA levels, suggesting a suppression of gene expression and protein synthesis concomitant to the increased receptor occupancy and internalization (116, 160). This biphasic mechanism might be due either to changing mRNA stability in the presence of different hormone concentrations or to effects on the regulatory elements in the promoter region of the FSH receptor gene. However, analysis of the FSH receptor promoter does not give any further clues, since no obvious regulatory elements can be detected therein (see Section IV). Whether the recently identified transcription factor ICER (110) can interact with promoter regions of the FSH receptor, and thereby be involved in the biphasic regulation, remains to be shown.
Examination of sexually mature adult rats during the 4-day estrous
cycle revealed the presence of FSH receptor in nearly all follicles
starting with only one layer of granulosa cells (151). The levels seem
to rise during follicular maturation, although some reports claim that
the steady state FSH receptor levels in healthy follicles do not
correlate with follicular size (163, 164, 165, 166), and decrease drastically in
the post-ovulatory follicle, after the LH surge (Fig. 6). With luteinization, FSH binding can
no longer be detected. In the bovine, full-length FSH receptor
transcripts are still detectable 1 day after luteinization, albeit at
low levels. By day 3, full-length transcripts are no longer detectable
but, surprisingly, expression of the extracellular domain persists
(117). This specific pattern of loss of FSH receptor gene expression
resembles, in reverse, the onset of expression in the ovary, again
suggesting that the two transcripts might have a different half-life
and/or be produced in a regulated succession. Follicular atresia is
associated with decreased responsiveness to FSH and reduced mRNA
receptor levels, due to a transcriptional down-regulation or decreased
stability of receptor mRNA (119, 165, 167). Although FSH receptor
expression has long been recognized to be under the control of FSH, a
recent study in hypogonadal mice lacking circulating gonadotropins
revealed the presence of FSH receptor mRNA in the ovary. Thus, factors
other than FSH may act on the induction of FSH receptor expression
(156).
|
4. Regulation of expression in vitro. Granulosa cells of immature, estrogen-treated rats contain FSH receptors that decline during culture. Treatment with FSH maintains the expression, suggesting that FSH increases the levels of its own receptor. Estrogens synergize with FSH in vitro to increase the number of receptors per granulosa cells, but alone do not alter the expression (152). The FSH-related increase of FSH receptor expression is dose-dependent and can be mimicked by the adenylyl cyclase activator forskolin and by cholera toxin, indicating that the gonadotropin can amplify its own action on granulosa cell differentiation and maturation. The suppression of FSH receptor reported in vivo by an ovulatory dose of FSH (116), however, is not observed in cultured granulosa cells. This discrepancy might be due to the absence in vitro of paracrine factors involved in the regulation of the FSH receptor expression in vivo or might be merely related to the immature status of the granulosa cells.
Paracrine factors are involved in the regulation of FSH receptor expression. Treatment of granulosa cells with epidermal growth factor, basic fibroblast growth factor, or insulin-like growth factor-1 attenuates the response to FSH but does not alter basal levels of expression, whereas GnRH completely suppresses the induction of FSH receptor mRNA by FSH (152). Other growth factors, such as transforming growth factor-ß and activin, are potent inducers of FSH receptor expression (159, 171, 172, 173). In the presence of FSH, activin has a biphasic action, which is inhibitory at low doses and stimulatory at high doses (174). The mechanism whereby transforming growth factor-ß and activin increase FSH receptor expression is still not clear. By acting via tyrosine kinase receptors, they do not directly increase intracellular cAMP accumulation, and it is therefore reasonable to assume two distinct pathways of FSH receptor induction.
In porcine granulosa cells not previously exposed to estrogens, the FSH receptor increases with time in the absence of FSH. FSH causes a dose-dependent decrease in FSH receptor binding, while stimulating FSH receptor mRNA levels (113). This differential regulation of protein and mRNA levels would not be mediated by cAMP, since cholera toxin increases FSH receptor both at the protein and mRNA level (166). However, the decrease of FSH binding might be due to the blocking of binding by unlabeled FSH and/or the FSH receptor is internalized and degraded as a consequence of FSH binding.
Future studies should clarify the biphasic mRNA expression in the estrous cycle. The identification of regulatory elements in the promoter region of the FSH receptor gene is therefore necessary. Paracrine factors and/or intracellular repressors involved in stage-specific receptor expression during the spermatogenic cycle should be identified in the male. In particular, it will be interesting to analyze the receptor expression in male infertility, especially considering that mutations of the FSH receptor are obviously rare in this condition (see Section X). To this end, sensitive detection methods capable of quantifying expression in biopsy material must be developed.
|
VI. Expression of the FSH Receptor in Cell Lines |
|---|
|
TopAbstractI. IntroductionII. Biochemical Properties of...III. Molecular Structure of...IV. The FSH Receptor...V. Expression of the...VI. Expression of the...VII. Structure-Function...VIII. Signal Transduction and...IX. Inhibitors and Modulators...X. Naturally Occurring Mutations...XI. ConclusionsReferences |
|---|
, including Refs.
175–183). Among the cloned FSH receptors from other species, the ovine
and the porcine receptor were expressed as well (72, 73).
|
summarizes the data obtained from the expression of the
rat and human FSH receptor in different cell types. With the exception
of one line obtained from immortalized granulosa cells (181), the cells
used do not derive from progenitors normally expressing the FSH
receptor, but possess functional Gs to which the
recombinant receptor couples.
In vitro expression is obtained by transfection of an
expanding cell population with a suitable vector where the receptor
cDNA is placed under the control of a strong promoter. As shown in
Table 2, the degree of expression, in terms of number of receptors per
cell, varies consistently between the different lines, and this
variability seems to be relatively independent of the type of promoter
driving the transcription. A factor potentially limiting the expression
of the FSH receptor cDNA is the above mentioned presence of several
stop codons in the 5'-untranslated sequence immediately preceding the
translational start site. The modification of this tract of sequence
(183) or the insertion therein of artificial intronic sequences (67)
has been a useful strategy for improving receptor expression.
In all the cell lines produced, the FSH receptor is coupled to
Gs and adenylyl cyclase, and the exposure of the
recombinant cells to FSH leads to a saturable, dose-dependent cAMP
production. As shown in Table 2, the ED50 of FSH-dependent
cAMP accumulation varies impressively between the various cell lines
and is little related to differences in receptor density. Since the
direct comparison of the ED50 values is hampered by the use
of different experimental systems, i.e., cells, constructs,
and FSH preparations (184, 185), the reasons for this variability are
not clear. Future experiments should explore the possibility that part
of the observed differences might be due to a different coupling
efficiency to adenylyl cyclase in the different cell lines. The
KD values obtained from binding studies with recombinant
receptors are usually in the nanomolar range, yet 10- to 1000-fold
higher than those found with membrane preparations of native rat and
calf receptors (56). However, FSH receptor binding is known to be
influenced by many factors, including temperature, ionic strength (186, 187), and time (188), and the direct comparison of the properties of
native and recombinant receptors under the same experimental conditions
has not yet been performed.
The first expression studies of the human FSH receptor cDNA suggested that there might be a certain degree of species specificity in the hormone-receptor interaction. In fact, the human recombinant receptor transiently transfected in 293 cells was shown to interact more effectively with human FSH compared with rat FSH and, even more, to FSH from other species. In contrast, both human and rat FSH showed similar affinity to rat receptors obtained from a crude preparation of testicular membranes, with the FSH preparations from other species having lower affinity (78). However, the species specificity of the two receptors was shown to be similar when both rat and human FSH receptor were of recombinant origin (176). Thus, the issue of species specificity is not clear, while the experimental approach used in these studies may not be completely appropriate. In
